Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Play to Win - counter-intuitive example

11 views
Skip to first unread message

The Name Forgotten

unread,
Feb 18, 2009, 8:08:43 AM2/18/09
to
As I mentioned in another post - this post will be about an example
where "Play-to-win" will lead you to worse results by following the
rule as it is written, then by not following the rule. To do this, I'm
going to use a thought experiment.

The situation arises when someone is placed in the corner-case
position where they have a certainity of getting 2 VP's over a very
minor chance of getting 3 VP's (i.e. a game win) or ending with only 1
VP. Most people would play it safe and go for the 2 VP's, because it's
a more strategic option in a tournament situation. However, Play-to-
win would force you to take the unlikely chance of getting a GW
instead of the certainty of the extra VP.

Let's imagine you are in this situation (you being player A)

Player A has 2 ready, untapped vampires, both on 1 blood, and with the
vote majority. It is your turn and you have 2 pool.
Player B has no ready minions and is on 3 pool.
Player C has 5 ready imbued, each with second sights and strike with
conviction, 2 are untapped and another 2 have laptops, and an
Unmasking in play. C is on 1 pool and has already ousted D (who ousted
E).

Player A is bleeding B who is bleeding C.

You have one stealth modifier in your hand at the moment (a forgotten
labyrinth), and you know you still have some in your deck, although it
is highly unlikely that you'll cycle into them by only playing a card
or two (let's say under 10% chance). You are not playing any intercept
cards. You do, however, have multiple Kine Resources Contested, and
all the votes you need to make it pass.

Your best bet of getting a Game Win and/or VPs is by calling a KRC
(unless someone can see a better way), which you successfully do with
the use of your forgotten labyrinth, and replace with cards that are
not stealth.

You now have the option of ousting any player on the table. In fact,
you could oust 2 players in one go.

You feel that it is highly unlikely (under 5% chance) that, should you
oust B, you will be able to call your second KRC and get it to pass
successfully. However, if C is not ousted by the end of your turn, C
will likely oust you and get the GW.

Your choice is to:

Option A - oust C now, along with B - leaving you with 2 VP's for
ousting B and surviving - which leaves you with the most VP's on the
table.

OR:

Option B - Take one point of damage yourself, leaving you with one
vampire left to act, a highly unlikely chance of ousting C (only 5% or
less chance of replacing with enough stealth), and a far greater
likelihood of being ousted by C, thus giving him the game win and the
most VP's.

Play-to-win would force you to choose Option B, because you still are
able to get the stealth you need, even though it is unlikely. This
seems to go against common-sense strategic play, which would be to
oust both of them in one go and end up with more VPs than anyone else.

I realise that the 'play-to-win' means going for GW's or VP's if it's
"reasonably possible", but with self-ousting examples "reasonably
possible" usually means "any chance of", which, if applied to this
situation, leads you towards making a poor decision. If, on the other
hand, 5% is not a reasonable chance, then you open Pandora's box
towards what is reasonable in a self-ousting situation.

Which brings me onto the same conclusion that Play-To-Win is a
problematic rule that should be discarded. I think if a judge was
called over, and had to enforce such a rule in the above example, it
would be very bad for the game, and certainly seems to run contrary to
common sense or good play.

Thoughts?

henrik

unread,
Feb 18, 2009, 8:21:01 AM2/18/09
to
On Feb 18, 2:08 pm, The Name Forgotten <tex...@telkomsa.net> wrote:

> The situation arises when someone is placed in the corner-case
> position where they have a certainity of getting 2 VP's over a very
> minor chance of getting 3 VP's (i.e. a game win) or ending with only 1
> VP. Most people would play it safe and go for the 2 VP's, because it's
> a more strategic option in a tournament situation. However, Play-to-
> win would force you to take the unlikely chance of getting a GW
> instead of the certainty of the extra VP.

> Thoughts?

I think you're interpreting the rule wrong (or at least bad).
The rules say that you should play for the game win if it's
"reasonably possible", not that you should play for the game win if
it's a very high risk that it makes you ousted with fewer victory
points than a safer option.
It is, of course, a judgement call (for the players at first, and for
judges if they get involved).


4.8. Play to Win
[...] For tournaments, playing to win means playing to get a Game Win
if it is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible.

http://www.white-wolf.com/vtes/index.php?line=veknRules

LSJ

unread,
Feb 18, 2009, 8:30:11 AM2/18/09
to
henrik wrote:
> On Feb 18, 2:08 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
>
>> The situation arises when someone is placed in the corner-case
>> position where they have a certainity of getting 2 VP's over a very
>> minor chance of getting 3 VP's (i.e. a game win) or ending with only 1
>> VP. Most people would play it safe and go for the 2 VP's, because it's
>> a more strategic option in a tournament situation. However, Play-to-
>> win would force you to take the unlikely chance of getting a GW
>> instead of the certainty of the extra VP.
>
>> Thoughts?
>
> I think you're interpreting the rule wrong (or at least bad).
> The rules say that you should play for the game win if it's
> "reasonably possible", not that you should play for the game win if
> it's a very high risk that it makes you ousted with fewer victory
> points than a safer option.

Correct.

Risk analysis is allowed; it is even reasonable.

James Coupe

unread,
Feb 18, 2009, 8:36:39 AM2/18/09
to
The Name Forgotten <tex...@telkomsa.net> wrote:
>As I mentioned in another post - this post will be about an example
>where "Play-to-win" will lead you to worse results by following the
>rule as it is written, then by not following the rule.

No, it won't.

>The situation arises when someone is placed in the corner-case
>position where they have a certainity of getting 2 VP's over a very
>minor chance of getting 3 VP's (i.e. a game win) or ending with only 1
>VP. Most people would play it safe and go for the 2 VP's, because it's
>a more strategic option in a tournament situation.

This is absolutely within the Play-to-Win rules.

Risk analysis and the consequences of failure are both part of the
judgment call required for PTW.

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/407887c25118873f

Players aren't forced to play for very minor chances that may blow up in
their face. During the final, the ordinary 'this might blow up in my
face' considerations don't apply in the same way (because the worst you
can do is come second), although you're not - of course - forced to play
for a GW now (which will lead to you exiting the game if you fail) when
you could bide your time and get the GW later.

>However, Play-to-
>win would force you to take the unlikely chance of getting a GW
>instead of the certainty of the extra VP.

No it wouldn't.

>Which brings me onto the same conclusion that Play-To-Win is a
>problematic rule that should be discarded.

Discarding a rule based on a flawed understanding of it seems unhelpful.

Things might be better if you understood how the rule worked. :-(

>I think if a judge was
>called over, and had to enforce such a rule in the above example, it
>would be very bad for the game, and certainly seems to run contrary to
>common sense or good play.

If a judge came over and enforced your twisted misunderstanding of the
rules, that would be bad, yes.

--
James Coupe
PGP Key: 0x5D623D5D YOU ARE IN ERROR.
EBD690ECD7A1FB457CA2 NO-ONE IS SCREAMING.
13D7E668C3695D623D5D THANK YOU FOR YOUR COOPERATION.

The Name Forgotten

unread,
Feb 18, 2009, 9:10:50 AM2/18/09
to

Ok, then my question is, what is "reasonable"? (or how should
judgements of reason be made?)

It seems like everybody missed the last part of my argument - whatever
line you draw for reasonable in this situation (whether it be a 5%,
50% or 99.9% chance of drawing the necessary stealth), the same would
have to apply to the "self-ousting" situation - i.e. when does a
person have a reasonable expectation to not get any VP's at all and be
free to play as he chooses? When he only has a 5% chance of getting no
VP's? A 50% chance?

So far, the general 'ruling' I've gotten on the self-ousting part is
that you only need to have a reasonable chance of getting a VP in
order to stop you from self-ousting or doing other crazy stuff - with
'reasonable' being closer to 'even a small chance in hell' vs 'a good
likely chance'.

Having a double standard of what reasonable is seems problematic to me
(i.e. 5% chance of getting a VP as opposed to 5% chance of getting a
GW). It looks to me like a Pandora's box situation with the way that
the rules are currently written. The moment you give too much leeway
in allowing a person to take 2 VP's instead of the potential of 3, you
are also giving that same amount of leeway to a person who throws up
his hands and says he has no chance of getting a VP with his S&B
kindred spirits deck because his prey has 5 intercept on the table, so
now he'll bleed out his predator instead (who stopped him from
bleeding out his prey earlier in the game), or some similar situation.

Of course, I will admit that there may be another standard to base
this decision on, but risk analysis, as LSJ talks about, seems to lead
to a double standard.

LSJ

unread,
Feb 18, 2009, 9:15:49 AM2/18/09
to
The Name Forgotten wrote:
> Of course, I will admit that there may be another standard to base
> this decision on, but risk analysis, as LSJ talks about, seems to lead
> to a double standard.

What double standard?

"Reasonable" is used in both cases, and used in the same way in both cases.

And, since actual probabilities cannot be computed, they aren't used (in either
case).

The Name Forgotten

unread,
Feb 18, 2009, 9:17:25 AM2/18/09
to
On Feb 18, 3:30 pm, LSJ <vtes...@white-wolf.com> wrote:
> henrik wrote:
> > On Feb 18, 2:08 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> >> The situation arises when someone is placed in the corner-case
> >> position where they have a certainity of getting 2 VP's over a very
> >> minor chance of getting 3 VP's (i.e. a game win) or ending with only 1
> >> VP. Most people would play it safe and go for the 2 VP's, because it's
> >> a more strategic option in a tournament situation. However, Play-to-
> >> win would force you to take the unlikely chance of getting a GW
> >> instead of the certainty of the extra VP.
>
>
>
> Risk analysis is allowed; it is even reasonable.
>

I agree with this point. The point I'm trying to make is that saying
this seems to lead to either (a) a double standard and a lack of
consistency in the rules or (b) it opens up too much leeway for
players to give up or do things that aren't in their best interests
because they don't have a 'reasonable' chance of getting a VP.

LSJ

unread,
Feb 18, 2009, 9:18:44 AM2/18/09
to

It seems not to do so, to me.

But then, I'm referring to cases where a judge is available and can apply judgment.

henrik

unread,
Feb 18, 2009, 9:21:31 AM2/18/09
to
On Feb 18, 3:10 pm, The Name Forgotten <tex...@telkomsa.net> wrote:

> Ok, then my question is, what is "reasonable"?

It's dependant on the situation. I don't think it's possible to put up
a universal rule (except for the one used atm).
If you're unable to get more victory points, you're allowed to
selfoust. Whether you're unable to get more victory points or not
depends on a lot of stuff:

Your deck, minions, hand, pool etc
Your opponents decks, minions, hands, pool etc

> (or how should
> judgements of reason be made?)

Look at the game situation and then think?

The Name Forgotten

unread,
Feb 18, 2009, 9:24:24 AM2/18/09
to
On Feb 18, 3:36 pm, James Coupe <ja...@zephyr.org.uk> wrote:
> The Name Forgotten <tex...@telkomsa.net> wrote:
>
> >Which brings me onto the same conclusion that Play-To-Win is a
> >problematic rule that should be discarded.
>
> Discarding a rule based on a flawed understanding of it seems unhelpful.
>
> Things might be better if you understood how the rule worked. :-(

I do, but, as I've said, to follow common sense and explanations on
how the rule worked would lead to a double standard. Things might also
be better if you understood my point 8) The rule should at least be
discarded in it's current state for this reason alone (i.e. leading to
a double standard).

> >I think if a judge was
> >called over, and had to enforce such a rule in the above example, it
> >would be very bad for the game, and certainly seems to run contrary to
> >common sense or good play.
>
> If a judge came over and enforced your twisted misunderstanding of the
> rules, that would be bad, yes.
>

Please untwist my misunderstanding, and explain to me how the 'double-
standard' issue is overcome.

I don't think that Play-to-win should effect my example, and, as per
the post you linked to:

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/407887c25118873f

However, it does seem to create a double standard - of allowing a lot
of leeway for someone choosing VP's over a chance of a GW, as opposed
to someone choosing to self oust instead of a chance at a VP, with
absolutely no foundation - the risk analysis provides the same chances
and the same percentage, but in one case it is argued as reasonable,
and in the other case it is argued as unreasonable.

The Name Forgotten

unread,
Feb 18, 2009, 9:40:03 AM2/18/09
to
> But then, I'm referring to cases where a judge is available and can apply judgment.- Hide quoted text -
>
> - Show quoted text -

Yes, and my question is: how is the judge expected to apply his
judgment - i.e. on what grounds.

Let's say, for example, I'm a judge in a tournament with 2 tables. On
the first table is the situation I described above. On the second
table is a situation where someone is purposefully bleeding backwards,
and refusing to bleed his prey because, he feels, he has no way of
ousting his prey.

Let's further say that I have all the time in the world to examine the
decks and do a complete analysis of people's odds in the game. Some
outside influences, surely, I cannot account for, but those outside
situations could happen at either table, so they are irrelevant for
the purposes of this argument.

Now, in Table A - the person has a 60% chance of drawing the stealth
card he needs by playing his action card to get a GW, and will be
ousted in the following turn if he does nothing (as opposed to
choosing the safer 2VP route).
In table B - the person has a 60% chance of drawing the stealth card
he needs by playing his action card to get a VP, and will be ousted in
the following turn if he does nothing (as opposed to bleeding his
predator).

From what I've understood from other posts - in Table A, it's still
acceptable to 'play it safe' and take the 2 VP route instead of a GW -
however, in Table B, giving up when you still have such a good chance
of earning a VP is not acceptable. If you are referring to risk
analysis as your primary method of making such a judgement - then
there certainly seems to be a double standard.

LSJ

unread,
Feb 18, 2009, 10:34:30 AM2/18/09
to
The Name Forgotten wrote:
> On Feb 18, 4:18 pm, LSJ <vtes...@white-wolf.com> wrote:
>> But then, I'm referring to cases where a judge is available and can apply judgment.- Hide quoted text -
>>
>> - Show quoted text -
>
> Yes, and my question is: how is the judge expected to apply his
> judgment - i.e. on what grounds.

On the grounds of his or her judgment. (And idea of reasonableness).

What I'm pretty sure this game doesn't need is a bunch of lawyers and tort law
and abridged encyclopedias of binding precedents brought to bear on this issue.

What we've got now may not be perfect, but it's better than the alternative.

The Name Forgotten

unread,
Feb 18, 2009, 11:19:30 AM2/18/09
to
On Feb 18, 5:34 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 18, 4:18 pm, LSJ <vtes...@white-wolf.com> wrote:
> >> But then, I'm referring to cases where a judge is available and can apply judgment.- Hide quoted text -
>
> >> - Show quoted text -
>
> > Yes, and my question is: how is the judge expected to apply his
> > judgment - i.e. on what grounds.
>
> On the grounds of his or her judgment. (And idea of reasonableness).
>
> What I'm pretty sure this game doesn't need is a bunch of lawyers and tort law
> and abridged encyclopedias of binding precedents brought to bear on this issue.

Which is why I'm suggesting we scrap the rule, or at least amend it to
close or accept the double standard. By this I specifically mean the
part quoted above - that being:

4.8. Play to Win
[...] For tournaments, playing to win means playing to get a Game Win
if it is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible.

Or, at the very least, scrapping the Game Win scenario and leaving it
just as playing to get as many VP's as possible will close that
loophole corner-case scenario. I have no problem with the rest of the
Play-to-Win ruling.

>
> What we've got now may not be perfect, but it's better than the alternative.

That's not really a useful argument. It's the same sort of argument
that someone might use to defend the use of fossil fuels (or the
Spanish Inquisition.)

My question still remains: Is it? I'm trying to make the point that
the alternative is better, even though, it too, is not perfect.
Removing the rule would stop judges from having the power to tell
people what strategy they must play (or risk disqualification or other
penalties), as well as stopping them from having that sort of
responsibility where they might be called to make such judgements.

How is it in any way better than not having the rule? I sincerely
doubt that the rule is preventing anyone from playing how they please
in a tournament situation (in any way that differs from what the
standard rules say about the matter anyhow). Forcing people to chase
after potential VP's just opens up all sorts of problems, as can be
witnessed by the amount of complaining and arguing that goes on in
this and other forums.

henrik

unread,
Feb 18, 2009, 11:33:52 AM2/18/09
to
On Feb 18, 5:19 pm, The Name Forgotten <tex...@telkomsa.net> wrote:

> Which is why I'm suggesting we scrap the rule, or at least amend it to
> close or accept the double standard. By this I specifically mean the
> part quoted above - that being:
>
> 4.8. Play to Win
> [...] For tournaments, playing to win means playing to get a Game Win
> if it is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible.
>
> Or, at the very least, scrapping the Game Win scenario and leaving it
> just as playing to get as many VP's as possible will close that
> loophole corner-case scenario. I have no problem with the rest of the
> Play-to-Win ruling.

Changing it like that will open up new loophole corner-case scenarios
though.

Matthew T. Morgan

unread,
Feb 18, 2009, 11:35:13 AM2/18/09
to
On Wed, 18 Feb 2009, The Name Forgotten wrote:

> Your choice is to:
>
> Option A - oust C now, along with B - leaving you with 2 VP's for
> ousting B and surviving - which leaves you with the most VP's on the
> table.

In addition to all the points made by other people, getting 2 VPs and more
VPs than anyone else at the table constitutes a game win. This absolutely
is the correct choice and there are no rules of any variety that would
prohibit this option.

> OR:

[make up rules that make no sense and then complain about them]

Matt Morgan

The Name Forgotten

unread,
Feb 18, 2009, 1:16:15 PM2/18/09
to
On Feb 18, 6:35 pm, "Matthew T. Morgan" <farq...@io.com> wrote:
> On Wed, 18 Feb 2009, The Name Forgotten wrote:
> > Your choice is to:
>
> > Option A - oust C now, along with B - leaving you with 2 VP's for
> > ousting B and surviving - which leaves you with the most VP's on the
> > table.
>
> In addition to all the points made by other people, getting 2 VPs and more
> VPs than anyone else at the table constitutes a game win.  This absolutely
> is the correct choice and there are no rules of any variety that would
> prohibit this option.

Point taken. For some reason I thought you needed more than 2 VPs. My
argument can still be true with one little adjustment, in that case -
that being that another player has 2 VPs.

I.e. Player A bleeding Player B (who has already ousted player D)
bleeding Player C (who has already ousted player E).

That way, if you oust both, player B has 2 VPs to tie with you, and if
you oust only B, C will get 3 VPs.

> > OR:
>
> [make up rules that make no sense and then complain about them]

Not a helpful comment. Cut-and-paste my argument and responses with
the new example. You're intelligent enough to realise that - don't
waste my time or yours.

The point still remains, the double standard still seems to exist -
whatever cut off point you have for chasing after the possibility of a
GW, risking VP's to get there, as opposed to the certainty of getting
VP's, you'd need to apply the same cut off point to when a person can
rightfully give up on a game and self-oust (or do some other crazy
action) - or risk maintaining a double standard.

Personally, I think almost all players (including myself) would think
that this double standard does exist - but it certainly brings into
question what it means for a person to reasonably expect that they
will not getting a VP - and give them far greater leeway than many
players would be happy with.

The Name Forgotten

unread,
Feb 18, 2009, 1:48:43 PM2/18/09
to

Such as? Would it create the double standard that this ruling seems to
do? Would it create controversy? I'm genuinely curious.

John Flournoy

unread,
Feb 18, 2009, 1:56:24 PM2/18/09
to
On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
> My question still remains: Is it? I'm trying to make the point that
> the alternative is better, even though, it too, is not perfect.
> Removing the rule would stop judges from having the power to tell
> people what strategy they must play (or risk disqualification or other
> penalties), as well as stopping them from having that sort of
> responsibility where they might be called to make such judgements.

Actually, it still leaves the judges in the same position - only now
they'd be having to go 'you must play this strategy here, because it
might give you more VPs, even if an alternate strategy will insure you
a game win but less VPs.'

The point of the game is to win. Saying 'you must try to maximize your
VPs, even if this means possibly not winning', is a sub-optimal rule.

(I'll give a scenario where going for more VPs might cost you the game
in a different post.)

> How is it in any way better than not having the rule? I sincerely
> doubt that the rule is preventing anyone from playing how they please
> in a tournament situation (in any way that differs from what the
> standard rules say about the matter anyhow). Forcing people to chase
> after potential VP's just opens up all sorts of problems, as can be
> witnessed by the amount of complaining and arguing that goes on in
> this and other forums.

It's better than not having the rule, because it allows a judge to
sanction somebody who is clearly not playing to win, such as the
following scenario:

Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
tournament.

Without "play to win", Player B can play to give his buddy (player A)
the game win in the hopes that A will make the finals (which B won't
be able to do no matter what.)

-John Flournoy

henrik

unread,
Feb 18, 2009, 1:58:20 PM2/18/09
to
On Feb 18, 7:48 pm, The Name Forgotten <tex...@telkomsa.net> wrote:

> > > Or, at the very least, scrapping the Game Win scenario and leaving it
> > > just as playing to get as many VP's as possible will close that
> > > loophole corner-case scenario. I have no problem with the rest of the
> > > Play-to-Win ruling.
>
> > Changing it like that will open up new loophole corner-case scenarios
> > though.
>
> Such as? Would it create the double standard that this ruling seems to
> do? Would it create controversy? I'm genuinely curious.

Such as situations where you're forced to a 2-2-1 split instead of
giving your prey a victory point and walk out with a 2-1-1-1 gamewin.
Right now, you can deal away victory points to other players, which I
think is a good thing. That won't be possible (or at least much
harder) with your proposed rule.

John Flournoy

unread,
Feb 18, 2009, 2:07:27 PM2/18/09
to
> > Changing it like that will open up new loophole corner-case scenarios
> > though.
>
> Such as? Would it create the double standard that this ruling seems to
> do? Would it create controversy? I'm genuinely curious.

Here's an a corner-case scenario.

Three players. Each has one pool. A bleeding B bleeding C.

Player A has a single minion. Player B has a single minion, tapped.
Player C has no minions.

Hero (Player A) has a KRC (with vote lock) in his hand. He already has
one VP. Player B also has one VP.

B has one card in his hand, which may or may not be a Wake. No other
relevant cards exist.

If the rule becomes 'you must play to maximize your VP and game win is
not a consideration', then player A would be prohibited from playing
his KRC - if he splits it between B and C, he will pass it, both will
be ousted, and the VP split will end up A3/B2.

If he instead tries to bleed B out, one of two things will happen:

1) B will not have a wake, and A will get all the remaining VP in
short order (4-1 split.)

2) B will have a wake, and block, and then B will oust C (at best a
3-2 split, and possibly losing 2-3 if B then ousts A).

By "you must try to maximize your VP", A _cannot_ play to take a
guaranteed win, but must make the bleed - even though that carries a
risk of losing the game.

And that seems (to me at least) like a rule that isn't clearly better
than what currently exists.

(By extension, there are lots of other scenarios where you might be
prohibited from ousting someone cross-table with a vote, because that
would reduce the maximum number of VPs you could have a reasonable
chance at getting.)

-John Flournoy

LSJ

unread,
Feb 18, 2009, 2:20:57 PM2/18/09
to
The Name Forgotten wrote:
> On Feb 18, 5:34 pm, LSJ <vtes...@white-wolf.com> wrote:
>> The Name Forgotten wrote:
>>> On Feb 18, 4:18 pm, LSJ <vtes...@white-wolf.com> wrote:
>>>> But then, I'm referring to cases where a judge is available and can apply judgment.- Hide quoted text -
>>>> - Show quoted text -
>>> Yes, and my question is: how is the judge expected to apply his
>>> judgment - i.e. on what grounds.
>> On the grounds of his or her judgment. (And idea of reasonableness).
>>
>> What I'm pretty sure this game doesn't need is a bunch of lawyers and tort law
>> and abridged encyclopedias of binding precedents brought to bear on this issue.
>
> Which is why I'm suggesting we scrap the rule, or at least amend it to

? Uh, that's the opposite to the "which" above.
The "which" above is what the game currently uses: reasonable judgment.

> close or accept the double standard. By this I specifically mean the
> part quoted above - that being:

You keep saying "double standard", but you don't actually say what the double
standard is. Reasonable judgment is used uniformly, not in a double standard
kind of way.

> 4.8. Play to Win
> [...] For tournaments, playing to win means playing to get a Game Win
> if it is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible.
>
> Or, at the very least, scrapping the Game Win scenario and leaving it
> just as playing to get as many VP's as possible will close that
> loophole corner-case scenario. I have no problem with the rest of the
> Play-to-Win ruling.

That would force players to sometime abandon a GW in favor of an additional VP.
The opposite of what "play to win" actually means.

2-1-1-1-0 is better (a GW) than 2.5-2.5-0-0-0 (no GW).

The obvious part (the part you've strangely done away with) is: play to win =
play to get the game win.

The secondary part - what to do when you can't get the game win - is a bit more
open to tinkering, but I think that the current (maximize VPs) is better than
the alternatives.

>> What we've got now may not be perfect, but it's better than the alternative.
>
> That's not really a useful argument.

Um.

Hmm. Pointing out that your suggestion to change what we've got to something
less good isn't useful?

> It's the same sort of argument
> that someone might use to defend the use of fossil fuels (or the
> Spanish Inquisition.)

No. Unless you're asserting there are no better alternatives to fossil fuels.

> My question still remains: Is it? I'm trying to make the point that
> the alternative is better, even though, it too, is not perfect.

Ah. That's exactly the same utility of the argument above. Then it's just a
matter of substance. If an alternative is shown to be better, then that's great.

So far, none are.

> Removing the rule would stop judges from having the power to tell
> people what strategy they must play (or risk disqualification or other
> penalties),

How so? You're proposed alternative has just the same amount of that.

> as well as stopping them from having that sort of
> responsibility where they might be called to make such judgements.

Again, not true. The proposed alternative doesn't lessen that.

> How is it in any way better than not having the rule? I sincerely
> doubt that the rule is preventing anyone from playing how they please
> in a tournament situation (in any way that differs from what the
> standard rules say about the matter anyhow). Forcing people to chase
> after potential VP's just opens up all sorts of problems, as can be
> witnessed by the amount of complaining and arguing that goes on in
> this and other forums.

Then why suggest an alternative rule whereby they have to chase even more VPs?

The Name Forgotten

unread,
Feb 18, 2009, 2:25:07 PM2/18/09
to

Untrue. That kind of play is illegal for different reasons - in that
it's tantamont to colluding (if not downright colluding). You don't
need a "play-to-win" rule to stop that from happening.

As I said, there are situation where a player might actually play to
lose, but I wonder if it will ever actually happen (at least in such a
way that it doesn't violate another rule).

LSJ

unread,
Feb 18, 2009, 2:28:22 PM2/18/09
to
The Name Forgotten wrote:
> On Feb 18, 8:56 pm, John Flournoy <carne...@gmail.com> wrote:
>> On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
>>
>> It's better than not having the rule, because it allows a judge to
>> sanction somebody who is clearly not playing to win, such as the
>> following scenario:
>>
>> Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
>> tournament.
>>
>> Without "play to win", Player B can play to give his buddy (player A)
>> the game win in the hopes that A will make the finals (which B won't
>> be able to do no matter what.)
>>
>
> Untrue. That kind of play is illegal for different reasons - in that
> it's tantamont to colluding (if not downright colluding). You don't
> need a "play-to-win" rule to stop that from happening.

It's not colluding (no conspiracy).

It's a violation of the sportsmanship rule, however. Namely the part of the
sportsmanship rule given by the current play to win rule.

> As I said, there are situation where a player might actually play to
> lose, but I wonder if it will ever actually happen (at least in such a
> way that it doesn't violate another rule).

The "another rule" you're hanging you argument on is the sportsmanship rule.
Which the "play to win" rule is merely a subset of. (The pertinent subset, in
this case.)

The Name Forgotten

unread,
Feb 18, 2009, 2:30:52 PM2/18/09
to

Can you run that scenario by me please? My head is hurting too much
from all the counter-arguments to even formulate that possibility. I
can't imagine a situation where you're left in a situation where your
only 2 options are to give one VP to your prey to get 2 VPs for a
gamewin, as opposed to not giving your prey a VP and giving it to
someone with a VP for a 2-2-1 split - AND you're forced to take the
2-2-1 split (given the context of this argument).

The Name Forgotten

unread,
Feb 18, 2009, 2:38:51 PM2/18/09
to
On Feb 18, 9:07 pm, John Flournoy <carne...@gmail.com> wrote:
> > > Changing it like that will open up new loophole corner-case scenarios
> > > though.
>
> > Such as? Would it create the double standard that this ruling seems to
> > do? Would it create controversy? I'm genuinely curious.
>
> Here's an a corner-case scenario.
>
> Three players. Each has one pool. A bleeding B bleeding C.
>
> Player A has a single minion. Player B has a single minion, tapped.
> Player C has no minions.
>
> Hero (Player A) has a KRC (with vote lock) in his hand. He already has
> one VP. Player B also has one VP.
>
> B has one card in his hand, which may or may not be a Wake. No other
> relevant cards exist.
>
> If the rule becomes 'you must play to maximize your VP and game win is
> not a consideration', then player A would be prohibited from playing
> his KRC - if he splits it between B and C, he will pass it, both will
> be ousted, and the VP split will end up A3/B2.

Not really true - it becomes a matter of risk assessment as LSJ
pointed out - but then that will lead to the same double standard -
which, leads me to believe that it's not a suitable compromise - so
I'll stick with my initial request - that the whole section about GWs
and VPs should be removed.

> And that seems (to me at least) like a rule that isn't clearly better
> than what currently exists.

Agreed. The compromise won't work - neither will the original - both
lead to a double standard. It should simply be removed.

> (By extension, there are lots of other scenarios where you might be
> prohibited from ousting someone cross-table with a vote, because that
> would reduce the maximum number of VPs you could have a reasonable
> chance at getting.)
>

Agreed. Thanks for bringing up a good, clear example.

LSJ

unread,
Feb 18, 2009, 2:45:06 PM2/18/09
to
The Name Forgotten wrote:
> Not really true - it becomes a matter of risk assessment as LSJ
> pointed out - but then that will lead to the same double standard -

What double standard?

> which, leads me to believe that it's not a suitable compromise - so
> I'll stick with my initial request - that the whole section about GWs
> and VPs should be removed.

To what purpose?

If you leave the sportsmanship rule, then the change accomplishes nothing
(except making it harder for judges to get the rule).

henrik

unread,
Feb 18, 2009, 2:55:00 PM2/18/09
to

Didn't have a drawn out scenario when I posted, but something like
this:

3 players left.
Just say you have 1 point, your prey has 0, your predator has 1 (and
your previous predator had 1 vp).
You can oust your prey, and by doing so allowing your predator to oust
you (yes, he will).
That'll leave you with a 2-2-1 split.

OR

You can let your prey oust your predator (yes, he will), and then oust
your prey (yes, you can).
That'll give you a 2-1-1-1 score, which gives you a game win.

Granted, my head is tired as well so there might be some flaw here.
But I'm sure you get the point.
The situation happens from time to time.

henrik

unread,
Feb 18, 2009, 3:19:23 PM2/18/09
to
On Feb 18, 8:55 pm, henrik <www.hen...@gmail.com> wrote:

> 3 players left.
> Just say you have 1 point, your prey has 0, your predator has 1 (and
> your previous predator had 1 vp).
> You can oust your prey, and by doing so allowing your predator to oust
> you (yes, he will).
> That'll leave you with a 2-2-1 split.
>
> OR
>
> You can let your prey oust your predator (yes, he will), and then oust
> your prey (yes, you can).
> That'll give you a 2-1-1-1 score, which gives you a game win.
>
> Granted, my head is tired as well so there might be some flaw here.
> But I'm sure you get the point.
> The situation happens from time to time.

After some thinking, that scenario might not be waterproof. Since
you'll get 2 victory points either way.
I'm pretty sure there's some variation which would show my point
though, I'll think about it for a while.

The Name Forgotten

unread,
Feb 18, 2009, 4:26:21 PM2/18/09
to
On Feb 18, 9:28 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 18, 8:56 pm, John Flournoy <carne...@gmail.com> wrote:
> >> On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> >> It's better than not having the rule, because it allows a judge to
> >> sanction somebody who is clearly not playing to win, such as the
> >> following scenario:
>
> >> Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
> >> tournament.
>
> >> Without "play to win", Player B can play to give his buddy (player A)
> >> the game win in the hopes that A will make the finals (which B won't
> >> be able to do no matter what.)
>
> > Untrue. That kind of play is illegal for different reasons - in that
> > it's tantamont to colluding (if not downright colluding). You don't
> > need a "play-to-win" rule to stop that from happening.
>
> It's not colluding (no conspiracy).

It seems like a very thin line to hide behind. While I agree with you
that it may not technically be colluding, it would have all the
hallmarks of colluding and in fact, it would be impossible to tell the
difference in any relevant practical way.

In other words, if another player had to call a judge over, and say
that he suspects A & B are colluding to get A into the final tables -
all the evidence would point to that being a fact. For example, if A &
B were, in fact, colluding, and met up 3 weeks before, or even 3
minutes before the game, and said they'd work together in that
situation - how would their in-game actions differ in any way
whatsoever.

While, as a judge, I might be lenient on A (given that you would find
it difficult to prove his involvement and that his actions don't
directly disrupt the game) the fact that B is acting in the actively
disruptive manner would certainly give cause for some sort of penalty
based on the high likelihood of collusion (there aren't many more ways
you could be certain then judging by their in-game activities, unless
they're really stupid enough to sign something, or get it on tape).

I think just the possibility that B's actions might be seen as
colluding, and the fact that it's pointed out as such will be enough
to at least make him try to play the game - at least, well enough that
you'd be satisfied under any 'play-to-win' condition as laid out by
the current ruling.

So, in short, while it might not actually be colluding - it is too
close to tell the difference - and any judge would (highly likely)
rule it as such, although probably not after giving a warning (just in
case the judge is wrong).

Sorry, but as a judge or a player on the same table as A & B, I'd have
no way of knowing the intentions of either A or B, or whether or not
they did, in fact, conspire or not. All the evidence present in the
game points to the fact that they did.

LSJ

unread,
Feb 18, 2009, 4:56:38 PM2/18/09
to
The Name Forgotten wrote:
> On Feb 18, 9:28 pm, LSJ <vtes...@white-wolf.com> wrote:
>> It's not colluding (no conspiracy).
>
> It seems like a very thin line to hide behind. While I agree with you
> that it may not technically be colluding, it would have all the
> hallmarks of colluding and in fact, it would be impossible to tell the
> difference in any relevant practical way.

As you like.

But fortunately, it isn't a line to hide behind at all, thanks to the blanket of
sportsmanship.

John Flournoy

unread,
Feb 18, 2009, 5:01:32 PM2/18/09
to
On Feb 18, 3:26 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
> Sorry, but as a judge or a player on the same table as A & B, I'd have
> no way of knowing the intentions of either A or B, or whether or not
> they did, in fact, conspire or not. All the evidence present in the
> game points to the fact that they did.

Let's say that player B announces "Hey, I'm going to throw the game to
make player A win."

Let's also assume that player A says "please don't, I don't want you
to do that."

And then player B begins to play to force a win for player A anyway
(at the detriment of his own game.). He does so completely politely
and civilly, despite repeated requests from player A to not play this
way

Without a section of the sportsmanship rules covering 'play to win',
what is player B violating by deliberately throwing the game to his
buddy in a clearly non-collusional way?

-John Flournoy

Matthew T. Morgan

unread,
Feb 18, 2009, 1:58:51 PM2/18/09
to
On Wed, 18 Feb 2009, The Name Forgotten wrote:

> On Feb 18, 6:35 pm, "Matthew T. Morgan" <farq...@io.com> wrote:
>> On Wed, 18 Feb 2009, The Name Forgotten wrote:
>>> Your choice is to:
>>
>>> Option A - oust C now, along with B - leaving you with 2 VP's for
>>> ousting B and surviving - which leaves you with the most VP's on the
>>> table.
>>
>> In addition to all the points made by other people, getting 2 VPs and more
>> VPs than anyone else at the table constitutes a game win.  This absolutely
>> is the correct choice and there are no rules of any variety that would
>> prohibit this option.
>
> Point taken. For some reason I thought you needed more than 2 VPs. My
> argument can still be true with one little adjustment, in that case -
> that being that another player has 2 VPs.

Okay, so now we're talking about:

A) 100% chance of a tie with 2 VPs.

B) 5% chance of a win with 3 VPs.

A is still a legal play, no matter how much you say it isn't.

> I.e. Player A bleeding Player B (who has already ousted player D)
> bleeding Player C (who has already ousted player E).
>
> That way, if you oust both, player B has 2 VPs to tie with you, and if
> you oust only B, C will get 3 VPs.
>
>>> OR:
>>
>> [make up rules that make no sense and then complain about them]
>
> Not a helpful comment. Cut-and-paste my argument and responses with
> the new example. You're intelligent enough to realise that - don't
> waste my time or yours.

Maybe that wasn't the most polite way to state it, but you are making up
rules. There is no rule that says you have to play the slightest odds to
go for the game win. Your insistence that there is is the only time
wasting going on here.

> The point still remains, the double standard still seems to exist -
> whatever cut off point you have for chasing after the possibility of a
> GW, risking VP's to get there, as opposed to the certainty of getting
> VP's, you'd need to apply the same cut off point to when a person can
> rightfully give up on a game and self-oust (or do some other crazy
> action) - or risk maintaining a double standard.
>
> Personally, I think almost all players (including myself) would think
> that this double standard does exist - but it certainly brings into
> question what it means for a person to reasonably expect that they
> will not getting a VP - and give them far greater leeway than many
> players would be happy with.

These points have been addressed in other responses in this thread. I
don't know why you feel that way, but you're wrong. The 2 VP play is
perfectly legal and valid.

I had this situation in the recent qualifier in Las Vegas:

There are three of us remaining in the game. I have vote lock, stealth
and a lot of pool. My prey has some intercept, lots of bounce and not
much pool. He also has a VP. My predator has one minion and one pool.
He's shown a little untap, intercept and combat, but I know his deck is
primarily a bleed deck. He's also indicated to me that he's at the end of
his rope (I've been trying to help him stay alive).

I draw into an Anarchist Uprising, which will oust both my prey and
predator, making me tie with 2 VPs with my prey. Now if I just bled,
there may have been some chance I'd draw into the super awesome hand and
be able to win the game, but I knew the chance was very low. It was much
more likely my bleed would be blocked (if no stealth) or bounced (if at
stealth) and neither would be good for me. If my prey gained 6, I'd lose
the opportunity to kill him with a single vote and still max out at 2 VPs.

I called the Anarchist Uprising and took my 2 VPs. This is perfectly
legal play. You can start all the threads you want about how it isn't,
but it is. Once you realize that, you'll realize there's no actual
problem here.

Matt Morgan

John Flournoy

unread,
Feb 18, 2009, 5:29:50 PM2/18/09
to
On Feb 18, 8:10 am, The Name Forgotten <tex...@telkomsa.net> wrote:

> Having a double standard of what reasonable is seems problematic to me
> (i.e. 5% chance of getting a VP as opposed to 5% chance of getting a
> GW). It looks to me like a Pandora's box situation with the way that
> the rules are currently written. The moment you give too much leeway
> in allowing a person to take 2 VP's instead of the potential of 3, you
> are also giving that same amount of leeway to a person who throws up
> his hands and says he has no chance of getting a VP with his S&B
> kindred spirits deck because his prey has 5 intercept on the table, so
> now he'll bleed out his predator instead (who stopped him from
> bleeding out his prey earlier in the game), or some similar situation.

I assume you mean 'take 2VPs and no game win instead of the potential
of a game win with however many VPs that includes'. Because if the
2VPs gives you a game win, you can absolutely always take that lesser
amount.

So, when talking about 'game win' or 'not':

If you have a reasonable chance to get a Game Win, you must try to do
so.
If you have a less-than-reasonable chance, you have leeway to do what
you like (as long as you're still getting as many VPs as you can.)

The exact same standard holds true to the 0VP/self-oust case:

If you have a reasonable chance to get VPs, you must try to do so.
If you have a less-than-reasonable chance to get any, you have leeway
to do what you like (including self-ousting.)

You keep referring to this as a double-standard, but it's not - it's
the same singular standard applied to two parallel cases (and a
standard where the threshold of "reasonable chance" is likely the
same.)

-John Flournoy

Frederick Scott

unread,
Feb 18, 2009, 5:35:06 PM2/18/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:67381b2c-a2bb-428b...@e3g2000vbe.googlegroups.com...

> On Feb 18, 8:56 pm, John Flournoy <carne...@gmail.com> wrote:
> > On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
> >
> > It's better than not having the rule, because it allows a judge to
> > sanction somebody who is clearly not playing to win, such as the
> > following scenario:
> >
> > Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
> > tournament.
> >
> > Without "play to win", Player B can play to give his buddy (player A)
> > the game win in the hopes that A will make the finals (which B won't
> > be able to do no matter what.)
>
> Untrue. That kind of play is illegal for different reasons - in that
> it's tantamont to colluding (if not downright colluding).

Nonsense. Colluding involves multiple people who are in collusion with
one another. If Player A had no intent to receive this boon then he'd
have done nothing wrong and he isn't colluding. Only Player B is causing
a problem and he isn't colluding, either, because he there's no person
for him to collude _with_. What he's doing wrong, precisely, is not
playing to advance his own game position - i.e., he's not playing to win.

(That said, I will comment that I'm not sure what the "Collusion to alter
the results of a game" actually means. Alter? In what way? Players
in a game are constantly attempting to alter its results - in their favor,
we would hope. They may even be colluding to alter its results if A
and B are cooperating against C in a way they perceive is to the benefit
of both of them. I'm not sure what the thing is in rule 5.1/"collusion"
clause that's identified as being 'bad'. LSJ?)

> You don't
> need a "play-to-win" rule to stop that from happening.

You don't need a "play-to-win' rule to stop anything bad from happening.
You could add rules willy-nilly to proscribe each specific form of bad
behavior you can identify. But the "play-to-win" rule addresses a class
of bad behaviors based on their common theme, and addresses the core of
the bad thing in all of them. Hence, PTW is a good way to address
players arbitrarily helping their buddies' tournament placement to their
own detriment. It still has the weaknesses you dislike about it - but
creating individual rules around individual types of bad behaviors would
have the same problem. How would a more specific rule routine and
concretely identify when Player B is trying to help Player A to his own
detriment any better than PTW does?

> As I said, there are situation where a player might actually play to
> lose, but I wonder if it will ever actually happen (at least in such a
> way that it doesn't violate another rule).

You haven't answered the question of what other rules cover the same ground.
Collusion doesn't. You'd have to demonstrate that there are other such
rules, first.

Fred


The Name Forgotten

unread,
Feb 18, 2009, 6:01:31 PM2/18/09
to

No stress. John (I think it was him anyway) provided a good example. I
realise the compromise wasn't a good one. Better to remove that part
of the rule altogether.

The Name Forgotten

unread,
Feb 18, 2009, 6:08:55 PM2/18/09
to
On Feb 18, 9:45 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > Not really true - it becomes a matter of risk assessment as LSJ
> > pointed out - but then that will lead to the same double standard -
>
> What double standard?

I'll just cut and paste my previous argument:


>


> > which, leads me to believe that it's not a suitable compromise - so
> > I'll stick with my initial request - that the whole section about GWs
> > and VPs should be removed.
>
> To what purpose?
>
> If you leave the sportsmanship rule, then the change accomplishes nothing
> (except making it harder for judges to get the rule).

It completely eliminates the double standard I mention above. It leads
to less debate and argument then to try to force players to continue
playing or be penalised, which is, more often than not, counter-
productive.

The Name Forgotten

unread,
Feb 18, 2009, 6:13:02 PM2/18/09
to

Sure, but then it seems to me that the question posed was, if the
blanket of sportmanship was removed, wouldn't players be able to help
other players win tables, even to their own detriment? To which my
response is, No, it wouldn't happen, because such acts display in
exactly the same manner that colluding would - leading the judge to
call it up as such.

The Name Forgotten

unread,
Feb 18, 2009, 6:17:06 PM2/18/09
to

How is it clearly non-collusional? It is only clear in A & B's head.
As a judge, or a fellow player, you are not inside player A or B's
head - you have no way of telling whether they have colluded or not,
except by their actions in the game. In fact, such a act could be
clearly staged by people who are, in fact colluding and trying to make
it look like they're not. As a judge, you'd go to B and say that
working to help A in such a manner could be considered colluding, as
all the evidence points that way - if he continues to act in such a
way, then the judgment of colluding will be placed squarely on his
shoulders.

LSJ

unread,
Feb 18, 2009, 6:29:07 PM2/18/09
to

Two things:

1) Your suggestion is to do away with the entire sportsmanship requirement?
I thought you were only suggesting doing away with the "must play to win" part.

2) This would then just be a matter changing the name by which the same rules
are enforced. Change for change's sake. So no need to do it.

The Name Forgotten

unread,
Feb 18, 2009, 6:36:22 PM2/18/09
to
On Feb 18, 8:58 pm, "Matthew T. Morgan" <farq...@io.com> wrote:
> On Wed, 18 Feb 2009, The Name Forgotten wrote:
> > On Feb 18, 6:35 pm, "Matthew T. Morgan" <farq...@io.com> wrote:
> >> On Wed, 18 Feb 2009, The Name Forgotten wrote:
> >>> Your choice is to:
>
> >>> Option A - oust C now, along with B - leaving you with 2 VP's for
> >>> ousting B and surviving - which leaves you with the most VP's on the
> >>> table.
>
> >> In addition to all the points made by other people, getting 2 VPs and more
> >> VPs than anyone else at the table constitutes a game win.  This absolutely
> >> is the correct choice and there are no rules of any variety that would
> >> prohibit this option.
>
> > Point taken. For some reason I thought you needed more than 2 VPs. My
> > argument can still be true with one little adjustment, in that case -
> > that being that another player has 2 VPs.
>
> Okay, so now we're talking about:
>
> A) 100% chance of a tie with 2 VPs.
>
> B) 5% chance of a win with 3 VPs.
>
> A is still a legal play, no matter how much you say it isn't.

I agree. I'd rule that it is legal. But, that being said, if a player
has the same 5% chance of getting a single VP, and decides to cross-
table destroy his grand-predator for no good reason whatsoever or self-
oust, and says, "hey look, it was reasonable to say I wasn't going to
get any VP's", it seems to fly in the face of the general perception
of what it means to not have a reasonable chance of getting a VP and
being able to do as you please. Most people would say the 5% chance of
winning with 3 VP's over the certainty of 2 VP's makes either a legal
play, but then most aren't willing to say the same when it comes to 5%
chance of winning 1 VP over 100% chance of getting 0 VPs, saying that
people MUST go for that 5% chance to 'maximise VPs' - although you
don't need to take that same chance to maximise GW.

> > Not a helpful comment. Cut-and-paste my argument and responses with
> > the new example. You're intelligent enough to realise that - don't
> > waste my time or yours.
>
> Maybe that wasn't the most polite way to state it, but you are making up
> rules.  There is no rule that says you have to play the slightest odds to
> go for the game win.  Your insistence that there is is the only time
> wasting going on here.

My insistence is that, as a judge, you are expected to hold the *same*
standard for both scenarios, as they both apply to the same ruling
and, thus, demand the same degree of reason to be applied. This,
however, flies in the face of common practice and understanding -
meaning that either the common understanding is wrong, the ruling is
wrong, or there is, in fact, a double standard (or alternately, that
I'm misunderstanding something somewhere).

>
> > The point still remains, the double standard still seems to exist -
> > whatever cut off point you have for chasing after the possibility of a
> > GW, risking VP's to get there, as opposed to the certainty of getting
> > VP's, you'd need to apply the same cut off point to when a person can
> > rightfully give up on a game and self-oust (or do some other crazy
> > action) - or risk maintaining a double standard.
>
> > Personally, I think almost all players (including myself) would think
> > that this double standard does exist - but it certainly brings into
> > question what it means for a person to reasonably expect that they
> > will not getting a VP - and give them far greater leeway than many
> > players would be happy with.
>
> These points have been addressed in other responses in this thread.

I don't feel that it has been adequately addressed in other posts.

>  I
> don't know why you feel that way, but you're wrong.  The 2 VP play is
> perfectly legal and valid.

As stated before, agreed.

>
> I had this situation in the recent qualifier in Las Vegas:
>
> There are three of us remaining in the game.  I have vote lock, stealth
> and a lot of pool.  My prey has some intercept, lots of bounce and not
> much pool.  He also has a VP.  My predator has one minion and one pool.
> He's shown a little untap, intercept and combat, but I know his deck is
> primarily a bleed deck.  He's also indicated to me that he's at the end of
> his rope (I've been trying to help him stay alive).
>
> I draw into an Anarchist Uprising, which will oust both my prey and
> predator, making me tie with 2 VPs with my prey.  Now if I just bled,
> there may have been some chance I'd draw into the super awesome hand and
> be able to win the game, but I knew the chance was very low.  It was much
> more likely my bleed would be blocked (if no stealth) or bounced (if at
> stealth) and neither would be good for me.  If my prey gained 6, I'd lose
> the opportunity to kill him with a single vote and still max out at 2 VPs.
>
> I called the Anarchist Uprising and took my 2 VPs.  This is perfectly
> legal play.  You can start all the threads you want about how it isn't,
> but it is.  

I agree that it is legal play - but it creates a double standard in
the process.

> Once you realize that, you'll realize there's no actual
> problem here.

The double standard problem remains - why is it perfectly fine in one
situation to force a person to take a 5% chance at a VP, but it's not
perfectly fine to force a 5% chance at a GW? If 5% is too low, then
increase the number to whatever you like. The fact remains that the
cutoff for the first situation seems to be much lower than the cutoff
for the second - at least in the general understanding of the rule.
I'm hoping that there is some other factor that can explain this
difference, but at the moment I'm failing to find it. If one cannot be
found, then I think the rule should be dismissed, at least as a hard
and fast rule with potential judgements and penalties attached.

LSJ

unread,
Feb 18, 2009, 6:42:01 PM2/18/09
to
The Name Forgotten wrote:
> On Feb 18, 9:45 pm, LSJ <vtes...@white-wolf.com> wrote:
>> The Name Forgotten wrote:
>>> Not really true - it becomes a matter of risk assessment as LSJ
>>> pointed out - but then that will lead to the same double standard -
>> What double standard?
>
> I'll just cut and paste my previous argument:
>
> Let's say, for example, I'm a judge in a tournament with 2 tables. On

An anecdote?
No description of what double standard you claim exists?

A concise description of what you mean would be nice, rather than leaving it to
be guessed at through the veil of a single example.

> the first table is the situation I described above. On the second

What "above"?

> table is a situation where someone is purposefully bleeding
> backwards,
> and refusing to bleed his prey because, he feels, he has no way of
> ousting his prey.
>
> Let's further say that I have all the time in the world to examine
> the
> decks and do a complete analysis of people's odds in the game. Some
> outside influences, surely, I cannot account for, but those outside
> situations could happen at either table, so they are irrelevant for
> the purposes of this argument.
>
> Now, in Table A - the person has a 60% chance of drawing the stealth
> card he needs by playing his action card to get a GW, and will be
> ousted in the following turn if he does nothing (as opposed to
> choosing the safer 2VP route).
> In table B - the person has a 60% chance of drawing the stealth card
> he needs by playing his action card to get a VP, and will be ousted
> in
> the following turn if he does nothing (as opposed to bleeding his
> predator).
>
> From what I've understood from other posts - in Table A, it's still
> acceptable to 'play it safe' and take the 2 VP route instead of a GW
> -
> however, in Table B, giving up when you still have such a good chance
> of earning a VP is not acceptable.

I don't know what the situation is in Table A, but I gather it is different in
the appropriate way (no double standard).

>>> which, leads me to believe that it's not a suitable compromise - so
>>> I'll stick with my initial request - that the whole section about GWs
>>> and VPs should be removed.
>> To what purpose?
>>
>> If you leave the sportsmanship rule, then the change accomplishes nothing
>> (except making it harder for judges to get the rule).
>
> It completely eliminates the double standard I mention above. It leads
> to less debate and argument then to try to force players to continue
> playing or be penalised, which is, more often than not, counter-
> productive.

I cannot see how. There is scant debate now. Removing the PTW description but
leaving the sportsmanship rule changes nothing -- the requirement to play to win
is inherent in sportsmanship -- so the debate level would not decrease, although
it may increase if the clarification of PTW is removed from the official document.

The Name Forgotten

unread,
Feb 18, 2009, 6:54:58 PM2/18/09
to
On Feb 19, 12:29 am, John Flournoy <carne...@gmail.com> wrote:
> On Feb 18, 8:10 am, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> I assume you mean 'take 2VPs and no game win instead of the potential
> of a game win with however many VPs that includes'.

Right. In this case the potential for 3 VPs and the GW as opposed to
2VPs and no GW (due to a tie).

> So, when talking about 'game win' or 'not':
>
> If you have a reasonable chance to get a Game Win, you must try to do
> so.
> If you have a less-than-reasonable chance, you have leeway to do what
> you like (as long as you're still getting as many VPs as you can.)
>
> The exact same standard holds true to the 0VP/self-oust case:
>
> If you have a reasonable chance to get VPs, you must try to do so.
> If you have a less-than-reasonable chance to get any, you have leeway
> to do what you like (including self-ousting.)
>
> You keep referring to this as a double-standard, but it's not - it's
> the same singular standard applied to two parallel cases (and a
> standard where the threshold of "reasonable chance" is likely the
> same.)
>

You've got my point in a nut-shell, except that I don't think that the
threshold of "reasonable chance" is the same in both situations -
hence the standard has 2 different values - one which is applied to
the one (GW) situation, and the other which is applied to the other
(self-oust) situation.

Trying to go through newsthreads doesn't seem to help clarify one way
or the other - it only reinforces the image that there is, in fact, a
double standard. If there isn't really a double standard, and the
value (as far as risk analysis) is the same, then it should be stated
as such - but I feel that it would either force players to go for GW's
or allow too many players the leeway of self-ousting or other nonsense
far too easily.

That's why I'm asking for the rule to be dropped. I understand why
people might ask for it to be replaced with something, although I'm
not convinced that it does - most of the problems is already covered
in other sections, and with very little re-editing, you could cover
them all.

The Name Forgotten

unread,
Feb 18, 2009, 7:15:03 PM2/18/09
to
On Feb 19, 12:35 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:67381b2c-a2bb-428b...@e3g2000vbe.googlegroups.com...

>
>
>
>
>
> > On Feb 18, 8:56 pm, John Flournoy <carne...@gmail.com> wrote:
> > > On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> > > It's better than not having the rule, because it allows a judge to
> > > sanction somebody who is clearly not playing to win, such as the
> > > following scenario:
>
> > > Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
> > > tournament.
>
> > > Without "play to win", Player B can play to give his buddy (player A)
> > > the game win in the hopes that A will make the finals (which B won't
> > > be able to do no matter what.)
>
> > Untrue. That kind of play is illegal for different reasons - in that
> > it's tantamont to colluding (if not downright colluding).
>
> Nonsense. Colluding involves multiple people who are in collusion with
> one another. If Player A had no intent to receive this boon then he'd
> have done nothing wrong and he isn't colluding. Only Player B is causing
> a problem and he isn't colluding, either, because he there's no person
> for him to collude _with_.

I've already agreed to this point. My counter point is that, to
everyone else, it would appear as if you are colluding, and you'd have
no way to prove them false, and they'd have every reason to suspect
you. I agree that Player A should not be called up on it, because he
hasn't directly done anything to change the game state in any way that
differs from how he would normally play (unless, of course, he has).
Player B on the other hand, has changed the game state to his own
detriment to favour another player.

At best, this would warrant a warning from the judge to Player B that
such actions appear to be colluding, and if Player B continues to
favour Player A in such a manner, then it will, in fact, be considered
colluding and Player B will be penalised for it, and, potentially, the
game state restored (to counteract how Player B detrimentally affected
the game).

> What he's doing wrong, precisely, is not
> playing to advance his own game position - i.e., he's not playing to win.

That is true as well.

> > You don't
> > need a "play-to-win" rule to stop that from happening.
>
> You don't need a "play-to-win' rule to stop anything bad from happening.
> You could add rules willy-nilly to proscribe each specific form of bad
> behavior you can identify. But the "play-to-win" rule addresses a class
> of bad behaviors based on their common theme, and addresses the core of
> the bad thing in all of them. Hence, PTW is a good way to address
> players arbitrarily helping their buddies' tournament placement to their
> own detriment. It still has the weaknesses you dislike about it - but
> creating individual rules around individual types of bad behaviors would
> have the same problem. How would a more specific rule routine and
> concretely identify when Player B is trying to help Player A to his own
> detriment any better than PTW does?

Well, as I said, it would be considered an act of colluding, whether
or not the intent on the side of both players is present. The bigger
problem, in my book, is how to stop players from doing nothing at all,
or an absolute minimum. I think that it's the rarest kind of player -
which is why I think it's better to abandon the PTW rule (at least the
tournament part) which seems to affect far more situations.

>
> > As I said, there are situation where a player might actually play to
> > lose, but I wonder if it will ever actually happen (at least in such a
> > way that it doesn't violate another rule).
>
> You haven't answered the question of what other rules cover the same ground.
> Collusion doesn't. You'd have to demonstrate that there are other such
> rules, first.

Collusion doesn't directly, agreed. Still, in your example, any judge
would rule that it was collusion - having no way to prove that it
wasn't and every bit of proof that it was - at least in the mind of
the player who gives away his VP - who, in every way, intended to work
with the other player to help him achieve his goal of winning -i.e. he
would've willingly colluded with his friend to help his friend win.

You could probably formulate a rule to prevent that situation, but
there would be no need, as, in practice, another rule already in
existence would do the same job, even if it wasn't intended to be that
way.

John Flournoy

unread,
Feb 18, 2009, 7:39:32 PM2/18/09
to
On Feb 18, 6:15 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
> On Feb 19, 12:35 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
>
>
>
> > Nonsense.  Colluding involves multiple people who are in collusion with
> > one another.  If Player A had no intent to receive this boon then he'd
> > have done nothing wrong and he isn't colluding.  Only Player B is causing
> > a problem and he isn't colluding, either, because he there's no person
> > for him to collude _with_.
>
> I've already agreed to this point. My counter point is that, to
> everyone else, it would appear as if you are colluding, and you'd have
> no way to prove them false, and they'd have every reason to suspect
> you. I agree that Player A should not be called up on it, because he
> hasn't directly done anything to change the game state in any way that
> differs from how he would normally play (unless, of course, he has).
> Player B on the other hand, has changed the game state to his own
> detriment to favour another player.
>

If you cannot sanction player A for collusion, than you cannot
sanction player B for it - collusion is something done in agreement by
two people.

If Player A says and does nothing to indicate that he is in collusion
with B (and tries to do and say things to disprove it), then A is not
provably colluding, and thus player B is not provably colluding
either.

Mind you, player B is certainly being unsportsmanlike because he is
not playing to win, clearly - but he is not provably in collusion with
player A.

> At best, this would warrant a warning from the judge to Player B that
> such actions appear to be colluding, and if Player B continues to
> favour Player A in such a manner, then it will, in fact, be considered
> colluding and Player B will be penalised for it, and, potentially, the
> game state restored (to counteract how Player B detrimentally affected
> the game).

Again, "collusion" is not something one person can do by themselves,
by definition.

> > What he's doing wrong, precisely, is not
> > playing to advance his own game position - i.e., he's not playing to win.
>
> That is true as well.

It is true _instead_, actually.

> > > You don't
> > > need a "play-to-win" rule to stop that from happening.

You do, because it's not covered by "collusion".

> > You don't need a "play-to-win' rule to stop anything bad from happening.
> > You could add rules willy-nilly to proscribe each specific form of bad
> > behavior you can identify.  But the "play-to-win" rule addresses a class
> > of bad behaviors based on their common theme, and addresses the core of
> > the bad thing in all of them. Hence, PTW is a good way to address
> > players arbitrarily helping their buddies' tournament placement to their
> > own detriment.  It still has the weaknesses you dislike about it - but
> > creating individual rules around individual types of bad behaviors would
> > have the same problem.  How would a more specific rule routine and
> > concretely identify when Player B is trying to help Player A to his own
> > detriment any better than PTW does?
>
> Well, as I said, it would be considered an act of colluding, whether
> or not the intent on the side of both players is present.

And again, you'd be mistaken about what "collusion" means.

> Collusion doesn't directly, agreed. Still, in your example, any judge
> would rule that it was collusion - having no way to prove that it
> wasn't and every bit of proof that it was - at least in the mind of
> the player who gives away his VP - who, in every way, intended to work
> with the other player to help him achieve his goal of winning -i.e. he
> would've willingly colluded with his friend to help his friend win.

Again, "collusion" does not mean what you claim it does.

> You could probably formulate a rule to prevent that situation, but
> there would be no need, as, in practice, another rule already in
> existence would do the same job, even if it wasn't intended to be that
> way.

In practice, it doesn't, because "Collusion" only applies to bad
sportsmanship between players clearly working together (not just one
player working for the benefit of another.)


Here's a specific example for you. A -> B -> C-> D-> E.

If Player D starts trying to oust player B with cross-table bleeds,
saying he wants to give the game to player A - and player A starts
telling him to stop that, that he wants no part of it, and even goes
so far as to start Eagle Sighting D's bleeds - then it's not
collusion.

It doesn't matter how absurdly in A's favor D is playing in that case;
A is amply stating and demonstrating that they did not in any way
agree to D's declarations, and that means that _neither_ A nor D are
guilty of collusion.

And thus the "Collusion" rule doesn't apply.

Just because D starts throwing a game doesn't automatically mean A
agreed to anything; and if you can't sanction player A for "making a
secret agreement with D", you can't sanction D for making the same
unprovable agreement with A.

(Of course, it's completely different if A gives you any reason to
think he knew about it and agreed to it in advance.)

-John Flournoy

John Flournoy

unread,
Feb 18, 2009, 7:43:14 PM2/18/09
to

Answered elsewhere in the thread.

Short version: "Collusion" does not mean "player B is helping player
A", it means "A and B both agreed that this would occur and are both
guilty". You cannot reasonably sanction someone for collusion if
you're not also reasonably sanctioning the other participant(s) as
well.

(You can certainly potentially warn B for other reasons, but "helping
another player" is not at all the same thing as "collusion".)

-John Flournoy

John Flournoy

unread,
Feb 18, 2009, 7:49:53 PM2/18/09
to

My view on it is this:

The rules apply the same standard: "Reasonable chance." They do not
create a double standard, because they use the same term for both
situations and do not say "that means A in one case and B in another."

Judges, however, may decide on their own that "reasonable chance" is
different in each case and apply different standards.

This is unpreventable; there is no way to quantify things in a manner
that people will agree to that prevents judges from making judgement
calls about how and when to apply the rules.

This is no different than the fact that the judges already get to set
whatever standards they like in terms of other basic aspects of the
rules, such as "when do I give a warning instead of a DQ" or the
explicit judgement call of "do I allow spectators to watch a game when
a player requests that they be asked not to" (rules 2.4, "a judge, who
may grant or deny the request as he or she sees fit.") or even "do I
as a head judge overrule what a subordinate judge has already ruled."

Judges having to apply their own set of standards to tournament rules
situations already exists, as does a difference in how two different
judges would apply the rules to the same situation.

That does not make the rules guilty of inherently creating a double
standard; that is the doing of the judges.

-John Flournoy

John Flournoy

unread,
Feb 18, 2009, 7:53:12 PM2/18/09
to
On Feb 18, 5:54 pm, The Name Forgotten <tex...@telkomsa.net> wrote:

> That's why I'm asking for the rule to be dropped. I understand why
> people might ask for it to be replaced with something, although I'm
> not convinced that it does - most of the problems is already covered
> in other sections, and with very little re-editing, you could cover
> them all.

If you have a specific idea on how such an edit could work to cover it
all, that'd be welcome I'm sure (at least for consideration and
further discussion.)

I think the general response to your suggestion of just dropping the
rule entirely without an edit, however, is insufficient to a lot of
people to cover the problems that led to the rule being spelled out in
the first place.

-John Flournoy

The Name Forgotten

unread,
Feb 18, 2009, 8:05:10 PM2/18/09
to
On Feb 19, 1:42 am, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 18, 9:45 pm, LSJ <vtes...@white-wolf.com> wrote:
> >> The Name Forgotten wrote:
> >>> Not really true - it becomes a matter of risk assessment as LSJ
> >>> pointed out - but then that will lead to the same double standard -
> >> What double standard?
>
> > I'll just cut and paste my previous argument:
>
> > Let's say, for example, I'm a judge in a tournament with 2 tables. On
>
> An anecdote?
> No description of what double standard you claim exists?

Great. Ok, to repeat myself again (which at this point I'm very tired
of doing).

The double standard is this: in one situation taking a 5% (or whatever
percentage you're comfortable with) is considered a reasonable chance
of getting a VP, so the PTW rule to maximise VP (or at least how
people have been playing it) says you shouldn't self-oust, however, in
a similar situation you have exactly the same percentage chance of
getting a GW, but this time it is not considered a reasonable chance,
so the PTW rule to maximise GW says you don't need to, as you do not
have a reasonable chance. However - the percentage chance is the same
in both situations and the rule is the same one - i.e. what PTW means
in a tournament. It is thus a double standard.

The example:

Table A:

Let's imagine you are in this situation (you being player A)


Player A has 2 ready, untapped vampires, both on 1 blood, and with
the
vote majority. It is your turn and you have 2 pool.
Player B has no ready minions and is on 3 pool, and has already ousted
D.
Player C has 5 ready imbued, each with second sights and strike with
conviction, 2 are untapped and another 2 have laptops, and an
Unmasking in play. C is on 1 pool and has already ousted E.


Player A is bleeding B who is bleeding C.


You have one stealth modifier in your hand at the moment (a forgotten
labyrinth), and you know you still have some in your deck, although
it
is highly unlikely that you'll cycle into them by only playing a card
or two (let's say under X% chance). You are not playing any intercept
cards. You do, however, have multiple Kine Resources Contested, and
all the votes you need to make it pass.


Your best bet of getting a Game Win and/or VPs is by calling a KRC
(unless someone can see a better way), which you successfully do with
the use of your forgotten labyrinth, and replace with cards that are
not stealth.


You now have the option of ousting any player on the table. In fact,
you could oust 2 players in one go.


You feel that it is highly unlikely (under X% chance) that, should
you
oust B, you will be able to call your second KRC and get it to pass
successfully. However, if C is not ousted by the end of your turn, C
will likely oust you and get the GW.

You choose to oust both B & C, which you do successfully rather than
take the X% chance at a GW.


Table B:

It is player A's turn.

Player A has 1 ready, untapped Malkavian with DEM OBF, and no other
disciplines, and is on 1 pool.
Player B has an untapped vampire with 4 Raven Spies, 4 other tapped
minions and 2 pool, 1 VP.
Player C has 2 tapped minions left, both have inherent stealth and is
on 1 pool, 1 VP.

Player A has a Kindred Spirits and a faceless night in his hand and no
untap/intercept in his deck. No-one has played any intercept beyond
what's on the table the entire game.

Player A has the option of trying to bleed out his prey, and hope that
he cycles into an Elder Impersonation to get past B's untapped minion
(an X% chance - exactly equal to the percentage in the table above).
Or Player A may vindictively oust his predator C, who has been giving
him grief.

Player A chooses to Kindred Spirits bleed his predator C and ousts his
predator, choosing to do so rather than take a X% chance of a VP.

The common perception is that the first situation (Table A) is
perfectly acceptable, whereas in the second situation (Table B) the
player should be forced to try to get a VP, according to the Play-to-
Win rule.

Bloody hell. I'm not repeating that, so I hope people are following
this thread.

> I don't know what the situation is in Table A, but I gather it is different in
> the appropriate way (no double standard).

Given the full situation - how does the situation differ in an
appropriate way.

> > It completely eliminates the double standard I mention above. It leads
> > to less debate and argument then to try to force players to continue
> > playing or be penalised, which is, more often than not, counter-
> > productive.
>
> I cannot see how. There is scant debate now. Removing the PTW description but
> leaving the sportsmanship rule changes nothing

It changes things significantly. I beg to differ on the scant debate.
It's a theme I see appear relatively regularly on this newsgroup, at
least moreso than most rule-based issues. And, more importantly, I've
heard it come up in more than one game when a player self-ousts or
does something equally stupid.

> -- the requirement to play to win
> is inherent in sportsmanship --

Agreed. But I don't think maximising GWs and VPs if you have a
reasonable chance is a good definition for it.

> so the debate level would not decrease, although
> it may increase if the clarification of PTW is removed from the official document.

Perhaps on the newsgroups, yes. But there would be no debate about it
at tournaments, and the other sideline issues that people might debate
about at tournaments, are being debated there already (e.g. one player
helping another to his own detriment, or one player targeting another
cross-table because of an out-of-game slight, etc.) People will not go
around saying "I had no reasonable chance of getting a VP so I could
self-oust" followed by "You could've cycled into something good - you
don't know" etc.

Maybe Peter's 'always go for your prey' approach is needed for these
rules 8)

John Flournoy

unread,
Feb 18, 2009, 8:07:38 PM2/18/09
to
On Feb 18, 5:36 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> I agree. I'd rule that it is legal. But, that being said, if a player
> has the same 5% chance of getting a single VP, and decides to cross-
> table destroy his grand-predator for no good reason whatsoever or self-
> oust, and says, "hey look, it was reasonable to say I wasn't going to
> get any VP's", it seems to fly in the face of the general perception
> of what it means to not have a reasonable chance of getting a VP and
> being able to do as you please. Most people would say the 5% chance of
> winning with 3 VP's over the certainty of 2 VP's makes either a legal
> play, but then most aren't willing to say the same when it comes to 5%
> chance of winning 1 VP over 100% chance of getting 0 VPs, saying that
> people MUST go for that 5% chance to 'maximise VPs' - although you
> don't need to take that same chance to maximise GW.

See, here's where your argument breaks down - I haven't seen "most
people" making the arguments you claim.

In fact, I haven't really seen any sort of mass of people saying that
"you have a 5% chance of getting 1VP versus zero prohibits you from
playing as if you will get zero", both in tournament settings and on
the newsgroup.

> My insistence is that, as a judge, you are expected to hold the *same*
> standard for both scenarios, as they both apply to the same ruling
> and, thus, demand the same degree of reason to be applied. This,
> however, flies in the face of common practice and understanding -
> meaning that either the common understanding is wrong, the ruling is
> wrong, or there is, in fact, a double standard (or alternately, that
> I'm misunderstanding something somewhere).

You claim that this application of a double standard is "common
practice". If it's common, can you give some specific examples of how
this has been applied (or stated by judges that they would apply it
differently?)

I'm not being facetious or rhetorical here - your perception is
clearly that the standard of "reasonable chance" is applied VERY
tightly on people with 0 VP, and my perception (and I suspect the
perception of many of the people I play with) is almost entirely
absent of this practice.

> The double standard problem remains - why is it perfectly fine in one
> situation to force a person to take a 5% chance at a VP, but it's not
> perfectly fine to force a 5% chance at a GW? If 5% is too low, then
> increase the number to whatever you like. The fact remains that the
> cutoff for the first situation seems to be much lower than the cutoff
> for the second - at least in the general understanding of the rule.
> I'm hoping that there is some other factor that can explain this
> difference, but at the moment I'm failing to find it. If one cannot be
> found, then I think the rule should be dismissed, at least as a hard
> and fast rule with potential judgements and penalties attached.

It's not a "hard and fast rule", because it explicitly relies on the
judgement of the judge(s) as to what a "reasonable chance" is. It is
in fact a very soft rule, as a result.

And again, I simply have never, ever, ever, ever in any tournament
I've attended seen a player with 0 VP forced to take a chance at
getting 1 VP and/or prohibited from self-ousting unless they were very
clearly (to the judges, onlookers and other players) still very much
capable of continuing. I've never seen a judge say "you still have a
reasonable chance to continue" and seen anyone disagree with that
ruling for more than the briefest moment once given, either at the
moment or after the tournament was done. And I can't recall it ever
being mentioned on the tournament reports on the newsgroups or in war-
story discussions with other players, either.

I _have_ seen judges called over and asked to verify that a player
didn't have a reasonable chance, and seen the judge say "I agree, his
chance is not reasonable" after an aside discussion (involving things
along the lines of "I know what's in my deck, and I lack what I need
to ever get past the defense my prey has showing clearly on the
table") - I believe I have even seen LSJ uphold a "no reasonable
chance" once.

Which isn't to say it doesn't occur - I'm just saying that in my
experience, it's very much not "common practice" or "general
understanding" or any sort of routinely applied double standard.

Feel free to provide anecdotes based on your experiences, though, as
specific examples of how these rules get applied in ways that the
people debating you in this thread seem not to experience would be
helpful.

-John Flournoy

The Name Forgotten

unread,
Feb 18, 2009, 8:15:38 PM2/18/09
to
On Feb 19, 1:29 am, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 18, 11:56 pm, LSJ <vtes...@white-wolf.com> wrote:
> >> The Name Forgotten wrote:
> >>> On Feb 18, 9:28 pm, LSJ <vtes...@white-wolf.com> wrote:
> >>>> It's not colluding (no conspiracy).
> >>> It seems like a very thin line to hide behind. While I agree with you
> >>> that it may not technically be colluding, it would have all the
> >>> hallmarks of colluding and in fact, it would be impossible to tell the
> >>> difference in any relevant practical way.
> >> As you like.
>
> >> But fortunately, it isn't a line to hide behind at all, thanks to the blanket of
> >> sportsmanship.
>
> > Sure, but then it seems to me that the question posed was, if the
> > blanket of sportmanship was removed, wouldn't players be able to help
> > other players win tables, even to their own detriment? To which my
> > response is, No, it wouldn't happen, because such acts display in
> > exactly the same manner that colluding would - leading the judge to
> > call it up as such.
>
> Two things:
>
> 1) Your suggestion is to do away with the entire sportsmanship requirement?
> I thought you were only suggesting doing away with the "must play to win" part.

No, as I've stated several times before only the part that states:
4.8. Play to Win
For tournaments, playing to win means playing to get a Game Win
if it is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible.

I didn't feel the need to cut and paste the relevant parts every time
I make mention of that section, but it appears that I must. I forget
that you probably read hundreds of posts and emails a day, and it gets
hard to keep track. Just remember that it gets incredibly tedious to
repeat myself all the time, and to make every last post entirely
enclusive - it's much easier if you were to follow the post. No
offence, but I'm not going to waste my time by repeating my points
continuously, so I'm probably not going to respond to points I've
already responded to in the future.

> 2) This would then just be a matter changing the name by which the same rules
> are enforced. Change for change's sake. So no need to do it.

God no. It would remove all the problems. However, my point is that
all the supposed problems that you and others say would suddenly arise
from removing that portion, are, in fact, already covered in other
rules (directly or indirectly) - although I do agree that a minor
adjustment to stop cross-table nonsense without cause or provocation
may be necessary for claritys sake, but that is about it.

LSJ

unread,
Feb 18, 2009, 8:24:22 PM2/18/09
to
The Name Forgotten wrote:
> Great. Ok, to repeat myself again (which at this point I'm very tired
> of doing).

Preaching to the choir, here.

> The double standard is this: in one situation taking a 5% (or whatever
> percentage you're comfortable with) is considered a reasonable chance
> of getting a VP, so the PTW rule to maximise VP (or at least how
> people have been playing it) says you shouldn't self-oust, however, in
> a similar situation you have exactly the same percentage chance of
> getting a GW, but this time it is not considered a reasonable chance,
> so the PTW rule to maximise GW says you don't need to, as you do not
> have a reasonable chance. However - the percentage chance is the same
> in both situations and the rule is the same one - i.e. what PTW means
> in a tournament. It is thus a double standard.

Ah, that's not a double standard.

This is where descriptions come in useful -- to help uncover such non-standard
usages of terms. Now we're getting somewhere.

In practice, this is called judgment.

But more to the point: in practice, you cannot calculate a percentage.

You just have to go with reasonable.

That's not a double standard.

Even the fact that different judges may judge it differently is not a double
standard. That's just the definition of judgment.

And doing away with the PTW clarification of the sportsmanship rule will not do
away with judgments (and thus, will not do away with what you call a double
standard).

Indeed, doing away will the whole of the sportsmanship rule will not accomplish
this (not that we'd ever do that, since it's such a bad idea for other reasons).
There's plenty left for judges to judge: stalling, was that action actually
declared, was that vote actually cast, how to fix a forgotten Army of Rats, and
so on. (All of which lead to your non-standard double standards)

LSJ

unread,
Feb 18, 2009, 8:28:42 PM2/18/09
to
The Name Forgotten wrote:
> On Feb 19, 1:29 am, LSJ <vtes...@white-wolf.com> wrote:
>> 1) Your suggestion is to do away with the entire sportsmanship requirement?
>> I thought you were only suggesting doing away with the "must play to win" part.
>
> No, as I've stated several times before only the part that states:
> 4.8. Play to Win
> For tournaments, playing to win means playing to get a Game Win
> if it is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible.

As I've stated several times before, that part is a redundant, superfluous,
unnecessary restatement of a portion of the general sportsmanship requirement.

>> 2) This would then just be a matter changing the name by which the same rules
>> are enforced. Change for change's sake. So no need to do it.
>
> God no. It would remove all the problems.

Removing the redundant clarification changes nothing.

Frederick Scott

unread,
Feb 18, 2009, 9:25:16 PM2/18/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:f8de0e5f-cf95-473b...@r15g2000prd.googlegroups.com...

> On Feb 19, 12:35 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
>> "The Name Forgotten" <tex...@telkomsa.net> wrote in
>> messagenews:67381b2c-a2bb-428b...@e3g2000vbe.googlegroups.com...
>>
>> > On Feb 18, 8:56 pm, John Flournoy <carne...@gmail.com> wrote:
>> > > On Feb 18, 10:19 am, The Name Forgotten <tex...@telkomsa.net> wrote:
>>
>> > > It's better than not having the rule, because it allows a judge to
>> > > sanction somebody who is clearly not playing to win, such as the
>> > > following scenario:
>>
>> > > Round 3 of a tournament. Player A has 1 GW, player B has 0VP for the
>> > > tournament.
>>
>> > > Without "play to win", Player B can play to give his buddy (player A)
>> > > the game win in the hopes that A will make the finals (which B won't
>> > > be able to do no matter what.)
>>
>> > Untrue. That kind of play is illegal for different reasons - in that
>> > it's tantamont to colluding (if not downright colluding).
>>
>> Nonsense. Colluding involves multiple people who are in collusion with
>> one another. If Player A had no intent to receive this boon then he'd
>> have done nothing wrong and he isn't colluding. Only Player B is causing
>> a problem and he isn't colluding, either, because he there's no person
>> for him to collude _with_.
>
> I've already agreed to this point. My counter point is that, to
> everyone else, it would appear as if you are colluding, and you'd have
> no way to prove them false, and they'd have every reason to suspect
> you.

No they wouldn't. They'd have every reason to see that B was not playing
to win but not that there was any collusion going on, as long as A was
playing to win himself. You can't - and therefore should not - be
suspected of colluding based on what some other player has done when
nothing you have done is wrong. It doesn't even matter what people
quote-"really"-unquote believe is going on. If player A doesn't look
like he's done anything wrong, then he hasn't. End of story.

> I agree that Player A should not be called up on it, because he
> hasn't directly done anything to change the game state in any way that
> differs from how he would normally play (unless, of course, he has).
> Player B on the other hand, has changed the game state to his own
> detriment to favour another player.

Right. And what you just demonstrated is exactly what everyone is
scream at you: that this is _not_ an example of collusion. It is
an example of Player B _not_ playing to win. And why we need a PTW
rule. A collusion rule won't do.

(Unless you accept LSJ's PoV, that "playing to win" ought to be
considered as part of the unspolken definition of "sportsmanship".
I can see the argument but I'm more comfortable with things more
explicitly defined than that.)

> At best, this would warrant a warning from the judge to Player B that
> such actions appear to be colluding,

It warrants some kind of action to the effect that Player B appears
not to be playing to win.

>> > You don't need a "play-to-win" rule to stop that from happening.
>>
>> You don't need a "play-to-win' rule to stop anything bad from happening.
>> You could add rules willy-nilly to proscribe each specific form of bad
>> behavior you can identify. But the "play-to-win" rule addresses a class
>> of bad behaviors based on their common theme, and addresses the core of
>> the bad thing in all of them. Hence, PTW is a good way to address
>> players arbitrarily helping their buddies' tournament placement to their
>> own detriment. It still has the weaknesses you dislike about it - but
>> creating individual rules around individual types of bad behaviors would
>> have the same problem. How would a more specific rule routine and
>> concretely identify when Player B is trying to help Player A to his own
>> detriment any better than PTW does?
>
> Well, as I said, it would be considered an act of colluding, whether
> or not the intent on the side of both players is present. The bigger
> problem, in my book, is how to stop players from doing nothing at all,
> or an absolute minimum. I think that it's the rarest kind of player -
> which is why I think it's better to abandon the PTW rule (at least the
> tournament part) which seems to affect far more situations.

I don't know what you mean by "hot to stop players from doing nothing
at all". Is this a real problem in your tournaments? Explain.

>> > As I said, there are situation where a player might actually play to
>> > lose, but I wonder if it will ever actually happen (at least in such a
>> > way that it doesn't violate another rule).
>>
>> You haven't answered the question of what other rules cover the same ground.
>> Collusion doesn't. You'd have to demonstrate that there are other such
>> rules, first.
>
> Collusion doesn't directly, agreed. Still, in your example, any judge
> would rule that it was collusion - having no way to prove that it
> wasn't

Huh?!? Totally disagree. No judge could call that collusion because
there would be no evidence of it. You can't simply conclude that because
Player B is doing something in Player A's interest that the two are
colluding. That's an assumption. Assumptions are not proof; ergo, no
collusion. No judge should rule that it is.

> You could probably formulate a rule to prevent that situation, but
> there would be no need, as, in practice, another rule already in
> existence would do the same job, even if it wasn't intended to be that
> way.

We HAVE a rule to prevent that situation. 4.8.

Fred


Juggernaut1981

unread,
Feb 18, 2009, 11:33:27 PM2/18/09
to
On Feb 19, 1:25 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:f8de0e5f-cf95-473b...@r15g2000prd.googlegroups.com...
> Fred- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -

From what I can see the entire thrust of the argument is intended to
be around the phrase including Game Wins as their own criteria for
PTW.
So, should PTW be rewritten to exclude GWs but to include VPs? The
intent of such a rule would be to say "If you can show you can't get a
VP, play as you want. Otherwise, you must play to win". Now I do not
know if that is the way Judges have practically applied PTW, in the
tournaments I have played in PTW hasn't really made an appearance.
But I do think that PTW should be set up to highlight a special case
from a default position. It does however flag you as a lost cause to
be "removed" from the table if you do call a judge over to adjudicate
that from your position you cannot win and can do whatever you want.
To be brutally honest, if there was a lost cause on the table and
removing them will balance the table out to increase my chances...
I'll take you out of the game myself.

Special Case = you cannot win, play how you want
Default position = play to gain VPs

(That could just also mean that in Sydney we're all nutters who will
rabidly chase our prey until we've burned our last pool.)

The Name Forgotten

unread,
Feb 19, 2009, 7:15:05 AM2/19/09
to
> -John Flournoy- Hide quoted text -

>
> - Show quoted text -

I've already agreed that it is not collusion.

However, I have also stated that because it so closely resembles
collusion as to be impossible to tell it apart from collusion, it
would certainly warrant a judge to warn (or technically caution) both
players that their play manifests in exactly the same way collusion
would, and should either party continue to play it such a manner
(fully knowing that it would be perceived as collusion) that it would,
in fact, be treated as such.

Given such a warning (caution), if this sort of play continues on both
of their sides, then they should both be penalised for collusion.

However, given the fact that D appears to have heeded the warning
(caution) by the fact that he has done nothing different than he would
in ordinary play, if A continues to act in such a way as a colluding
player would, then he alone should be penalised for it.

Still, I agree that there are far better ways of dealing with this
corner-case situation that could deal with the matter directly, rather
than fall back on the 'appearance' of collusion. I will hazard an
attempt at a better rewriting of 'play-to-win', even though it'll
likely be a futile attempt. I'm sure there are people who could write
up an even better version than I could - people with resources and
time who are in a position to do such things. I'm just suggesting that
the current ruling is not a good one, it should be dropped, and that
alternatives should be looked into.

The Name Forgotten

unread,
Feb 19, 2009, 7:29:47 AM2/19/09
to

True, you'd caution both - and if one of them appears not to have
taken your caution seriously - you apply penalties to that person - if
both haven't heeded your caution, you apply penalties to both. If
however, one appears to attempt to rectify the situation, and gives
you every evidence that he is not attempting to play in any way that
is in conflict with the rules of the game, then he should, by no
means, be penalised further.

For example, If Player A & Player B have colluded, and the judge is
given the exact same play, the judge (never being able to judge intent
with certainty) will caution both players. Player A heeds the warning,
and plays without any further acts of collusion. Player B, on the
other hand, ignores the warning and continues to play in the same
manner (that being, as if Player A were still colluding with him) -
Player B should be penalised for collusion and Player A need not be.

> (You can certainly potentially warn B for other reasons, but "helping
> another player" is not at all the same thing as "collusion".)

The only relevant way that it is different is in intent. As I said,
you will never be able to determine intent, as you will never be able
to be inside their heads.

The Name Forgotten

unread,
Feb 19, 2009, 7:39:33 AM2/19/09
to

Agreed. Except that this seems to be a case where the SAME judge would
rule differently in both cases - even though the risk analysis would
lead to the exact same result.

While I agree that a judge could be hardline about it and rule the
same way in both cases - my general perception is that almost all
judges (and players) would rule differently - they'd give far more
lenience to the person who could get a potential GW then they would to
the person who could get a potential VP. While the rule itself isn't
explicit on this double standard - the vagueness of what is
'reasonable', the general determination to have players continue games
in a competitive fashion even though they do not have a 100% certain
chance of getting a VP (from their point of view), and the lack of
determination to have players go for GWs at the risk of potential VPs
seem to lead this rule to be used as a double standard.

The Name Forgotten

unread,
Feb 19, 2009, 7:57:51 AM2/19/09
to
On Feb 19, 2:53 am, John Flournoy <carne...@gmail.com> wrote:
> On Feb 18, 5:54 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
>
> > That's why I'm asking for the rule to be dropped. I understand why
> > people might ask for it to be replaced with something, although I'm
> > not convinced that it does - most of the problems is already covered
> > in other sections, and with very little re-editing, you could cover
> > them all.
>
> If you have a specific idea on how such an edit could work to cover it
> all, that'd be welcome I'm sure (at least for consideration and
> further discussion.)
>

I've highlighted the changes I'd make with the ***. It seems to be a
better alternative to me, but I'm sure it can be refined alot more
with constructive criticism. My challenge to you is to find ways that
it is problematic, and see if there are possible additions that could
be made (excluding any attempt to go back to the original, of course).


4.8. Play to Win

One aspect of sportsmanlike conduct is that players must not play
toward goals that conflict with the goal of the game as stated in the
V:TES rulebook (e.g., attacking certain players on the basis of their
V:EKN ratings or overall tournament standing, etc.).

*** or attacking certain players on the basis of out of game
considerations, e.g. race, gender, animosity, etc. Also, a player will
not play to benefit certain players on the basis of their V:EKN
ratings, overall tournament standings, or other out-of-game
considerations, eg. friendship, fame, lack of time etc.

*** For tournaments, play to win also means that a player make a
reasonable effort to play cards in any given round of the tournament.

Neither the basic game rules nor the tournament rules enforce or
regulate deals made between players. The tournament rules acknowledge
deals, however, in that a deal which represents the best interests of
the players involved at the time the deal is made is allowed to be
honored, even when the normal play to win rule would indicate that a
deal should be broken. This only applies to deal that are in the best
interests of the players involved at the time the deal is made. That
is, it applies only when making the deal is playing to win. (It is
also allowable to break such a deal, of course).

Exception: when only two Methuselahs remain, the tournament rules no
longer acknowledge any deals. Prior deals are voided, even if they
were play to win when made. When only two Methuselahs remain, both
Methuselahs must play to win based only on game state, without regard
to any deals.

LSJ

unread,
Feb 19, 2009, 8:09:10 AM2/19/09
to
The Name Forgotten wrote:
> I've highlighted the changes I'd make with the ***. It seems to be a
> better alternative to me, but I'm sure it can be refined alot more
> with constructive criticism. My challenge to you is to find ways that
> it is problematic, and see if there are possible additions that could
> be made (excluding any attempt to go back to the original, of course).
>
> 4.8. Play to Win
>
> One aspect of sportsmanlike conduct is that players must not play
> toward goals that conflict with the goal of the game as stated in the
> V:TES rulebook (e.g., attacking certain players on the basis of their
> V:EKN ratings or overall tournament standing, etc.).
>
> *** or attacking certain players on the basis of out of game
> considerations, e.g. race, gender, animosity, etc. Also, a player will
> not play to benefit certain players on the basis of their V:EKN
> ratings, overall tournament standings, or other out-of-game
> considerations, eg. friendship, fame, lack of time etc.

This is just adding to the examples, not adding to the rules, given above.

That is, this is no change.

Unless you're saying that the player's otherwise-equal-in-game choices are to be
artificially limited by out-of-game criteria (given the choice in a lost
position of helping either his prey, Jessica Alba, or his predator, Joe Blow,
the player must not help Jessica Alba, given the out-of-game criterion that she
is more famous than Joe Blow).

That would just cause more problems, especially when every situation/choice can
be described by interested parties as possibly being influenced by some
out-of-game consideration.

> *** For tournaments, play to win also means that a player make a
> reasonable effort to play cards in any given round of the tournament.

Eh?

What does this accomplish?

Adding more goals (e.g., move cards, gain blood, take actions, block actions,
torporize vampires, burn locations, etc.) doesn't help the situation. It hurts
the situation.

Jozxyqk

unread,
Feb 19, 2009, 8:16:37 AM2/19/09
to
LSJ <vte...@white-wolf.com> wrote:
> Unless you're saying that the player's otherwise-equal-in-game choices are to be
> artificially limited by out-of-game criteria (given the choice in a lost
> position of helping either his prey, Jessica Alba, or his predator, Joe Blow,
> the player must not help Jessica Alba, given the out-of-game criterion that she
> is more famous than Joe Blow).

If Jessica Alba plays VTES, *everybody* wins!

Kevin M.

unread,
Feb 19, 2009, 8:16:22 AM2/19/09
to
The Name Forgotten wrote:
> Maybe Peter's 'always go for your prey' approach
> is needed for these rules 8)

It sounds like you don't know how to play the game yet.

Have you read the rulebook for "Object of the Game", 2.2.1, 9.9.1, and
"Rules Glossary -- Prey"? I'd suggest that you do, so that you learn the
most efficient, most direct way to do well in VTES -- ousting your prey.

Or, as a wise man once said:
"You should spend more time figuring out how to win instead of figuring out
how to lose."


Kevin M., Prince of Las Vegas
"Know your enemy, and know yourself; in one-thousand battles
you shall never be in peril." -- Sun Tzu, *The Art of War*
"Contentment... Complacency... Catastrophe!" -- Joseph Chevalier
Please visit VTESville daily! http://vtesville.myminicity.com/
Las Vegas NAQ 2009! http://members.cox.net/vtesinlv/


The Name Forgotten

unread,
Feb 19, 2009, 8:26:51 AM2/19/09
to

I will try and look for such examples in this newsgroup. I have stated
that my perception may be incorrect, and that judges are applying the
exact same risk-assessment unilaterally to both cases (even though
different judges may use different thresholds for that limit). My
common sense and instincts guide me towards being more lenient towards
the going for GW scenario as opposed to the going for VP scenario,
but, for the rule not to have a double standard, I would need to be as
firm with both case, given the same percentage chance at either the GW
or VP. It would be a difficult pill to swallow - but, at least, it
would serve as a guideline for future judges.

(In that, they'd have to apply the same standard of risk-analysis to a
player who wants to self-oust instead of taking a chance at a VP, as
they would to the person who wants to go for a VP instead of taking a
chance at a GW).

> > The double standard problem remains - why is it perfectly fine in one
> > situation to force a person to take a 5% chance at a VP, but it's not
> > perfectly fine to force a 5% chance at a GW? If 5% is too low, then
> > increase the number to whatever you like. The fact remains that the
> > cutoff for the first situation seems to be much lower than the cutoff
> > for the second - at least in the general understanding of the rule.
> > I'm hoping that there is some other factor that can explain this
> > difference, but at the moment I'm failing to find it. If one cannot be
> > found, then I think the rule should be dismissed, at least as a hard
> > and fast rule with potential judgements and penalties attached.
>
> It's not a "hard and fast rule", because it explicitly relies on the
> judgement of the judge(s) as to what a "reasonable chance" is. It is
> in fact a very soft rule, as a result.

Yes, but the judgement of the judge(s) should be held unilaterally and
without bias for certain situations. A judge must use the same degree
of "reasonable chance" that affects one situation in a similar
situation that the judge is judging, otherwise you open the door to
gross misconduct.

For example, a judge rule on one table that a player should be
penalised and disqualified for playing with marked cards, whereas on a
second table the same judge rules that another player should be
warned, and given time to sleeve their deck, when they are found to be
playing with marked cards, even though the decks are equally marked
(heck, it could even be the same deck). Although such an incident
could be granted (even though it could be contested and brought up)
should there be different judges involved, as each judge has leeway to
determine the necessary penalty - the same judge making such differing
judgements on the same situation is definitely problematic.

> And again, I simply have never, ever, ever, ever in any tournament
> I've attended seen a player with 0 VP forced to take a chance at
> getting 1 VP and/or prohibited from self-ousting unless they were very
> clearly (to the judges, onlookers and other players) still very much
> capable of continuing. I've never seen a judge say "you still have a
> reasonable chance to continue" and seen anyone disagree with that
> ruling for more than the briefest moment once given, either at the
> moment or after the tournament was done. And I can't recall it ever
> being mentioned on the tournament reports on the newsgroups or in war-
> story discussions with other players, either.
>
> I _have_ seen judges called over and asked to verify that a player
> didn't have a reasonable chance, and seen the judge say "I agree, his
> chance is not reasonable" after an aside discussion (involving things
> along the lines of "I know what's in my deck, and I lack what I need
> to ever get past the defense my prey has showing clearly on the
> table") - I believe I have even seen LSJ uphold a "no reasonable
> chance" once.
>
> Which isn't to say it doesn't occur - I'm just saying that in my
> experience, it's very much not "common practice" or "general
> understanding" or any sort of routinely applied double standard.

I agree. My experience certainly isn't all inclusive and is based on
my perceptions. My perceptions could be flawed, and if that is the
general case world-wide, then I withdraw my objections. I thought
bringing this matter up on an international forum might give me the
world-wide perception of whether or not this is the case. So far,
you're the only one who has given me a real-world experience response
to this. I'll look up other posts and see if I can show where the
standard for self-ousting seems to lie (admittedly I tend to skim over
those posts). Generally though, my gut reaction is to say that there
is one.

> Feel free to provide anecdotes based on your experiences, though, as
> specific examples of how these rules get applied in ways that the
> people debating you in this thread seem not to experience would be
> helpful.
>

I agree. I'll try (when I'm not busy responding ;)

The Name Forgotten

unread,
Feb 19, 2009, 9:53:54 AM2/19/09
to
On Feb 19, 3:24 am, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > Great. Ok, to repeat myself again (which at this point I'm very tired
> > of doing).
>
> Preaching to the choir, here.
>
> > The double standard is this: in one situation taking a 5% (or whatever
> > percentage you're comfortable with) is considered a reasonable chance
> > of getting a VP, so the PTW rule to maximise VP (or at least how
> > people have been playing it) says you shouldn't self-oust, however, in
> > a similar situation you have exactly the same percentage chance of
> > getting a GW, but this time it is not considered a reasonable chance,
> > so the PTW rule to maximise GW says you don't need to, as you do not
> > have a reasonable chance. However - the percentage chance is the same
> > in both situations and the rule is the same one - i.e. what PTW means
> > in a tournament. It is thus a double standard.
>
> Ah, that's not a double standard.
>
> This is where descriptions come in useful -- to help uncover such non-standard
> usages of terms. Now we're getting somewhere.
>
> In practice, this is called judgment.
>
> But more to the point: in practice, you cannot calculate a percentage.
>
> You just have to go with reasonable.
>
> That's not a double standard.
>
> Even the fact that different judges may judge it differently is not a double
> standard. That's just the definition of judgment.

I've responded to this argument:

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/3d8505f02504be25?hl=en

>
> And doing away with the PTW clarification of the sportsmanship rule will not do
> away with judgments (and thus, will not do away with what you call a double
> standard).
>
> Indeed, doing away will the whole of the sportsmanship rule will not accomplish
> this (not that we'd ever do that, since it's such a bad idea for other reasons).
> There's plenty left for judges to judge: stalling, was that action actually
> declared, was that vote actually cast, how to fix a forgotten Army of Rats, and
> so on. (All of which lead to your non-standard double standards)

The point is that the SAME judge will rule differently based on the
SAME percentage of risk assessment based on the PTW-clarification of
the sportsmanship rule. That is a double standard (unless there is
some other standard to measure by).

While I agree that you cannot make such an accurate assessment in real-
life play - my point is that even if you could, it appears to me that
almost all judges would act more leniently towards the situation I
raised in:

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/1ff55b7b840b7d5d?hl=en

The Name Forgotten

unread,
Feb 19, 2009, 9:58:17 AM2/19/09
to
On Feb 19, 3:28 am, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 19, 1:29 am, LSJ <vtes...@white-wolf.com> wrote:
> >> 1) Your suggestion is to do away with the entire sportsmanship requirement?
> >> I thought you were only suggesting doing away with the "must play to win" part.
>
> > No, as I've stated several times before only the part that states:
> > 4.8. Play to Win
> > For tournaments, playing to win means playing to get a Game Win
> > if it is reasonably possible, and when a Game Win is not reasonably
> > possible, then playing to get as many Victory Points as possible.
>
> As I've stated several times before, that part is a redundant, superfluous,
> unnecessary restatement of a portion of the general sportsmanship requirement.

And yet, it alone is what people point to in terms of 'rules' to tell
people that they should continue playing instead of self-ousting. It
seems wholly necessary for those arguments.

> >> 2) This would then just be a matter changing the name by which the same rules
> >> are enforced. Change for change's sake. So no need to do it.
>
> > God no. It would remove all the problems.
>
> Removing the redundant clarification changes nothing.

It has the very clear effect of not having people point to it and say
that it's a standard which must be followed.

The Name Forgotten

unread,
Feb 19, 2009, 10:36:32 AM2/19/09
to
On Feb 19, 4:25 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:f8de0e5f-cf95-473b...@r15g2000prd.googlegroups.com...

Well, then how would a judge determine that collusion is taking place?
If you have to have evidence like a written agreement or the
equivalent, it would be far too easy for people to meet up on the sly,
collude with each other, and play the game pretending that they made
no such agreement. Sure, play-to-win covers some of these situations,
but if it covered all of them, there'd be no need for a collusion rule
(such as 2 vote deck players agreeing to work together and split the
tables).

The only reasonable way a judge can suspect and call up collusion is
by the way in which people play - if people start playing in such a
way that it benefits another player to their own detriment, then that
certainly appears to be collusion. However, without a written
agreement or the equivalent, it would be overly harsh to simply
outright judge them guilty. Call them up on it - say to them it looks
like collusion - which they will likely vehemently deny, or make
excuses for their actions (unless they scummy enough to collude, but
honest enough to admit to it when they're asked). Then whoever still
acts in a collusive manner should be penalised for doing so.

Still, I gave a simplier response and clarifaction that can get rid of
the problematic clause in the PTW section, and still avoid this
situation:

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/8c5ceb5fdf552051?hl=en

I admit, it's not perfect, but it's a start.

> > I agree that Player A should not be called up on it, because he
> > hasn't directly done anything to change the game state in any way that
> > differs from how he would normally play (unless, of course, he has).
> > Player B on the other hand, has changed the game state to his own
> > detriment to favour another player.
>
> Right.  And what you just demonstrated is exactly what everyone is
> scream at you: that this is _not_ an example of collusion.  It is
> an example of Player B _not_ playing to win.  And why we need a PTW
> rule.  A collusion rule won't do.

Yes, I've already agreed repeatedly that it is not collusion by the
strict definition of it - but the example I gave above could have
played out exactly like that if people were colluding - in fact, I'd
go so far as to say that there's an equally likely chance of it
happening if players were colluding as if they were not.

> (Unless you accept LSJ's PoV, that "playing to win" ought to be
> considered as part of the unspolken definition of "sportsmanship".
> I can see the argument but I'm more comfortable with things more
> explicitly defined than that.)

Me neither.

> > At best, this would warrant a warning from the judge to Player B that
> > such actions appear to be colluding,
>
> It warrants some kind of action to the effect that Player B appears
> not to be playing to win.
>

Also agreed. My point is merely that what you, as a judge, would use
for your case to prove that a player is colluding is exactly the same
evidence you would use in this case where B appears not to be playing
to win (although in the example, he is doing more than that, B is
playing to make A win).


> >> > You don't need a "play-to-win" rule to stop that from happening.
>
> >> You don't need a "play-to-win' rule to stop anything bad from happening.
> >> You could add rules willy-nilly to proscribe each specific form of bad
> >> behavior you can identify.  But the "play-to-win" rule addresses a class
> >> of bad behaviors based on their common theme, and addresses the core of
> >> the bad thing in all of them. Hence, PTW is a good way to address
> >> players arbitrarily helping their buddies' tournament placement to their
> >> own detriment.  It still has the weaknesses you dislike about it - but
> >> creating individual rules around individual types of bad behaviors would
> >> have the same problem.  How would a more specific rule routine and
> >> concretely identify when Player B is trying to help Player A to his own
> >> detriment any better than PTW does?
>
> > Well, as I said, it would be considered an act of colluding, whether
> > or not the intent on the side of both players is present. The bigger
> > problem, in my book, is how to stop players from doing nothing at all,
> > or an absolute minimum. I think that it's the rarest kind of player -
> > which is why I think it's better to abandon the PTW rule (at least the
> > tournament part) which seems to affect far more situations.
>
> I don't know what you mean by "hot to stop players from doing nothing
> at all".  Is this a real problem in your tournaments?  Explain.

"I think that it's the rarest kind of player "

In fact, I have never seen someone join a tournament and do absolutely
nothing, nor do I think I ever will. In fact, I'd be pretty surprised
if anyone encountered a player like this. Although, without the
motivation to go for VP's or GW's, a player would be perfectly
entitled to do so (although there might already be another rule that
directly addresses this issue that I might be unaware of).

> >> > As I said, there are situation where a player might actually play to
> >> > lose, but I wonder if it will ever actually happen (at least in such a
> >> > way that it doesn't violate another rule).
>
> >> You haven't answered the question of what other rules cover the same ground.
> >> Collusion doesn't.  You'd have to demonstrate that there are other such
> >> rules, first.
>
> > Collusion doesn't directly, agreed. Still, in your example, any judge
> > would rule that it was collusion - having no way to prove that it
> > wasn't
>
> Huh?!?  Totally disagree.  No judge could call that collusion because
> there would be no evidence of it.  You can't simply conclude that because
> Player B is doing something in Player A's interest that the two are
> colluding.  That's an assumption.  Assumptions are not proof; ergo, no
> collusion.  No judge should rule that it is.

Many rulings are based on assumptions. It just a matter of whether or
not the assumptions are reasonable, given the evidence. For example -
a player might be caught with a marked card - it's up to the judge to
decide whether or not it was purposefully done so - it would be very
difficult to prove intent other than by basing it on the actions of
said player. If the card always appeared at the top of his deck, you'd
have every right to say that it was an intentionally marking and
penalise accordingly (along with other penalties he might have
incurred). If the card appeared randomly in play, and the player
looked surprised when it was pointed out by a person cross table from
him, then it's less likely that he should be penalised for it,
although he might be asked to swap out the card, or some other
reasonable action.

The proof of a players intention can only be proved by the way they
play, and the reasonable assumptions that arise from it. When one
player actively helps another to win, at the detriment to his own
game, a judge can reasonably assume that the players are colluding -
although, based on the lack of certainty, and possible innocent
explanations of that game play, the judge would probably only caution
the players (whereas a signed agreement to collude might warrant an
outright disqualification).

> > You could probably formulate a rule to prevent that situation, but
> > there would be no need, as, in practice, another rule already in
> > existence would do the same job, even if it wasn't intended to be that
> > way.
>
> We HAVE a rule to prevent that situation.  4.8.
>

Again, not a helpful comment in the context of an argument that says:
we don't need 4.8 to stop players from going to tournaments and
helping their buddies win.

The Name Forgotten

unread,
Feb 19, 2009, 10:40:59 AM2/19/09
to
> rabidly chase our prey until we've burned our last pool.)- Hide quoted text -

>
> - Show quoted text -

This point was raised and answered in this thread already.
Unfortunately I don't have the time to find it now. Maybe someone else
will be kind enough to point you to the exact post.

The Name Forgotten

unread,
Feb 19, 2009, 10:52:47 AM2/19/09
to
On Feb 19, 3:09 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > I've highlighted the changes I'd make with the ***. It seems to be a
> > better alternative to me, but I'm sure it can be refined alot more
> > with constructive criticism. My challenge to you is to find ways that
> > it is problematic, and see if there are possible additions that could
> > be made (excluding any attempt to go back to the original, of course).
>
> > 4.8. Play to Win
>
> > One aspect of sportsmanlike conduct is that players must not play
> > toward goals that conflict with the goal of the game as stated in the
> > V:TES rulebook (e.g., attacking certain players on the basis of their
> > V:EKN ratings or overall tournament standing, etc.).
>
> > *** or attacking certain players on the basis of out of game
> > considerations, e.g. race, gender, animosity, etc. Also, a player will
> > not play to benefit certain players on the basis of their V:EKN
> > ratings, overall tournament standings, or other out-of-game
> > considerations, eg. friendship, fame, lack of time etc.
>
> This is just adding to the examples, not adding to the rules, given above.
>
> That is, this is no change.

Like you said, this would be a clarification - which, in effect, would
have a real consequence in gameplay and judgement calls - that is, it
could cause a change in gameplay. If it conflicts with the rules
'given above', then the rules 'given above' should change to include
such examples.

>
> Unless you're saying that the player's otherwise-equal-in-game choices are to be
> artificially limited by out-of-game criteria (given the choice in a lost
> position of helping either his prey, Jessica Alba, or his predator, Joe Blow,
> the player must not help Jessica Alba, given the out-of-game criterion that she
> is more famous than Joe Blow).
>
> That would just cause more problems, especially when every situation/choice can
> be described by interested parties as possibly being influenced by some
> out-of-game consideration.

This sounds exactly like an argument I made before in another post. It
wouldn't cause any problems, as someone would have to prove that you
helped Jessica Alba in direct conflict with the rules - if you're odds
are equal in game terms, then who you choose could be chosen by any
criterion, as the choice does not conflict with any rules.

>
> > *** For tournaments, play to win also means that a player make a
> > reasonable effort to play cards in any given round of the tournament.
>
> Eh?
>
> What does this accomplish?
>
> Adding more goals (e.g., move cards, gain blood, take actions, block actions,
> torporize vampires, burn locations, etc.) doesn't help the situation. It hurts
> the situation.

Yeah, I'm not too sure on the wording of that, I'm sure someone can
write it better than me. I just want to stop someone from sitting at a
table and doing nothing the entire game (or just discard cards or
whatever). Although, to be honest, I think just being in the
tournament should be enough to assume that the person is going to play
his deck (in the same way that being in the final table should give
you enough of a reason to assume the person is playing to win). It's
not really necessary.

Frederick Scott

unread,
Feb 19, 2009, 11:23:49 AM2/19/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:98d0d8f2-afd5-4580...@n2g2000vba.googlegroups.com...

On Feb 19, 2:43 am, John Flournoy <carne...@gmail.com> wrote:
> > Short version: "Collusion" does not mean "player B is helping player
> > A", it means "A and B both agreed that this would occur and are both
> > guilty". You cannot reasonably sanction someone for collusion if
> > you're not also reasonably sanctioning the other participant(s) as
> > well.
>
> True, you'd caution both

No, you wouldn't. You don't caution a player (A) who has done nothing
wrong in any way.

And I hope you never judge. Or if you do, I hope you never have to judge
this particular situation - because your stubbornness about seeing culpability
in A's behavior indicates you'd screw it up.

Fred


Frederick Scott

unread,
Feb 19, 2009, 11:47:48 AM2/19/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:b587f8b8-fa04-4454...@m29g2000prd.googlegroups.com...

You'd look for evidence that both players are colluding.

> If you have to have evidence like a written agreement or the
> equivalent, it would be far too easy for people to meet up on the sly,
> collude with each other, and play the game pretending that they made
> no such agreement.

That may be true. None the less, this does not justify sanctioning people
who have nothing wrong. Player B may well be violating PTW, depending on
the exact situation. But Player A has done nothing wrong to that point and
thus cannot be cautioned, warned, or have any other sanction against him.

> Sure, play-to-win covers some of these situations,
> but if it covered all of them, there'd be no need for a collusion rule

Well, gee...that's what *I'VE* been saying, if you read my original response
to this situation. The situation you just posed runs afoul of PTW. Why
are you so hung up on the collusion thing? The collusion thing is only a
reinforcement to PTW and PTW is the much more generally written rule that
covers these kinds of situations much better. It's like you're trying to
argue by sowing confusion.

> The only reasonable way a judge can suspect and call up collusion is
> by the way in which people play - if people start playing in such a
> way that it benefits another player to their own detriment, then that
> certainly appears to be collusion. However, without a written
> agreement or the equivalent, it would be overly harsh to simply
> outright judge them guilty. Call them up on it - say to them it looks
> like collusion - which they will likely vehemently deny, or make
> excuses for their actions (unless they scummy enough to collude, but
> honest enough to admit to it when they're asked). Then whoever still
> acts in a collusive manner should be penalised for doing so.

Yep. So conclusion: the collusion rule is hard to enforce. PTW can also
be hard to enforce but not as hard.

So your original argument was: "We don't need PTW because PTW is covered
in other rules".

My response was: "What other rules?"

Your response was: "Collusion in the Sportsmanship rule".

Now, subsequent discussion uncovers one major weakness in proving collusion
when the beneficiary has actually done nothing wrong and you can't prove
that he and his benefactor actually colluded.

My response is: "Yep. That's a problem with trying to enforce a rule about
conclusion. PTW is actually the rule that got violated anyway so why even
worry about collusion in this case?"

Your response is: ...well...I can't even tell where you're now trying to go
with this whole argument at this point. Can you enlighten us?

> Still, I gave a simplier response and clarifaction that can get rid of
> the problematic clause in the PTW section, and still avoid this
> situation:

First of all, what "problematic clause"? You haven't shown that any clause
in PTW is "problematic" AFAICS. Let's establish that any clause in PTW is
problematic in a way that can be fixed, first.

> Also agreed. My point is merely that what you, as a judge, would use
> for your case to prove that a player is colluding is exactly the same
> evidence you would use in this case where B appears not to be playing
> to win (although in the example, he is doing more than that, B is
> playing to make A win).

So one obvious question is, how did we even start talking about the
collusion clause in the Sportsmanship rule? Answer: we started talking
about it - in preference to PTW - because *YOU* insisted that we didn't
need PTW because we had that clause. Now, I'm *TOTALLY* confused where
you're going with this whole argument. In the paragraph above, you seem
to be admitting that the collusion clause isn't worth bothering with and
that we need something like...rule 4.8, Play to Win.

> > > Collusion doesn't directly, agreed. Still, in your example, any judge
> > > would rule that it was collusion - having no way to prove that it
> > > wasn't
> >
> > Huh?!? Totally disagree. No judge could call that collusion because
> > there would be no evidence of it. You can't simply conclude that because
> > Player B is doing something in Player A's interest that the two are
> > colluding. That's an assumption. Assumptions are not proof; ergo, no
> > collusion. No judge should rule that it is.
>
> Many rulings are based on assumptions. It just a matter of whether or
> not the assumptions are reasonable, given the evidence. For example -
> a player might be caught with a marked card - it's up to the judge to
> decide whether or not it was purposefully done so -

There's a difference between assuming what the real physical evidence of a
marked card means vs. assuming that someone who has done nothing bad at
all might have, in fact, intended for something bad to happen and therefore
is culpable. As a judge, you can do the first thing, you can't do the
second thing. If you're judging and doing the second thing, please stop it.

> > > You could probably formulate a rule to prevent that situation, but
> > > there would be no need, as, in practice, another rule already in
> > > existence would do the same job, even if it wasn't intended to be that
> > > way.
> >
> > We HAVE a rule to prevent that situation. 4.8.
>
> Again, not a helpful comment in the context of an argument that says:
> we don't need 4.8 to stop players from going to tournaments and
> helping their buddies win.

I'd agree except for one thing: you're the person who has driven this
discussion around this U-turn, not me. You're the one who raised the
collusion clause as a reason we don't need 4.8 and then turned around and
demonstrated why the collusion clause is insufficient. I'm just saying,
"Yea, you're right. It *IS* insufficient. Fortunately, we have 4.8."

Fred


The Name Forgotten

unread,
Feb 19, 2009, 1:11:34 PM2/19/09
to
On Feb 19, 6:23 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:98d0d8f2-afd5-4580...@n2g2000vba.googlegroups.com...

> On Feb 19, 2:43 am, John Flournoy <carne...@gmail.com> wrote:
>
> > > Short version: "Collusion" does not mean "player B is helping player
> > > A", it means "A and B both agreed that this would occur and are both
> > > guilty". You cannot reasonably sanction someone for collusion if
> > > you're not also reasonably sanctioning the other participant(s) as
> > > well.
>
> > True, you'd caution both
>
> No, you wouldn't.  You don't caution a player (A) who has done nothing
> wrong in any way.

Before I continue, what do you mean by sanction? Telling a player you
suspect that there might be foul play (i.e. caution them) and you're
just going to check out how the game progresses to see if there is
actual foul play involved hardly seems like any sort of sanctioning to
me. To you, this seems to equate to "I know you're guilty of colluding
and I'm penalising you for it!"

I don't agree that both people *have* to be sanctioned. I do, however,
think that the situation warrants a closer view. In fact, Player B
(who self-sacrifices himself) could have secretly made a plan with
Player C, who is also Player A's friend, and who would prefer Player A
in the final with him (or whatever other good reason). As a judge, it
would be difficult for you to say Player A was colluding, if he simply
acts in a way that is considered normal play. Player B is the one
acting suspiciously. Player A would certainly be my first suspect for
collusion with Player B, and saying that something smells fishy, and
that you suspect that they might be colluding and so you want to see
what their on-table actions are like, and see if it matches the
behaviour of a colluding player, doesn't seem like anything overtly
harsh or 'sanctioning'. When player B continues to act like a player
who is colluding (by helping A win), even though you've expressly told
him that behaviour is identical to a colluder's behaviour, then you
can sanction him as such - although, without having any evidence from
gameplay (or any other kind of evidence) to suspect A had any real
involvement, and you're unsure of who else player B is colluding with
(or you don't have any evidence at all to support other suspicions),
you wouldn't sanction anyone else.

> And I hope you never judge.  Or if you do, I hope you never have to judge
> this particular situation - because your stubbornness about seeing culpability
> in A's behavior indicates you'd screw it up.

Not a useful comment. I avoid Ad Hominem arguments, so don't waste
your, my or other people's time by resorting to it. Whether or not I
judge at all, or judge any specific situation, or am stubborn, or even
make the occasional bad call doesn't have any relevance to this
argument. The fact that I'm trying to come to terms with this issue
through debate on this forum indicates that I'm a judge who tries to
fully comprehend the rules I'm enforcing rather than dogmatic
bludgeoning people over the head with them. That will be all I have to
say about that.

As for A's culpability - you give me an example clearly arguing that
he is not culpable - I say, given the evidence, I'd state my concerns
to the player, if he is culpable, then so be it, but, by your own
example, he would appear not to be culpable, and so shouldn't be
punished. That seems like a reasonable, careful approach to the
matter.

Frederick Scott

unread,
Feb 19, 2009, 3:48:40 PM2/19/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:6eeb723e-649d-45ed...@u39g2000prn.googlegroups.com...

> On Feb 19, 6:23 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > "The Name Forgotten" <tex...@telkomsa.net> wrote in
> > messagenews:98d0d8f2-afd5-4580...@n2g2000vba.googlegroups.com...
> > > On Feb 19, 2:43 am, John Flournoy <carne...@gmail.com> wrote:
> > > > Short version: "Collusion" does not mean "player B is helping player
> > > > A", it means "A and B both agreed that this would occur and are both
> > > > guilty". You cannot reasonably sanction someone for collusion if
> > > > you're not also reasonably sanctioning the other participant(s) as
> > > > well.
> >
> > > True, you'd caution both
> >
> > No, you wouldn't. You don't caution a player (A) who has done nothing
> > wrong in any way.
>
> Before I continue, what do you mean by sanction?

The applicable definition:

sanc - tion [sangk-shuhn]
...
-verb (used with object)
...
8. to impose a sanction on; penalize, esp. by way of discipline.

> Telling a player you
> suspect that there might be foul play (i.e. caution them) and you're
> just going to check out how the game progresses to see if there is
> actual foul play involved hardly seems like any sort of sanctioning to
> me.

It is an official sanction. There is no basis for it. Ergo, as a judging
action, it is improper. It is a mistake. Even were it an "unofficial"
caution, it's still a mistake. You caution people when their actions (note
*THEIR* actions, not someone else's actions) are on the border of being
illegal or look like they might have an element of illegality to them.
But Player A hasn't done anything. You can't caution him because he's
done nothing you can caution him *ABOUT*. What's he done that's on the
border of being illegal or appears to maybe have an element of illegality
to it? Nothing.

> I don't agree that both people *have* to be sanctioned. I do, however,
> think that the situation warrants a closer view.

As a judge, you are always free to apply your closest attention to
whichever games seem to warrant it. You do not need to officially
caution anyone to do this. Official cautions should be only be
applied where players' actual action warrent them. Player A did nothing
to warrant a caution; ergo, it would be an error to caution Player A.

> > And I hope you never judge. Or if you do, I hope you never have to judge
> > this particular situation - because your stubbornness about seeing culpability
> > in A's behavior indicates you'd screw it up.
>
> Not a useful comment. I avoid Ad Hominem arguments,

It is a very useful comment. It is not an ad hominem attack because
I believe it speaks to some of the errors you're making in your reasoniong
about PTW in general. I note that despite the fact that I've challenged
you on this repeatedly, we're still bickering about the collusion clause of
the Sportsmanlike Conduct rule and not simply conceding that the rule which
applies to this situation is PTW. Applied to Player B, only. That would
not error in judgement.

But aside from that, I'm sincere. You don't seem be able to distinguish
between a suspicious situation and suspicious behavior on the part of
someone involved in the suspicious situation. And as judge, you need
to be able to make such distinctions. If you ever judge any tournaments
I play in, I will feel better if I know you're thought point through
and understand the criticism of your reasoning lodged by John Flournoy
and myself.

> The fact that I'm trying to come to terms with this issue
> through debate on this forum indicates that I'm a judge who tries to
> fully comprehend the rules I'm enforcing rather than dogmatic
> bludgeoning people over the head with them.

OK, that part's commendable. But you seem to be stubbornly resisting
a very valid point and that isn't so commendable.

Fred


The Name Forgotten

unread,
Feb 19, 2009, 3:48:26 PM2/19/09
to
On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:b587f8b8-fa04-4454...@m29g2000prd.googlegroups.com...

That's why I said, caution the player and the most likely suspect -
and watch the game.

What if you could tell for certain that one person was colluding, but
you had no way of telling who the second person is? Would you throw up
your hands and say, "well, I can't tell who you are colluding with, so
I'm going to let you off"?

For example, what if you had a tape recording of two players
colluding, but you're unable to identify the one player. Would you let
the one you can identify get away with it because you couldn't
identify the other? This question points at the issue that you don't
need to sanction two (or more people) in order to sanction one for
colluding.

> > If you have to have evidence like a written agreement or the
> > equivalent, it would be far too easy for people to meet up on the sly,
> > collude with each other, and play the game pretending that they made
> > no such agreement.
>
> That may be true. None the less, this does not justify sanctioning people
> who have nothing wrong. Player B may well be violating PTW, depending on
> the exact situation. But Player A has done nothing wrong to that point and
> thus cannot be cautioned, warned, or have any other sanction against him.

I see your point. However, I don't, as a judge, need to caution, warn
or place any other sanction on Player A, unless I feel that his
actions warrant it. I could see B acting unreasonably favorably
towards A (to the point of helping him out). Say "this looks an awful
lot like collusion". If it appears that Player A's in-game actions or
words have given me and other around the suspicion that they may be
colluding together, than give a caution to both of them. If Player A's
in-game actions and words gives no impression that he is colluding,
then don't caution him.

If you have reason to believe that people are colluding by the actions
they have taken or the words they have spoken, you should caution them
at the very least. It means absolutely nothing if they are not
colluding, and hopefully squashes their idea of colluding further if
that was their initial plan. I really can't see the harm in it.

> > Sure, play-to-win covers some of these situations,
> > but if it covered all of them, there'd be no need for a collusion rule
>
> Well, gee...that's what *I'VE* been saying, if you read my original response
> to this situation.

Sorry, I'm not sure I understand - have you been saying that we should
keep the PTW rule, but there is no need for the collusion rule? I
don't think that's what you mean, but your statement makes it sound
that way.

> The situation you just posed runs afoul of PTW. Why
> are you so hung up on the collusion thing? The collusion thing is only a
> reinforcement to PTW and PTW is the much more generally written rule that
> covers these kinds of situations much better.

Well, if LSJ is to be followed - PTW is just a clarification on a
rule. Collusion is a rule. You're right though, I've spent far too
much time on the collusion thing that it actually detracts away from
the proper argument.

> It's like you're trying to
> argue by sowing confusion.

Sorry, I don't mean to confuse, I sometimes am too through for my own
good and I try to respond to unnecessary issues.

>
> > The only reasonable way a judge can suspect and call up collusion is
> > by the way in which people play - if people start playing in such a
> > way that it benefits another player to their own detriment, then that
> > certainly appears to be collusion. However, without a written
> > agreement or the equivalent, it would be overly harsh to simply
> > outright judge them guilty. Call them up on it - say to them it looks
> > like collusion - which they will likely vehemently deny, or make
> > excuses for their actions (unless they scummy enough to collude, but
> > honest enough to admit to it when they're asked). Then whoever still
> > acts in a collusive manner should be penalised for doing so.
>
> Yep. So conclusion: the collusion rule is hard to enforce. PTW can also
> be hard to enforce but not as hard.

Agreed.

> So your original argument was: "We don't need PTW because PTW is covered
> in other rules".

No. My original argument was we shouldn't have PTW because it leads to
a double standard.
My secondary argument was the only things we need PTW for is IMO
already covered by other rules - although I did say that my
understanding and perception of the issue is possibly limited. I asked
for counter examples.

Your response (or someones response) was to give such a counter
example.

My response was that the counter example is already covered by another
rule - that being collusion.

Dammit, I'm being meticulous again...

> My response was: "What other rules?"
>
> Your response was: "Collusion in the Sportsmanship rule".
>
> Now, subsequent discussion uncovers one major weakness in proving collusion
> when the beneficiary has actually done nothing wrong and you can't prove
> that he and his benefactor actually colluded.
>
> My response is: "Yep. That's a problem with trying to enforce a rule about
> conclusion. PTW is actually the rule that got violated anyway so why even
> worry about collusion in this case?"
>
> Your response is: ...well...I can't even tell where you're now trying to go
> with this whole argument at this point. Can you enlighten us?

My argument at this point was to agree that it was an issue, and there
is a better alternative to the Play-to-Win rule to stop this from
happening (that wouldn't lead to a double standard.)
But then I mistakenly went on a tangent as well. I said that the
collusion rule could still be used by a judge to slam a player who
plays only to benefit another player - basing this argument mostly on
the fact that while intent is necessary for collusion, you'd never be
able to prove intent as a judge, and the only way you can reasonably
expect to conclude the intent is through a players actions - in this
case, the actions would be identical to someone who was, in fact
colluding.

>
> > Still, I gave a simplier response and clarifaction that can get rid of
> > the problematic clause in the PTW section, and still avoid this
> > situation:
>
> First of all, what "problematic clause"? You haven't shown that any clause
> in PTW is "problematic" AFAICS. Let's establish that any clause in PTW is
> problematic in a way that can be fixed, first.

I have, in fact brought up this clause, and made my case in the very
first post of this thread.

After that, I've even given it an imperfect rewrite (mostly to expand
the examples) and to remove the specific issue that I find
problematic. A better version is what I've done below, but it is, by
no means, a perfect solution. I'll probably edit and refine it with
feedback.:

4.8. Play to Win


One aspect of sportsmanlike conduct is that players must not play
toward goals that conflict with the goal of the game as stated in the
V:TES rulebook (e.g., attacking certain players on the basis of their
V:EKN ratings or overall tournament standing, etc.).


*** or attacking certain players on the basis of out of game
considerations, e.g. race, gender, animosity, etc. Also, a player
will
not play to benefit certain players on the basis of their V:EKN
ratings, overall tournament standings, or other out-of-game
considerations, eg. friendship, fame, lack of time etc.


*** (Remove section about VP's & GW's)

Neither the basic game rules nor the tournament rules enforce or
regulate deals made between players. The tournament rules acknowledge
deals, however, in that a deal which represents the best interests of
the players involved at the time the deal is made is allowed to be
honored, even when the normal play to win rule would indicate that a
deal should be broken. This only applies to deal that are in the best
interests of the players involved at the time the deal is made. That
is, it applies only when making the deal is playing to win. (It is
also allowable to break such a deal, of course).


Exception: when only two Methuselahs remain, the tournament rules no
longer acknowledge any deals. Prior deals are voided, even if they
were play to win when made. When only two Methuselahs remain, both
Methuselahs must play to win based only on game state, without regard
to any deals.


>


> > Also agreed. My point is merely that what you, as a judge, would use
> > for your case to prove that a player is colluding is exactly the same
> > evidence you would use in this case where B appears not to be playing
> > to win (although in the example, he is doing more than that, B is
> > playing to make A win).
>
> So one obvious question is, how did we even start talking about the
> collusion clause in the Sportsmanship rule? Answer: we started talking
> about it - in preference to PTW - because *YOU* insisted that we didn't
> need PTW because we had that clause.

I didn't insist this, and certainly didn't say it was preferable to
PTW in this situation. I've stated my argument above already.

> Now, I'm *TOTALLY* confused where
> you're going with this whole argument. In the paragraph above, you seem
> to be admitting that the collusion clause isn't worth bothering with and
> that we need something like...rule 4.8, Play to Win.

I probably should've been more specific in the beginning, and I've
write about PTW - but really I'm only talking about the one tiny
section that talks specifically about what PTW means in a tournament -
i.e. maximising GW's then VP's. I continue to state that I'm perfectly
happy with the rest of 4.8 - it's only that one clause I have issue
with, which I'll clarify for the purposes of this response below:

"For tournaments, playing to win means playing to get a Game Win if it
is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible."

> > > > Collusion doesn't directly, agreed. Still, in your example, any judge


> > > > would rule that it was collusion - having no way to prove that it
> > > > wasn't
>
> > > Huh?!? Totally disagree. No judge could call that collusion because
> > > there would be no evidence of it. You can't simply conclude that because
> > > Player B is doing something in Player A's interest that the two are
> > > colluding. That's an assumption. Assumptions are not proof; ergo, no
> > > collusion. No judge should rule that it is.
>
> > Many rulings are based on assumptions. It just a matter of whether or
> > not the assumptions are reasonable, given the evidence. For example -
> > a player might be caught with a marked card - it's up to the judge to
> > decide whether or not it was purposefully done so -
>
> There's a difference between assuming what the real physical evidence of a
> marked card means vs. assuming that someone who has done nothing bad at
> all might have, in fact, intended for something bad to happen and therefore
> is culpable. As a judge, you can do the first thing, you can't do the
> second thing. If you're judging and doing the second thing, please stop it.

This is off of my topic as well. The point I was trying to make
(rather poorly), is that sometimes, as a judge, you have to decide
what physical evidence you need to prove intent on the players behalf.
If all the physical evidence (through gameplay etc.) is pointing to
the high likelihood that a player has got the requisite
"intention" (because you'll never prove it with a certainity unless
you are in his head), then you should sanction him for it. To ensure
that you make such a call correctly, it would be wise to make your
concerns known, because there is a small chance that despite all the
physical evidence pointing in one direction, it may still be an
unintentional coincidence. If the player is aware that all the
physical evidence is pointing towards his intention to do something
that is against the rules, he now has the option to stop giving more
evidence to further prove his intent by changing his behaviour.

For example, a judge could be called because someone noticed someone
was playing with a marked card and it appeared twice at the top of his
opponents deck at the start of 2 different games. The judge could
point this out - unsure of whether or not the player had intended to
do so or not. If, however, the situation is not rectified at the start
of the third game (by having a judge shuffle or something similar),
then it seems that the player clearly intends to cheat with his marked
card.

BTW, yes I know marked cards is against the rules - but there is a
different penalty associated with it if it is an intentional act, or
an unintentional one. Like unintentionally 'colluding' (or acting in a
way that appears to be colluding) has no penalty, but intentional
colluding has a penalty associated with it. Unintentionally playing
with marked cards (and no pattern) has a caution involved, but using a
marked card intentionally bears the penalty of cheating, which is
typically disqualification. You have to look at the game play to make
such a judgement of intent in this case, just as you would in the case
of collusion - intentional or not.

> > > > You could probably formulate a rule to prevent that situation, but
> > > > there would be no need, as, in practice, another rule already in
> > > > existence would do the same job, even if it wasn't intended to be that
> > > > way.
>
> > > We HAVE a rule to prevent that situation. 4.8.
>
> > Again, not a helpful comment in the context of an argument that says:
> > we don't need 4.8 to stop players from going to tournaments and
> > helping their buddies win.
>
> I'd agree except for one thing: you're the person who has driven this
> discussion around this U-turn, not me. You're the one who raised the
> collusion clause as a reason we don't need 4.8 and then turned around and
> demonstrated why the collusion clause is insufficient. I'm just saying,
> "Yea, you're right. It *IS* insufficient. Fortunately, we have 4.8."

Ok. You've made your point. I believe I've made mine.

The Name Forgotten

unread,
Feb 19, 2009, 4:35:19 PM2/19/09
to
On Feb 19, 10:48 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:6eeb723e-649d-45ed...@u39g2000prn.googlegroups.com...

Fine.

> Even were it an "unofficial"
> caution, it's still a mistake.

I don't find that telling people that there is a concern about
collusion happening at the table, and who it might involve is a
'mistake' - unless you go in there with the mindset of accusing people
of wrong-doing without investigating the matter properly. If you are
simply bringing up the concern, looking at the evidence, and saying
you'll look at future evidence, then I don't see a problem with that.
I do not mean 'caution' them in the sanctioning method, I even used
the word warn in the first few posts, which I also didn't mean in the
official sanctioning sense. You don't have to officially 'caution'
players to investigate a matter on a given table. You can simply say
that there is enough evidence to warrant that you stick around and see
if more evidence of colluding appears.

> You caution people when their actions (note
> *THEIR* actions, not someone else's actions) are on the border of being
> illegal or look like they might have an element of illegality to them.

I've already agreed to this. If A's actions don't look illegal, then
don't 'caution' him in any way, official, or unofficial. If there is
some reason to suspect that his actions may have an element of
illegality to it, then caution him - playing in such a way as to block
B's bleeds of A's prey seems pretty suspicious to me (which I think
was the original case to which I said both should be cautioned).

> But Player A hasn't done anything.  You can't caution him because he's
> done nothing you can caution him *ABOUT*.  What's he done that's on the
> border of being illegal or appears to maybe have an element of illegality
> to it?  Nothing.

I've already agreed to this.

>
> > I don't agree that both people *have* to be sanctioned. I do, however,
> > think that the situation warrants a closer view.
>
> As a judge, you are always free to apply your closest attention to
> whichever games seem to warrant it.  You do not need to officially
> caution anyone to do this.  Official cautions should be only be
> applied where players' actual action warrent them.  Player A did nothing
> to warrant a caution; ergo, it would be an error to caution Player A.

I've made my point above already.

> > > And I hope you never judge. Or if you do, I hope you never have to judge
> > > this particular situation - because your stubbornness about seeing culpability
> > > in A's behavior indicates you'd screw it up.
>
> > Not a useful comment. I avoid Ad Hominem arguments,
>
> It is a very useful comment.  It is not an ad hominem attack because
> I believe it speaks to some of the errors you're making in your reasoniong
> about PTW in general.  I note that despite the fact that I've challenged
> you on this repeatedly, we're still bickering about the collusion clause of
> the Sportsmanlike Conduct rule and not simply conceding that the rule which
> applies to this situation is PTW.  Applied to Player B, only.  That would
> not error in judgement.

I have already agreed to that. PTW is a better rule for dealing with
THAT situation. However, PTW, as it is written has other problems -
which was the main thrust of my argument, and so I was trying to look
at what would happen if the specific clause i.e. "For tournaments,


playing to win means playing to get a Game Win if it is reasonably
possible, and when a Game Win is not reasonably possible, then playing

to get as many Victory Points as possible," were removed what other
effects it may have on the game.

I'm playing Devil's advocate to see if we can use as much of the rules
currently in existence to overcome the problems that would appear to
arise out of abandoning such a 'clarification'. I've already agreed
repeatedly that collusion isn't the best answer to that issue (being
the helping a friend get a GW) - but it probably could be used to
overcome it.

> But aside from that, I'm sincere.  You don't seem be able to distinguish
> between a suspicious situation and suspicious behavior on the part of
> someone involved in the suspicious situation.  And as judge, you need
> to be able to make such distinctions.  If you ever judge any tournaments
> I play in, I will feel better if I know you're thought point through
> and understand the criticism of your reasoning lodged by John Flournoy
> and myself.
>

I understand it, but you are over-exaggerating my point (on the
collusion issue), and I'm trying to point that out. I think, given a
tournament situation, I would judge in exactly the manner you are
proposing, and that the judgment calls I make you'll be perfectly fine
with.

> > The fact that I'm trying to come to terms with this issue
> > through debate on this forum indicates that I'm a judge who tries to
> > fully comprehend the rules I'm enforcing rather than dogmatic
> > bludgeoning people over the head with them.
>
> OK, that part's commendable.  But you seem to be stubbornly resisting
> a very valid point and that isn't so commendable.
>

I'm not stubbornly resisting - I've agreed to your points already -
repeatedly I might add. The only part I'm arguing with you on is the
philosophical one of how do you prove 'intent'? I say you need to look
at the game state and judge accordingly - which is exactly what you
seem to be saying.

I agree that I sometimes make controversial points - I don't always
see every angle of every situation. I like to give every viewpoint a
fair go before dismissing it. I'm very grateful for your and Johns
counter-arguments - they are really good points (most of the time
8'p ) I've already dismissed standpoints I've made along the way, and
will continue to do so (thanks to your input). You and John have
definitely made the point that the Collusion rule would not be enough
if you remove the PTW clause I spoke about. I agree.

But you are only arguing about the issue of collusion and the 'friend
helping another friend get a GW' situation - and then saying that
there can't be anything wrong with the PTW rule in it's entirety
because it can be used to benefit in some situations.

My main point, that PTW is still problematic because of the double
standard associated with it, is what I'd really like to look at. What
are the alternatives available? Can we come up with a better
alternative? Relying on the other existing rules don't seem to be
enough, but there certainly seems to be a way of clarifying the
situation such that it can't lead to a double standard if you are
following your gut intuition.

BTW you'll be pleased to know that I very, very rarely make judgments
on my own - I usually consult several others to make sure that the
judgment is good and fair. I'll be honest and say that I've made a few
bad calls, but they've all been minor ones (and ones that no-one, at
the time, picked up). I never blindly make rulings without consulting
someone - and I certainly wouldn't sanction someone without others
agreeing. I'm not one to think that any single person should make
judgements without consulting anyone (unless, of course, that person
is LSJ). I make my ideas and theories known, and give people a fair
chance to respond. I'm very lucky to have a great group of players
here who don't cause any major problems (and certainly never
intentionally). My viewpoints are sometimes clouded by that.

But, I digress.

Kevin M.

unread,
Feb 19, 2009, 4:46:05 PM2/19/09
to
The Name Forgotten wrote:
> BTW you'll be pleased to know that I very, very rarely make
> judgments on my own - I usually consult several others to make
> sure that the judgment is good and fair.

and

> I never blindly make rulings without consulting someone - and
> I certainly wouldn't sanction someone without others agreeing.

and

> I'm not one to think that any single person should make judgements
> without consulting anyone (unless, of course, that person is LSJ).

OMG. Now I agree with Fred and hope you never, ever judge (although for
different reasons).

Frederick Scott

unread,
Feb 19, 2009, 5:02:38 PM2/19/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:130898ea-7fe0-4c63...@p23g2000prp.googlegroups.com...

> On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
>> > Sure, play-to-win covers some of these situations,
>> > but if it covered all of them, there'd be no need for a collusion rule
>>
>> Well, gee...that's what *I'VE* been saying, if you read my original response
>> to this situation.
>
> Sorry, I'm not sure I understand - have you been saying that we should
> keep the PTW rule, but there is no need for the collusion rule?

Yep. To tell the truth, until you quoted from the judge's manual, I had no
idea what the Collusion Clause in the Sportsmanship Rule was actually for.
It's written terribly, doesn't define what it means very well, and - as
written - could be applied to any two players who struck any deal in any
game in any tournament. Given the judge's manual, I now believe I understand
the spirit of that clause. But the clause itself should be clarified or
totally eliminated because its purpose is already covered by rule 4.8, PTW.
My preference would be the latter unless someone can demonstrate any good
reason to have it.

> I don't think that's what you mean, but your statement makes it sound
> that way.

That's what I mean, yes. One lesson to take away from all this: not all
who oppose you in a debate necessarily agree with each other. In face, every
individual will have their own take on a complex issue. Don't conflate my
position with, for instance, LSJ's. He'll probably thank you not conflate
his with mine - especially in view of the fact that his represents the
"official" way of ruling VEKN tournaments. My comments can't speak to that.

>> So your original argument was: "We don't need PTW because PTW is covered
>> in other rules".
>
> No. My original argument was we shouldn't have PTW because it leads to
> a double standard.

I have to confess, I've never understood your double-standard argument.
When you're challenged to relate it, you give an example so complex and
detailed, I lose the essence of why you think one situation would be
allowed and another not allowed. My suspicion is that I'd either disagree
that both were PTW violations or else that both were not PTW violations or
alternatively that I'd disagree that a correct judge should invoke the PTW
rule in one case and not the other.

It is possible that some PTW violations are much more difficult to detect
than others and to this extent, de facto double standards may exist. I'd
file that point under the heading, "it can't be helped." I'd still rather
have the rule than not have the rule, double standard and all. As long
as it never leads to accusing innocent players of violations that are
innately non-disprovable. I don't think this is an issue with PTW,
properly enforced.

> My secondary argument was the only things we need PTW for is IMO
> already covered by other rules - although I did say that my
> understanding and perception of the issue is possibly limited. I asked
> for counter examples.

I think the example you gave of Player B helping Player A was exactly
a counter example: neither player could be accused of collusion. But
Player B can be accused of voilating PTW. Of what else could he be
accused?

>> My response was: "What other rules?"
>>
>> Your response was: "Collusion in the Sportsmanship rule".
>>
>> Now, subsequent discussion uncovers one major weakness in proving collusion
>> when the beneficiary has actually done nothing wrong and you can't prove
>> that he and his benefactor actually colluded.
>>
>> My response is: "Yep. That's a problem with trying to enforce a rule about
>> conclusion. PTW is actually the rule that got violated anyway so why even
>> worry about collusion in this case?"
>>
>> Your response is: ...well...I can't even tell where you're now trying to go
>> with this whole argument at this point. Can you enlighten us?
>
> My argument at this point was to agree that it was an issue, and there
> is a better alternative to the Play-to-Win rule to stop this from
> happening (that wouldn't lead to a double standard.)

OK. I guess I can't respond to that because a double standard can't be
demonstrated, IMO. Well, again, maybe some double standards exist where one
violation is much more visible than another. But I'm challenged to
comprehend how we could ever fix that. PTW, by it's very nature, is hard to
define outside of how it's succintly defined. And it has no objective litmus
tests, or only completely worthless ones. Even so, it still fills an
important purpose.

>> > Still, I gave a simplier response and clarifaction that can get rid of
>> > the problematic clause in the PTW section, and still avoid this
>> > situation:
>>
>> First of all, what "problematic clause"? You haven't shown that any clause
>> in PTW is "problematic" AFAICS. Let's establish that any clause in PTW is
>> problematic in a way that can be fixed, first.
>
> I have, in fact brought up this clause, and made my case in the very
> first post of this thread.
>
> After that, I've even given it an imperfect rewrite (mostly to expand
> the examples) and to remove the specific issue that I find
> problematic. A better version is what I've done below, but it is, by
> no means, a perfect solution. I'll probably edit and refine it with
> feedback.:
>
> 4.8. Play to Win
>
> One aspect of sportsmanlike conduct is that players must not play
> toward goals that conflict with the goal of the game as stated in the
> V:TES rulebook (e.g., attacking certain players on the basis of their
> V:EKN ratings or overall tournament standing, etc.).
>
> *** or attacking certain players on the basis of out of game
> considerations, e.g. race, gender, animosity, etc. Also, a player
> will
> not play to benefit certain players on the basis of their V:EKN
> ratings, overall tournament standings, or other out-of-game
> considerations, eg. friendship, fame, lack of time etc.

When you toss in "other out-of-game" considerations, you've made
all the rest superfluous, AFAICS. In order to be a rule, you have
to define what is illegal behavior. As I've pointed out to you elsewhere,
examples do not define. They merely clarify definitions. I suppose you
could add a line that reads something like, "Examples of out-of-game
considerations include (but are not limited to): V:EKN ratings, overall
tournament standings, race, gender, animosity, friendship, fame, or
lack of time." But in fact, this addition adds nothing to the actual
rule.

> *** (Remove section about VP's & GW's)

Removing this, on the other, is a positive detriment to the rule. The
section about VPs and GWs is very important. It defines what, exactly,
is meant by "playing in goals which are in conflict with the goal of the
game as stated in the V:TES rulebook". Notice that playing towards
any of the goals cited about isn't actually illegal under 4.8. It's
only when such goals are *IN CONFLICT* with the rulebook goals that
a violation has occurred. But what are the rulebook goals with which
the other goals must not conflict? It's defined in that section! Why
are we getting rid of this crucial piece?

Nothing you added fixes the problem you just caused by removing the
section about VPs and GWs. What you added talks about the potentially
conflicting goals, not the ones which must not be conflicted.

Nor do you explain why you see that section as problematic. It's
actually the easiest part of the rule to understand. I don't see what
you have against it. I don't see why it causes any sort of double
standard. If anything, concrete, objective definition of the "rulebook
rules goals" will eliminate double standards, not cause them.

...


> I probably should've been more specific in the beginning, and I've
> write about PTW - but really I'm only talking about the one tiny
> section that talks specifically about what PTW means in a tournament -
> i.e. maximising GW's then VP's. I continue to state that I'm perfectly
> happy with the rest of 4.8 - it's only that one clause I have issue
> with, which I'll clarify for the purposes of this response below:
>
> "For tournaments, playing to win means playing to get a Game Win if it
> is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible."

OK. So why is this bad?

> This is off of my topic as well. The point I was trying to make
> (rather poorly), is that sometimes, as a judge, you have to decide
> what physical evidence you need to prove intent on the players behalf.
> If all the physical evidence (through gameplay etc.) is pointing to
> the high likelihood that a player has got the requisite
> "intention" (because you'll never prove it with a certainity unless
> you are in his head), then you should sanction him for it.

I don't agree. When the intention was not known, you should rule with
the understanding that the intention was not known. In the case of a
marked card, the evidence you _do_ have is that the player has a
marked card. If you have no evidence, you sanction him ONLY for having
the marked card; you can make no assumptions as to why it's marked.

Now, that said, mind you if a card is obviously marked, then we're
still talking about a pretty serious situation. It should be expected
that players can arrive at a tournament with an unmarked deck. And
when a player fails to do that, I think it's pretty reasonable to
make the penalties rather stiff. If you tolerate obviously marked
cards, the _potential_ for cheating - if not the actuality of it - is
bad. There's no reason players can't conform to this rule and should
be expected to and should be penalized when they fail to do that.

(Mind you, I'm talking about an obvious deliberate markings here. A
dog-eared card sleeve, for instance, could happen quite by accident
and would certain warrant a far milder sanction, lacking any
evidence of deliberated cheating.)

> For example, a judge could be called because someone noticed someone
> was playing with a marked card and it appeared twice at the top of his
> opponents deck at the start of 2 different games. The judge could
> point this out - unsure of whether or not the player had intended to
> do so or not. If, however, the situation is not rectified at the start
> of the third game (by having a judge shuffle or something similar),
> then it seems that the player clearly intends to cheat with his marked
> card.

Why not simply _tell_ the player that he has to fix the problem and
precisely what he has to do to fix it, in your judgement? Why even
wait to see it in a second game? Tell him the first time you see it.
And then inspect his fix. Then, there need be no future problem. That
seems a much more straightforward way to deal with this than anything else.

Fred


The Name Forgotten

unread,
Feb 20, 2009, 4:15:26 AM2/20/09
to
On Feb 19, 11:46 pm, "Kevin M." <youw...@imaspammer.org> wrote:
> The Name Forgotten wrote:
> > BTW you'll be pleased to know that I very, very rarely make
> > judgments on my own - I usually consult several others to make
> > sure that the judgment is good and fair.
>
> and
>
> > I never blindly make rulings without consulting someone - and
> > I certainly wouldn't sanction someone without others agreeing.
>
> and
>
> > I'm not one to think that any single person should make judgements
> > without consulting anyone (unless, of course, that person is LSJ).
>
> OMG.  Now I agree with Fred and hope you never, ever judge (although for
> different reasons).
>

Ok, I'm curious as to why you feel this way. Either start a seperate
post about this (as it'll go way off topic) or PM me about it. What
reasons do you have for hoping that I never judge? I have a feeling
you misunderstand my point, or that I've expressed it rather badly.
I'm trying to say that I don't come to a decision without first
listening to the opinions of different people first (or without
consulting a rule-book or errata list, or whatever else, which is,
essentially, consulting LSJ - maybe that's your issue??) Anyway, even
if I do consult a rule-book, etc, I still consult either the player in
question about a particular card-ruling or the table at whole to make
sure I (A) understand the query and (B) give an adequate response
(i.e. one that they understand). Perhaps you are deriving a different
meaning out of what I'm saying?

The Name Forgotten

unread,
Feb 20, 2009, 5:42:13 AM2/20/09
to
On Feb 20, 12:02 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:130898ea-7fe0-4c63...@p23g2000prp.googlegroups.com...

>
> > On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> >> > Sure, play-to-win covers some of these situations,
> >> > but if it covered all of them, there'd be no need for a collusion rule
>
> >> Well, gee...that's what *I'VE* been saying, if you read my original response
> >> to this situation.
>
> > Sorry, I'm not sure I understand - have you been saying that we should
> > keep the PTW rule, but there is no need for the collusion rule?
>
> Yep.  To tell the truth, until you quoted from the judge's manual, I had no
> idea what the Collusion Clause in the Sportsmanship Rule was actually for.
> It's written terribly, doesn't define what it means very well, and - as
> written - could be applied to any two players who struck any deal in any
> game in any tournament.  

No. It can't. It lists many exceptions. Namely:

Players participating in standard table talk or in-game agreements
should not be considered in violation of this rule as long as they
meet the following criteria:

No player introduces incentives outside the current game such as cash,
cards, or other items.
No part of the agreement has been secret or has taken place outside of
the current game.
No part of the agreement involves a random selection of the winner.
The agreement does not otherwise violate section 5.2 of the V:EKN
Tournament Rules.

> Given the judge's manual, I now believe I understand
> the spirit of that clause.  

Evidently you don't.

> But the clause itself should be clarified or
> totally eliminated because its purpose is already covered by rule 4.8, PTW.
> My preference would be the latter unless someone can demonstrate any good
> reason to have it.

It is already clarified.

>
> > I don't think that's what you mean, but your statement makes it sound
> > that way.
>
> That's what I mean, yes.  One lesson to take away from all this: not all
> who oppose you in a debate necessarily agree with each other.  In face, every
> individual will have their own take on a complex issue.  Don't conflate my
> position with, for instance, LSJ's.  He'll probably thank you not conflate
> his with mine - especially in view of the fact that his represents the
> "official" way of ruling VEKN tournaments.  My comments can't speak to that.

I don't conflate other people's arguments. I conflate my own to
include what everyone is saying. I know you're standpoint is different
from LSJ's. I only want my answer to be able to respond to both your
concern and LSJ's one.

> >> So your original argument was: "We don't need PTW because PTW is covered
> >> in other rules".
>
> > No. My original argument was we shouldn't have PTW because it leads to
> > a double standard.
>
> I have to confess, I've never understood your double-standard argument.

Therein lies the problem. If you cannot understand my issue with the
PTW, then any further debate is pointless. I'd rather spend the time
focusing on helping you understand my point there, before I argue on a
seperate issue that relies on that point.

> When you're challenged to relate it, you give an example so complex and
> detailed, I lose the essence of why you think one situation would be
> allowed and another not allowed.  

I don't. I think there should be one standard. However, enforcing this
single standard seems to be counter-intuitive to me, especially given
the 'people shouldn't self-oust unless they stand almost no chance of
getting a VP' stance.

> My suspicion is that I'd either disagree
> that both were PTW violations or else that both were not PTW violations or
> alternatively that I'd disagree that a correct judge should invoke the PTW
> rule in one case and not the other.

That's my point exactly. Still, I'd be going against my intuitions
about the spirit of PTW if I either enforced both violations or
disagreed that they were both violations - given that the relevant
risks were identical in terms of the rules. I'm guessing other judges
might feel similarly inclined, and might in fact, rule differently in
those situations (or at least want to).

> It is possible that some PTW violations are much more difficult to detect
> than others and to this extent, de facto double standards may exist.  I'd
> file that point under the heading, "it can't be helped."  I'd still rather
> have the rule than not have the rule, double standard and all.  As long
> as it never leads to accusing innocent players of violations that are
> innately non-disprovable.  I don't think this is an issue with PTW,
> properly enforced.

Agreed, but not my point.

> > My secondary argument was the only things we need PTW for is IMO
> > already covered by other rules - although I did say that my
> > understanding and perception of the issue is possibly limited. I asked
> > for counter examples.
>
> I think the example you gave of Player B helping Player A was exactly
> a counter example: neither player could be accused of collusion.  But
> Player B can be accused of voilating PTW.  Of what else could he be
> accused?
>

The point was taken. I went on a tangent (which I shouldn't of) about
how one determines intent as a judge.

>
> >> My response was: "What other rules?"
>
> >> Your response was: "Collusion in the Sportsmanship rule".
>
> >> Now, subsequent discussion uncovers one major weakness in proving collusion
> >> when the beneficiary has actually done nothing wrong and you can't prove
> >> that he and his benefactor actually colluded.
>
> >> My response is: "Yep.  That's a problem with trying to enforce a rule about
> >> conclusion.  PTW is actually the rule that got violated anyway so why even
> >> worry about collusion in this case?"
>
> >> Your response is: ...well...I can't even tell where you're now trying to go
> >> with this whole argument at this point.  Can you enlighten us?
>
> > My argument at this point was to agree that it was an issue, and there
> > is a better alternative to the Play-to-Win rule to stop this from
> > happening (that wouldn't lead to a double standard.)
>
> OK.  I guess I can't respond to that because a double standard can't be
> demonstrated, IMO.  

It can be demonstrated through a thought experiment. I'll try clarify
my point in a different way below.

> Well, again, maybe some double standards exist where one
> violation is much more visible than another.  But I'm challenged to
> comprehend how we could ever fix that.  

That is the exact challenge I'm proposing. I do understand that there
might not be a better alternative. I'd like to see if there is, and
like to see if other people can explore the issue. If there isn't then
fine - at least I've tried to find a better option, and can understand
why we need to have a less-than-perfect rule.

> PTW, by it's very nature, is hard to
> define outside of how it's succintly defined.  And it has no objective litmus
> tests, or only completely worthless ones.  Even so, it still fills an
> important purpose.

Agreed, it does.

Uhm... duh??? What does eg. mean to you? The "other out-of-game"
considerations is exactly the change that adds something to the actual
rule. I think I've lost your point here. Sure VEKN ratings and overall
tournament are further examples of out of game considerations.

> > *** (Remove section about VP's & GW's)
>
> Removing this, on the other, is a positive detriment to the rule.  The
> section about VPs and GWs is very important.  It defines what, exactly,
> is meant by "playing in goals which are in conflict with the goal of the
> game as stated in the V:TES rulebook".  

No, the rules in the V:TES rulebook define what the goals of the game
are.

> Notice that playing towards
> any of the goals cited about isn't actually illegal under 4.8.  It's
> only when such goals are *IN CONFLICT* with the rulebook goals that
> a violation has occurred.  

Which was my point to LSJ. I'm agree with you on this.

> But what are the rulebook goals with which
> the other goals must not conflict?  

Well - this is the rulebook goal with which the other goals must no
conflict (it's one of the first things in the rulebook):

Object of the Game
Your goal is to accumulate the most victory points by destroying the
influence held by rival Methuselahs. You accomplish this goal by using
your influence to gain control of younger vampires and using those
vampires to take actions to reduce the influence of rival Methuselahs.
Influence is represented by pool counters (see Equipment Needed, sec.
1.2), the main currency of the game. When a Methuselah runs out of
pool counters, she* is bereft of influence and ousted from the game.
As Methuselahs are ousted, players earn victory points; the winner is
the player with the most victory points at the end of the game (see
Victory Points, sec. 9.1).


> It's defined in that section!  Why
> are we getting rid of this crucial piece?

Because the Game Win section seems to create a double standard. (I'll
try explain this at the end).

Note that going for a Game Win is not an out-of-game consideration,
so, if anything, the Play-to-Win rule should include an exception
clause that allows for in-game considerations, such as it's already
present deal-making exception, for going for Game Wins as opposed to
VPs - which may be a problem in some corner-case examples.

> Nothing you added fixes the problem you just caused by removing the
> section about VPs and GWs.  What you added talks about the potentially
> conflicting goals, not the ones which must not be conflicted.

Sorry, I assumed that people were already familiar with the Object of
the Game, as stated in the rulebook.

> Nor do you explain why you see that section as problematic.  

I've tried and I'll try explain again.

> It's
> actually the easiest part of the rule to understand.  I don't see what
> you have against it.  I don't see why it causes any sort of double
> standard.  If anything, concrete, objective definition of the "rulebook
> rules goals" will eliminate double standards, not cause them.
>
> ...

Moot.

> > I probably should've been more specific in the beginning, and I've
> > write about PTW - but really I'm only talking about the one tiny
> > section that talks specifically about what PTW means in a tournament -
> > i.e. maximising GW's then VP's. I continue to state that I'm perfectly
> > happy with the rest of 4.8 - it's only that one clause I have issue
> > with, which I'll clarify for the purposes of this response below:
>
> > "For tournaments, playing to win means playing to get a Game Win if it
> > is reasonably possible, and when a Game Win is not reasonably
> > possible, then playing to get as many Victory Points as possible."
>
> OK.  So why is this bad?

Will explain below.

> > This is off of my topic as well. The point I was trying to make
> > (rather poorly), is that sometimes, as a judge, you have to decide
> > what physical evidence you need to prove intent on the players behalf.
> > If all the physical evidence (through gameplay etc.) is pointing to
> > the high likelihood that a player has got the requisite
> > "intention" (because you'll never prove it with a certainity unless
> > you are in his head), then you should sanction him for it.
>
> I don't agree.  When the intention was not known, you should rule with
> the understanding that the intention was not known.  In the case of a
> marked card, the evidence you _do_ have is that the player has a
> marked card.  If you have no evidence, you sanction him ONLY for having
> the marked card; you can make no assumptions as to why it's marked.
>
> Now, that said, mind you if a card is obviously marked, then we're
> still talking about a pretty serious situation.  It should be expected
> that players can arrive at a tournament with an unmarked deck.  And
> when a player fails to do that, I think it's pretty reasonable to
> make the penalties rather stiff.  If you tolerate obviously marked
> cards, the _potential_ for cheating - if not the actuality of it - is
> bad.  There's no reason players can't conform to this rule and should
> be expected to and should be penalized when they fail to do that.
>
> (Mind you, I'm talking about an obvious deliberate markings here.  A
> dog-eared card sleeve, for instance, could happen quite by accident
> and would certain warrant a far milder sanction, lacking any

> evidence of deliberated ...

Welcome to my point - how does a judge decide what an obvious
deliberate marking is? The judge looks at the likelihood of something
like that happening by accident. If it seems highly unlikely that
something like that will happen by accident, then the judge penalizes
the player because the judge has strong reason to believe that it was
intentional (even though there is a one-in-a-billion chance that the
player had not intended this). If he doesn't - then a player could
continuously try to cheat, and fall back to the 'I didn't mean to
cheat' defense to get away with a lesser penalty or no penalty at all.

The same applies in collusion - if the judge has strong reason to
believe that a player is colluding (i.e. the intention to collude is
there) then the judge should act upon it, regardless of whether or not
the player claims innocence. My point was that, at the end of the day,
the actual intention of the player may be relevant in terms of the
rules, but the only way a judge can determine intention is through the
evidence presented to the judge. If that gives the judge strong reason
to believe collusion is taking place, then the judge should act on it.
There is a small chance that the judge may be incorrect, and that the
player is truly innocent - but, as you said, if the judge tolerates
obvious cases of collusion, the _potential_ for cheating, if not the


actuality of it - is bad.

Ok, as promised, I'll try to explain my issue with:

"For tournaments, playing to win means playing to get a Game Win if it
is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible."

Which I'll refer to as the PTW clause.

The problem is that it leads to these two similar situations:

A) Going for the GW if reasonably possible, or, if not, a fall-back
position of going for VP's
B) Going for the VP if reasonably possible, or, if not, a fall-back
position of being able to do anything.

In other words, it places the same emphasis on going for a GW as it
places on going for VPs (at least by my understanding of it).

However, when a player is in the corner-case position of a player
going for VP's instead of a chance for a GW, the threshold of what is
considered to be reasonably possible (and thus requiring the player to
try for the GW) is different from the threshold of what is considered
to be reasonably possible (and thus requiring the player to go for the
VP) for a player who is placed in a position of being able to do
whatever the player wants instead of going for the chance of a VP.

I tried to use an example that relied on the chances of a single card
draw to highlight this issue - because, in both cases, it has an
equally measurable amount of risk. Even in a game, when a judge is
called over, the player can state his situation, and say, for example,
I need to draw a specific card to get the GW in the former, or VP in
the latter - I have three cards left in my library - and one of them
is it - giving me a 33% chance of getting what I need. The same judge
might be guided by his intuition that, in the former example, the 33%
chance is too risky and tell the player they are free to play either
way, while, in the latter example he might be guided by his intuition
to say that 33% chance isn't risky enough, so the player should go for
the VP.

While I agree that the judge *should* hold up the same standard in
both situations, given the way that the PTW clause is written at the
moment - it seems counter-intuitive. Going for the safety of a VP as
opposed to the chance of a GW seems like someone who is, in fact,
playing to win (because the certain VP would lead him to tie with
someone else and get to coin-flip against them - a 50% chance as
opposed to a 33% chance- or lead to a better tournament position).
However, in the latter case, it seems clear that if the player doesn't
take (that equal) chance at a VP, he is not actually playing to win -
as not going for the chance gains him nothing (except maybe relieving
a feeling of vengence against another player).

Hence, we are stuck in a situation that we either hold a single
counter-intuitive standard, or we have a double standard.

That's my problem in a nutshell, and that's why I'm exploring
alternatives to the way the PTW clause is written.

Hopefully, it's a little more clear?

LSJ

unread,
Feb 20, 2009, 6:35:30 AM2/20/09
to
The Name Forgotten wrote:
> The problem is that it leads to these two similar situations:
>
> A) Going for the GW if reasonably possible, or, if not, a fall-back
> position of going for VP's
> B) Going for the VP if reasonably possible, or, if not, a fall-back
> position of being able to do anything.
>
> In other words, it places the same emphasis on going for a GW as it
> places on going for VPs (at least by my understanding of it).

Not at all. The priority is on the GW, just as you wrote above.

Just as the criteria for going to the finals places a priority on number of GWs
and uses VP as a secondary criteria (and TPs as a third).

> However, when a player is in the corner-case position of a player
> going for VP's instead of a chance for a GW, the threshold of what is
> considered to be reasonably possible (and thus requiring the player to
> try for the GW) is different from the threshold of what is considered
> to be reasonably possible (and thus requiring the player to go for the
> VP) for a player who is placed in a position of being able to do
> whatever the player wants instead of going for the chance of a VP.

Nothing makes those thresholds different, other than (possibly) actual differing
circumstances in the game or (possibly) sub-optimal judging.

> I tried to use an example that relied on the chances of a single card
> draw to highlight this issue - because, in both cases, it has an
> equally measurable amount of risk. Even in a game, when a judge is
> called over, the player can state his situation, and say, for example,
> I need to draw a specific card to get the GW in the former, or VP in
> the latter - I have three cards left in my library - and one of them
> is it - giving me a 33% chance of getting what I need. The same judge
> might be guided by his intuition that, in the former example, the 33%
> chance is too risky and tell the player they are free to play either
> way, while, in the latter example he might be guided by his intuition
> to say that 33% chance isn't risky enough, so the player should go for
> the VP.

Sure, No double standard. That same judge could judge the former as a reasonable
chance while judging that latter as unreasonable. That is, he could err the
other way, too. Or he could judge both to be reasonable, or both to be unreasonable.

None of this has anything to do with the existence of the PTW rule; it only
involves a criticism of relying on judges. That is a valid criticism. However,
we do need judges -- that bit is unavoidable.

>
> While I agree that the judge *should* hold up the same standard in
> both situations, given the way that the PTW clause is written at the
> moment - it seems counter-intuitive. Going for the safety of a VP as
> opposed to the chance of a GW seems like someone who is, in fact,
> playing to win (because the certain VP would lead him to tie with
> someone else and get to coin-flip against them - a 50% chance as
> opposed to a 33% chance- or lead to a better tournament position).

A 33% shot at a VP vs. a 33% shot at a GW leads the player to have to go for the
GW. Having a 90% chance at a VP and a 5% chance at a GW leaves the player free
to go for the VP (assuming the judge judges the 5% chance at the GW to be not a
reasonable chance).

(Aside: of course, the numbers cannot be computed so precisely in practice.)

> However, in the latter case, it seems clear that if the player doesn't
> take (that equal) chance at a VP, he is not actually playing to win -

> as not going for the chance gains him nothing [snip distraction].

If the player choses a VP over an equally-likely GW, he or she is violating the
PTW rule, yes. No double standard.

> Hence, we are stuck in a situation that we either hold a single
> counter-intuitive standard, or we have a double standard.

How? The only problem you've shown is that we need judges. You haven't shown any
problem with the PTW rule.

> That's my problem in a nutshell, and that's why I'm exploring
> alternatives to the way the PTW clause is written.

The alternatives would still leave the need for judges.

The Name Forgotten

unread,
Feb 20, 2009, 8:53:17 AM2/20/09
to
On Feb 20, 1:35 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > The problem is that it leads to these two similar situations:
>
> > A) Going for the GW if reasonably possible, or, if not, a fall-back
> > position of going for VP's
> > B) Going for the VP if reasonably possible, or, if not, a fall-back
> > position of being able to do anything.
>
> > In other words, it places the same emphasis on going for a GW as it
> > places on going for VPs (at least by my understanding of it).
>
> Not at all. The priority is on the GW, just as you wrote above.
>
> Just as the criteria for going to the finals places a priority on number of GWs
> and uses VP as a secondary criteria (and TPs as a third).

That's not what I'm contending. The priority is to go for a GW -
agreed. By 'same emphasis' - I mean it uses the exact same standard
("reasonably possible") to measure with the exact same strength of
measurement, for when a person should go for a Game Win, as it does
for when a person should go for a VP. It implies that you are using
the same yard-stick with the exact same cut-off point.

To use an analogy:

I ask you to send me 3000 sheets of paper, under certain conditions. I
ask you to give priority to sending me 3000 reams of A4 paper, or if
you are unable to send me 3000 reams of A4 paper, then I want you to
send me 3000 reams of A3 paper, or if you are unable then I want you


to do as you please.

The standard of 3000 sheets of paper (or 'reasonably possible') is the
same, even though priority is given to A4 paper ("GW").

So it is with this - the same standard of what can be considered
'reasonably possible' is expected to be used to measure whether a
person should go for a GW, or go for VPs. However, applying the same
standard leads to counter-intuitive results.

> > However, when a player is in the corner-case position of a player
> > going for VP's instead of a chance for a GW, the threshold of what is
> > considered to be reasonably possible (and thus requiring the player to
> > try for the GW)  is different from the threshold of what is considered
> > to be reasonably possible (and thus requiring the player to go for the
> > VP) for a player who is placed in a position of being able to do
> > whatever the player wants instead of going for the chance of a VP.
>
> Nothing makes those thresholds different, other than (possibly) actual differing
> circumstances in the game or (possibly) sub-optimal judging.
>
> > I tried to use an example that relied on the chances of a single card
> > draw to highlight this issue - because, in both cases, it has an
> > equally measurable amount of risk. Even in a game, when a judge is
> > called over, the player can state his situation, and say, for example,
> > I need to draw a specific card to get the GW in the former, or VP in
> > the latter - I have three cards left in my library - and one of them
> > is it - giving me a 33% chance of getting what I need. The same judge
> > might be guided by his intuition that, in the former example, the 33%
> > chance is too risky and tell the player they are free to play either
> > way, while, in the latter example he might be guided by his intuition
> > to say that 33% chance isn't risky enough, so the player should go for
> > the VP.
>
> Sure, No double standard.

That is, by it's very definition, a double standard. He is using the
same 'risk-assessment' measuring tool - coming to the same result, yet
using that result to judge in favour one way in one situation, and
using it to judge in favour the other way in the next. Unless there is
something else that can be brought to bare on the situation in a
relevant manner (i.e. is backed by rules, or clarifications or
whatever), another thing to be considered to measure what is
'reasonable' that doesn't rely on that kind of risk-assessment, the
player has every right to say that he is being unfairly discriminated
against, because the player is.

Personally, I feel that their must be some consideration paid to the
fact that the person is risking the much better chance of getting VPs
to go after the GW, as opposed to the player who is risking nothing at
all (as he's in a losing position already) to go after a VP. However,
this consideration doesn't seem to be validated at all by rules,
clarifications, or anything else that I've found to date. This is
solely because of the way that clause is structured - in that a person
only needs to go for VPs once a person can't get a GW.

A person should always go for VPs, although a person should be given
the option to go for a GW even if it conflicts with getting VPs. That
would close the double standard quite neatly, IMO.

> That same judge could judge the former as a reasonable
> chance while judging that latter as unreasonable. That is, he could err the
> other way, too. Or he could judge both to be reasonable, or both to be unreasonable.
>
> None of this has anything to do with the existence of the PTW rule; it only
> involves a criticism of relying on judges. That is a valid criticism. However,
> we do need judges -- that bit is unavoidable.

Moot.

> > While I agree that the judge *should* hold up the same standard in
> > both situations, given the way that the PTW clause is written at the
> > moment - it seems counter-intuitive. Going for the safety of a VP as
> > opposed to the chance of a GW seems like someone who is, in fact,
> > playing to win (because the certain VP would lead him to tie with
> > someone else and get to coin-flip against them - a 50% chance as
> > opposed to a 33% chance- or lead to a better tournament position).
>
> A 33% shot at a VP vs. a 33% shot at a GW leads the player to have to go for the
> GW. Having a 90% chance at a VP and a 5% chance at a GW leaves the player free
> to go for the VP (assuming the judge judges the 5% chance at the GW to be not a
> reasonable chance).
>
> (Aside: of course, the numbers cannot be computed so precisely in practice.)

At least, not in almost all cases. Unless, like I said, you end up in
a "I need to draw the right card for this to work" scenario. It
*might* be more plausable to calculate odds in a draft scenario in
real-life practice (at least, so has been my experience).

I do grant a certain amount of leeway for judging consideration - for
example, you could give general probablities, like "Near-impossible",
"about 50-50" etc. But so long as the odds fall in the same range (in
real life practice), the same standard should be upheld.

> > However, in the latter case, it seems clear that if the player doesn't
> > take (that equal) chance at a VP, he is not actually playing to win -
> > as not going for the chance gains him nothing [snip distraction].
>
> If the player choses a VP over an equally-likely GW, he or she is violating the
> PTW rule, yes. No double standard.

This is where my problem is - the current PTW clause says nothing
about you having to pursue VP's *until* you can no longer pursue a GW.
There is nothing that I can find that leads me to believe that the
pursuit of VPs should enter into consideration until the player can no
longer go for GW. However, I feel that intuitively and in the spirit
of PTW, there should be.

> > Hence, we are stuck in a situation that we either hold a single
> > counter-intuitive standard, or we have a double standard.
>
> How? The only problem you've shown is that we need judges. You haven't shown any
> problem with the PTW rule.

Moot.

> > That's my problem in a nutshell, and that's why I'm exploring
> > alternatives to the way the PTW clause is written.
>
> The alternatives would still leave the need for judges.

Of course, but moot.

LSJ

unread,
Feb 20, 2009, 9:16:16 AM2/20/09
to
The Name Forgotten wrote:
> A person should always go for VPs, although a person should be given
> the option to go for a GW even if it conflicts with getting VPs. That
> would close the double standard quite neatly, IMO.

Going for the GW you don't yet have doesn't conflict with getting VPs.

How do you go for a GW without going for VPs (unless you've already got the GW
sewn up)?

The Name Forgotten

unread,
Feb 20, 2009, 10:28:29 AM2/20/09
to

I think John provided a good example in this thread:

http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/57fd9ed0ec587267?hl=en

There are some bizarre situations that make it better for you to not
pursue VPs in order to make sure you get a GW. Of course these
generally involve either you having VP's already, or you getting VPs
(although you could get more by taking a riskier action).

Basic situation:
A (with 2 VPs) bleeding B bleeding C.

A has the chance to oust himself and C with a Domain Challenge, thus
giving himself 2VPs & the GW and everyone else 1 VP. He can take a
riskier action to try oust B (thus gaining a VP), but it's highly
unlikely to work, at which point B will oust both C and A with
certainty, giving B 3VPs and a GW instead of A.

It shouldn't be an either/or situation - a person should be free to
choose to go for VPs or a GW in any situation. However, with such a
low-tolerance towards people who want to self-oust because they cannot
reasonably get a VP, you'd need something to stop an equally low
tolerance from affecting people's decisions when they need to choose
between VPs or GW, especially since those decisions will be based on
the spirit of playing to win.

The Name Forgotten

unread,
Feb 20, 2009, 10:34:06 AM2/20/09
to
On Feb 20, 5:28 pm, The Name Forgotten <tex...@telkomsa.net> wrote:
> On Feb 20, 4:16 pm, LSJ <vtes...@white-wolf.com> wrote:
>
> > The Name Forgotten wrote:
> > > A person should always go for VPs, although a person should be given
> > > the option to go for a GW even if it conflicts with getting VPs. That
> > > would close the double standard quite neatly, IMO.
>
> > Going for the GW you don't yet have doesn't conflict with getting VPs.
>
> > How do you go for a GW without going for VPs (unless you've already got the GW
> > sewn up)?
>
> I think John provided a good example in this thread:
>
> http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/57fd...

>
> There are some bizarre situations that make it better for you to not
> pursue VPs in order to make sure you get a GW. Of course these
> generally involve either you having VP's already, or you getting VPs
> (although you could get more by taking a riskier action).
>
> Basic situation:
> A (with 2 VPs) bleeding B bleeding C.
>
> A has the chance to oust himself and C with a Domain Challenge, thus
> giving himself 2VPs & the GW and everyone else 1 VP. He can take a
> riskier action to try oust B (thus gaining a VP), but it's highly
> unlikely to work, at which point B will oust both C and A with
> certainty, giving B 3VPs and a GW instead of A.
>
> It shouldn't be an either/or situation - a person should be free to
> choose to go for VPs or a GW in any situation. However, with such a
> low-tolerance towards people who want to self-oust because they cannot
> reasonably get a VP, you'd need something to stop an equally low
> tolerance from affecting people's decisions when they need to choose
> between VPs or GW, especially since those decisions will be based on
> the spirit of playing to win.

After thinking about this further maybe a better way of phrasing the
play to win section would be:

"For tournaments, playing to win means playing to get a Game Win if it

is reasonably possible, and/or playing to get as many Victory Points
as possible."

Instead of:

"For tournaments, playing to win means playing to get a Game Win if it
is reasonably possible, and when a Game Win is not reasonably
possible, then playing to get as many Victory Points as possible."

Can anyone see any problems with that?

LSJ

unread,
Feb 20, 2009, 10:35:54 AM2/20/09
to

OK. So you allow for risk analysis in determining whether or not to go for the 3
VP game win over the 2VP game win. So does the current PTW rule. So how does
your rule differ from the current PTW rule?

Your proposed rule:


A person should always go for VPs, although a person should be given
the option to go for a GW even if it conflicts with getting VPs. That
would close the double standard quite neatly, IMO.

sounds like a restatement of the current rule (except in cases of already having
the GW, which your rule seems to mandate the player not be allowed to roll
over after getting the game win).

LSJ

unread,
Feb 20, 2009, 10:40:12 AM2/20/09
to
The Name Forgotten wrote:
> After thinking about this further maybe a better way of phrasing the
> play to win section would be:
>
> "For tournaments, playing to win means playing to get a Game Win if it
> is reasonably possible, and/or playing to get as many Victory Points
> as possible."
>
> Instead of:
>
> "For tournaments, playing to win means playing to get a Game Win if it
> is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible."
>
> Can anyone see any problems with that?

It seems to change nothing except the words used. The effect is the same, except
that it occasionally (for extremely low values of occasionally) allows a
player with a reasonable chance at a GW to throw the game in favor of more
(though no more reasonable) VPs, which isn't necessarily an interesting change
to make.

Frederick Scott

unread,
Feb 20, 2009, 1:07:10 PM2/20/09
to

"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:fad3ba08-d0b4-4c58...@f24g2000vbf.googlegroups.com...

> On Feb 20, 12:02 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > "The Name Forgotten" <tex...@telkomsa.net> wrote in
> > messagenews:130898ea-7fe0-4c63...@p23g2000prp.googlegroups.com...
> >
> > > On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > >> > Sure, play-to-win covers some of these situations,
> > >> > but if it covered all of them, there'd be no need for a collusion rule
> >
> > >> Well, gee...that's what *I'VE* been saying, if you read my original response
> > >> to this situation.
> >
> > > Sorry, I'm not sure I understand - have you been saying that we should
> > > keep the PTW rule, but there is no need for the collusion rule?
> >
> > Yep. To tell the truth, until you quoted from the judge's manual, I had no
> > idea what the Collusion Clause in the Sportsmanship Rule was actually for.
> > It's written terribly, doesn't define what it means very well, and - as
> > written - could be applied to any two players who struck any deal in any
> > game in any tournament.
>
> No. It can't. It lists many exceptions. Namely:
>
> Players participating in standard table talk or in-game agreements
> should not be considered in violation of this rule as long as they
> meet the following criteria:

Where are you getting this? Looking at the TOURNAMENT RULES, it merely
says, "Cheating includes, but is not limited to, the following intentional
activities...Collusion to alter the results of a game".

Try to remember, the Judge's Guide is *NOT* part of the tournament rules.

> > Given the judge's manual, I now believe I understand
> > the spirit of that clause.
>
> Evidently you don't.

I certainly do. But whatever is written in the Judge's Guide doesn't
alter the fact: the rule *ITSELF* is terribly written. The fact that it's
intended meaning is clarified in the Judge's Guide doesn't make it all better.
The Judge's Guide is not an obligatory document. Judges should be able to
understand the meaning of a rule based on how the rule itself is written.
Frankly, I hate rules that can so easily misinterpreted. That's why I said:
this rule needs to be clarified (in the *TOURNAMENT RULES*) or eliminated.
Since I really don't see what function it serves, given PTW, I'd say it should
be eliminated.

> > >> So your original argument was: "We don't need PTW because PTW is covered
> > >> in other rules".
> >
> > > No. My original argument was we shouldn't have PTW because it leads to
> > > a double standard.
> >
> > I have to confess, I've never understood your double-standard argument.
>
> Therein lies the problem. If you cannot understand my issue with the
> PTW, then any further debate is pointless. I'd rather spend the time
> focusing on helping you understand my point there, before I argue on a
> seperate issue that relies on that point.

I think it should give you pause to consider whether your idea that the
clause you want to eliminate actually creates double standards. I mean,
to reduce your argument down to it's essence, this is what I've gotten
to date: What the Play-to-Win rule actually says is, "It should be
considered unsportsmanlike conduct to play a game towards goals NOT written
the rulebook when they're in conflict with the goals written in the rulebook."
In essence, it says, you can't play toward "X" kinds of goals where they're in
conflict with "Y" kinds of goals - "X" being out-of-game game goals, "Y"
being the goals written in the rulebook. Then, it gives a couple examples
(but not limited to...) of "X" goals: playing for VEKN ratings or playing for
tournament standings. Then, it has a clause which better defines what "Y"
goals mean: playing for a game win if obtainable or, failing that, playing
for the most victory points.

So, claiming that the last clause - which more concretely defines what are
the "Y" goals - creates double standards, essentially you wish to get rid of
it and replace it with more examples of possible "X" goals. What is wrong
with this picture? But this is all any reasonable person can get from your
reasoning so far.

> > My suspicion is that I'd either disagree
> > that both were PTW violations or else that both were not PTW violations or
> > alternatively that I'd disagree that a correct judge should invoke the PTW
> > rule in one case and not the other.
>
> That's my point exactly. Still, I'd be going against my intuitions
> about the spirit of PTW if I either enforced both violations or
> disagreed that they were both violations - given that the relevant
> risks were identical in terms of the rules. I'm guessing other judges
> might feel similarly inclined, and might in fact, rule differently in
> those situations (or at least want to).

Well, I don't know why you'd think that. You're asserting you think other
judges might be tempted to rule the situations differently when, in fact,
they're the same and therefore the rule creates a double standard. It's
not a very convincing argument. If anything, it might be an argument for
adding more definition and clarity to the rule - but certainly not taking
away the clause that defines what is playing towards the goals listed in
the rulebook.

> > > My argument at this point was to agree that it was an issue, and there
> > > is a better alternative to the Play-to-Win rule to stop this from
> > > happening (that wouldn't lead to a double standard.)
> >
> > OK. I guess I can't respond to that because a double standard can't be
> > demonstrated, IMO.
>
> It can be demonstrated through a thought experiment. I'll try clarify
> my point in a different way below.

...


> > > 4.8. Play to Win
> >
> > > One aspect of sportsmanlike conduct is that players must not play
> > > toward goals that conflict with the goal of the game as stated in the
> > > V:TES rulebook (e.g., attacking certain players on the basis of their
> > > V:EKN ratings or overall tournament standing, etc.).
> >
> > > *** or attacking certain players on the basis of out of game
> > > considerations, e.g. race, gender, animosity, etc. Also, a player
> > > will
> > > not play to benefit certain players on the basis of their V:EKN
> > > ratings, overall tournament standings, or other out-of-game
> > > considerations, eg. friendship, fame, lack of time etc.
> >
> > When you toss in "other out-of-game" considerations, you've made
> > all the rest superfluous, AFAICS. In order to be a rule, you have
> > to define what is illegal behavior. As I've pointed out to you elsewhere,
> > examples do not define. They merely clarify definitions. I suppose you
> > could add a line that reads something like, "Examples of out-of-game
> > considerations include (but are not limited to): V:EKN ratings, overall
> > tournament standings, race, gender, animosity, friendship, fame, or
> > lack of time." But in fact, this addition adds nothing to the actual
> > rule.
>
> Uhm... duh??? What does eg. mean to you?

Essentially, "for instance". But my version was more consise and made
better sense than your's. Your's said, "(for instance) race, gender
animosity, or other-of-game considerations..." adding another 'e.g.'
there but essentially means everything else that's not an in-game
consideration. Everything you mentioned was an out-of-game consideration
so...why not just say THAT?!? Basically, you're not suppose to play
for out-of-game considerations when they're in conflict with rulebook
game goals - and here's a lot of possible examples. That's much more
straightforward than what you wrote.

> The "other out-of-game"
> considerations is exactly the change that adds something to the actual
> rule. I think I've lost your point here. Sure VEKN ratings and overall
> tournament are further examples of out of game considerations.

I have lost the point, you're right. The rule already said that. All
you did was to add was some more examples of things that might be
out-of-game considerations.

> > > *** (Remove section about VP's & GW's)
> >
> > Removing this, on the other, is a positive detriment to the rule. The
> > section about VPs and GWs is very important. It defines what, exactly,
> > is meant by "playing in goals which are in conflict with the goal of the
> > game as stated in the V:TES rulebook".
>
> No, the rules in the V:TES rulebook define what the goals of the game
> are.

Not in that precise manner, they don't. They define winning the game. They
define victory points. But they're ambiguous about how these two goals
relate to one another. What's more important? Maximizing my chances for
a sweep? Or just making sure I get the most victory points. Also, since
the tournament structure has its own definition of a what a "game win" is
and it has the concept of a game timing out and separately defines how many
victory points are awarded when a game times out. Because the rulebook
rules don't recognize any of these points, the PTW rule needs to make some
allowance for them or players might be in jeopardy of having to play towards
truly counterproductive goals (as "playing for tournament standings"). For
instance, if I have two victory points, I don't need to do anything else
except to make sure no single player gains a second one to get a game win.
The PTW rule should recognize this point. The rulebook rules by themselves
do not. On the other hand, if I have only one victory point and no one else
has any and the game is in danger of timing out, I would fail to get a game
win by tournament rules but not by rulebook rules. It should not be OK for
me to play for a timeout in this situation.

This is an important clause. It creates no double standard that didn't
already exist. I'm completely confused why it bothers you.

> > Notice that playing towards
> > any of the goals cited about isn't actually illegal under 4.8. It's
> > only when such goals are *IN CONFLICT* with the rulebook goals that

> > a violation has occurred. But what are the rulebook goals with which


> > the other goals must not conflict?
>
> Well - this is the rulebook goal with which the other goals must no
> conflict (it's one of the first things in the rulebook):
>
> Object of the Game
> Your goal is to accumulate the most victory points by destroying the
> influence held by rival Methuselahs. You accomplish this goal by using
> your influence to gain control of younger vampires and using those
> vampires to take actions to reduce the influence of rival Methuselahs.
> Influence is represented by pool counters (see Equipment Needed, sec.
> 1.2), the main currency of the game. When a Methuselah runs out of
> pool counters, she* is bereft of influence and ousted from the game.
> As Methuselahs are ousted, players earn victory points; the winner is
> the player with the most victory points at the end of the game (see
> Victory Points, sec. 9.1).

Right. You'll notice there are two separate goals listed. Because
there are two, it's ambiguous from the clause above, which is actually
the more important goal. So removing the clause in the PTW rule which
sorts this out would return this situation to its ambiguous state.

Furthermore, the game win rule in VEKN tournaments was create for some
very good reasons. But they don't exist in the rulebook rules. So,
left to itself, it could be considered to be an "out-of-game" consideration
under PTW to play for a tournament game win without this clause. That
would be bad. The clause exists, in part, to give recognition to the
point that players are supposed to be playing for a tournament game win
and that goal should considered an in-game goal.

> > It's defined in that section! Why
> > are we getting rid of this crucial piece?
>
> Because the Game Win section seems to create a double standard. (I'll
> try explain this at the end).
>
> Note that going for a Game Win is not an out-of-game consideration,

Ah, but without this clause, it is!

> > so, if anything, the Play-to-Win rule should include an exception
> > clause that allows for in-game considerations, such as it's already
> > present deal-making exception, for going for Game Wins as opposed to
> > VPs - which may be a problem in some corner-case examples.

I'm not sure what you're saying here. The clause you want to eliminate
is the clause that creates the exception that allows players to go for
a game win - and that's a _good_ thing.

> > Nothing you added fixes the problem you just caused by removing the
> > section about VPs and GWs. What you added talks about the potentially
> > conflicting goals, not the ones which must not be conflicted.
>
> Sorry, I assumed that people were already familiar with the Object of
> the Game, as stated in the rulebook.

I'm sorry, that response makes no sense. You added more examples to
what can be considered out-of-game goals. That's all you did AFAICS.
My familiarity with the rulebook OotG has nothing to do with this.

Huh?!? I don't think that's a valid question. I don't see it being
very hard to determine what is a deliberate, intentional mark and what
appears to be an accident. I guess if it could reasonably be an accident,
like a dog-eared card sleeve, you have to assume it was an accident. If
it couldn't be an accident, you assume it was deliberate. Players won't
usually be very successful at cheating by marking their cards in ways
that could be mistaken as accidental because such markings are very easy
to detect. If that's the cheater's game, then you may have to assume it
was an accident the first time you see it but keep an eye on him. If
he continually does this, you nab him with a serious of escalating warnings
and decreasing tolerance for "accidental" markings. One way or another,
you'll stop the cheating.

> The judge looks at the likelihood of something
> like that happening by accident. If it seems highly unlikely that
> something like that will happen by accident, then the judge penalizes
> the player because the judge has strong reason to believe that it was
> intentional (even though there is a one-in-a-billion chance that the
> player had not intended this). If he doesn't - then a player could
> continuously try to cheat, and fall back to the 'I didn't mean to
> cheat' defense to get away with a lesser penalty or no penalty at all.

See above. Common sense usually provides an answer.

> The same applies in collusion -

Nope. To collude, both players have to be involved in attempting to
collude. It's really very hard to prove that with the kind of
collusion you're talking about. But that's OK. It's relatively easy to
prove that Player B was not playing to win (in the situation you
described) - so just enforce the PTW rule and you'll have effectively
stopped the collusion...if it was collusion.

> Ok, as promised, I'll try to explain my issue with:
>
> "For tournaments, playing to win means playing to get a Game Win if it
> is reasonably possible, and when a Game Win is not reasonably
> possible, then playing to get as many Victory Points as possible."
>
> Which I'll refer to as the PTW clause.
>
> The problem is that it leads to these two similar situations:
>
> A) Going for the GW if reasonably possible, or, if not, a fall-back
> position of going for VP's
> B) Going for the VP if reasonably possible, or, if not, a fall-back
> position of being able to do anything.
>
> In other words, it places the same emphasis on going for a GW as it
> places on going for VPs (at least by my understanding of it).
>
> However, when a player is in the corner-case position of a player
> going for VP's instead of a chance for a GW, the threshold of what is
> considered to be reasonably possible (and thus requiring the player to
> try for the GW) is different from the threshold of what is considered
> to be reasonably possible (and thus requiring the player to go for the
> VP) for a player who is placed in a position of being able to do
> whatever the player wants instead of going for the chance of a VP.

Perhaps. But how is this different from players who have to make a
decision to go for a single sure VP vs. a sweep (3 more VPs)? Eliminating
the clause you want to eliminate doesn't fix this problem. If anything,
it excaberates the problem. And then you have the problem that playing
for a tournament-defined game win could be ruled an "out-of-game goal",
which would be bad.

...


> While I agree that the judge *should* hold up the same standard in
> both situations, given the way that the PTW clause is written at the
> moment - it seems counter-intuitive.

I don't know what to say. If other judges understand the meaning of
the rule, then they, too, should uphold the same standard in both
situations (if they were truly the same situations). It doesn't
make sense to change a rule out of fear that judges will rule it
wrongly. If that's your fear, then you ask that the same rule be
maintained but rewritten in a way that's more clear. Your proposal
actually changes the rule, so wouldn't be the correct resposne to
problem you're citing.

> Going for the safety of a VP as
> opposed to the chance of a GW seems like someone who is, in fact,
> playing to win (because the certain VP would lead him to tie with
> someone else and get to coin-flip against them - a 50% chance as
> opposed to a 33% chance- or lead to a better tournament position).

The safe-VP-vs.-going-for-more judgement is always going to be in the
game. This is a judgement for a player to make and the game win rule
should never be invoked to sanction a player when he made a judgement
that the judge doesn't agree with. Your proposed change does not alter
this point, AFAICS.

> That's my problem in a nutshell, and that's why I'm exploring
> alternatives to the way the PTW clause is written.
>
> Hopefully, it's a little more clear?

I guess that's more clear, if I understood your explanation. But I think
you may misunderstand the application of the PTW rule if you think it's
causing a problem in this sense. It shouldn't be.

Fred


Frederick Scott

unread,
Feb 20, 2009, 1:17:32 PM2/20/09
to
"LSJ" <vte...@white-wolf.com> wrote in message
news:4Xynl.17326$yr3....@nlpi068.nbdc.sbc.com...

I think the essence of what he's getting at is that one sometimes has
to decide between a surer single victory point vs. forgoing the surer
victory point to increase your chances of eventually getting a
game win. It's not that one can go for a game win without going for
victory points but that going for a game win may cost you the certainty
of getting a single victory point.

(I don't think his proposed change fixes this, but this answers the
question you posed above.)

In American football, the analogous situation would be fourth and
goal at the one yardline, your team is down three points, and there
are two minutes left in overtime (NFL rules - you'll tie the game
if time expires when the score is tied). You're asking the equivalent of,
"How can your team go for a win without getting points?" But "getting
points" isn't the issue. It's what chance do you have of getting how
many points and how does this affect your long-term chance of winning
the game or tying the game. It's not clear which goal is more
important within the rules of the game itself.

Fred


The Name Forgotten

unread,
Feb 20, 2009, 2:31:29 PM2/20/09
to
On Feb 20, 8:07 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:fad3ba08-d0b4-4c58...@f24g2000vbf.googlegroups.com...

>
>
>
>
>
> > On Feb 20, 12:02 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > > "The Name Forgotten" <tex...@telkomsa.net> wrote in
> > > messagenews:130898ea-7fe0-4c63...@p23g2000prp.googlegroups.com...
>
> > > > On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > > >> > Sure, play-to-win covers some of these situations,
> > > >> > but if it covered all of them, there'd be no need for a collusion rule
>
> > > >> Well, gee...that's what *I'VE* been saying, if you read my original response
> > > >> to this situation.
>
> > > > Sorry, I'm not sure I understand - have you been saying that we should
> > > > keep the PTW rule, but there is no need for the collusion rule?
>
> > > Yep. To tell the truth, until you quoted from the judge's manual, I had no
> > > idea what the Collusion Clause in the Sportsmanship Rule was actually for.
> > > It's written terribly, doesn't define what it means very well, and - as
> > > written - could be applied to any two players who struck any deal in any
> > > game in any tournament.
>
> > No. It can't. It lists many exceptions. Namely:
>
> > Players participating in standard table talk or in-game agreements
> > should not be considered in violation of this rule as long as they
> > meet the following criteria:
>
> Where are you getting this?  Looking at the TOURNAMENT RULES, it merely
> says, "Cheating includes, but is not limited to, the following intentional
> activities...Collusion to alter the results of a game".
>
> Try to remember, the Judge's Guide is *NOT* part of the tournament rules.

From the introduction of the V:EKN Tournament Rules:

"Players who violate sections of the V:EKN Tournament Rules will be
subject to the appropriate provisions of the V:EKN Judges' Guide."

The Judge's Guide is very much part of the tournament rules.

> > > Given the judge's manual, I now believe I understand
> > > the spirit of that clause.
>
> > Evidently you don't.
>
> I certainly do.  But whatever is written in the Judge's Guide doesn't
> alter the fact: the rule *ITSELF* is terribly written.  The fact that it's
> intended meaning is clarified in the Judge's Guide doesn't make it all better.
> The Judge's Guide is not an obligatory document.  Judges should be able to
> understand the meaning of a rule based on how the rule itself is written.
> Frankly, I hate rules that can so easily misinterpreted.  That's why I said:
> this rule needs to be clarified (in the *TOURNAMENT RULES*) or eliminated.
> Since I really don't see what function it serves, given PTW, I'd say it should
> be eliminated.

Well, where it's stated isn't important - the tournament rules don't
repeat the rulebook either - but it does mention it must be followed -
just as it does with the judge's guide. If a Judge is not familiar
with the Judge's Guide, then the judge certainly shouldn't be judging
(at least not on their own). Try say that 10 times fast!

No, the "Y" goals are the ones that are already in the rulebook. If
you're a sucker for repetition, you could recopy the whole section
there if you'd like. The "X" goals just specify conditions that would
be different for a tournament situation, including additional
considerations for tournament play that would likely not appear, or be
excused in casual play.

> > > My suspicion is that I'd either disagree
> > > that both were PTW violations or else that both were not PTW violations or
> > > alternatively that I'd disagree that a correct judge should invoke the PTW
> > > rule in one case and not the other.
>
> > That's my point exactly.  Still, I'd be going against my intuitions
> > about the spirit of PTW if I either enforced both violations or
> > disagreed that they were both violations - given that the relevant
> > risks were identical in terms of the rules. I'm guessing other judges
> > might feel similarly inclined, and might in fact, rule differently in
> > those situations (or at least want to).
>
> Well, I don't know why you'd think that.  You're asserting you think other
> judges might be tempted to rule the situations differently when, in fact,
> they're the same and therefore the rule creates a double standard.  It's
> not a very convincing argument.  If anything, it might be an argument for
> adding more definition and clarity to the rule - but certainly not taking
> away the clause that defines what is playing towards the goals listed in
> the rulebook.
>

Which, I believe is the point I was making. I was saying, let's
imagine taking away the rule (because it appears to be problematic) -
then let's look if doing so would create any legitimate problems (not
covered by other rules) - if not, let's just drop the rule, if so,
then let's see if there is a better alternative.

You took this to mean that I'm being hardline about it and reject any
other possibility than dropping the rule.

I've said repeatedly, I think we can do better than this rule (or
clarification or whatever), but I admit that maybe we can't. It should
be explored because the rule appears to be problematic in some corner-
case examples.

True, my version could be more refined. It was written quickly and
indelicately. It was written like that because I wanted to keep as
much of the original ruling intact without inserting bit's and pieces
in several different areas. I think the relevant point was made
clearly enough to be understood.

> > The "other out-of-game"
> > considerations is exactly the change that adds something to the actual
> > rule. I think I've lost your point here. Sure VEKN ratings and overall
> > tournament are further examples of out of game considerations.
>
> I have lost the point, you're right.  The rule already said that.  All
> you did was to add was some more examples of things that might be
> out-of-game considerations.

Yes, to close the hole that 'a player helping his friend win' creates
(by adding in helping as well as attacking). Sure, I added in a few
unnecessary examples if you really want to complain about it, but
talking about attacking players on the basis of V:EKN ratings etc.
could already be considered unnecessary examples. I thought the
examples might be useful, you think they're not. Ok, I get that. We'll
just have to agree to disagree - it's not the main issue here.

> > > > *** (Remove section about VP's & GW's)
>
> > > Removing this, on the other, is a positive detriment to the rule. The
> > > section about VPs and GWs is very important. It defines what, exactly,
> > > is meant by "playing in goals which are in conflict with the goal of the
> > > game as stated in the V:TES rulebook".
>
> > No, the rules in the V:TES rulebook define what the goals of the game
> > are.
>
> Not in that precise manner, they don't.  They define winning the game.  They
> define victory points.  But they're ambiguous about how these two goals
> relate to one another.  

Huh? How are these two below sentences ambiguous about how victory
points and winning relate at all?

"Your goal is to accumulate the most victory points by destroying the
influence held by rival Methuselahs."

And

"the winner is the player with the most victory points at the end of
the game"

It seems to inextricably link winning with earning the most VPs.

> What's more important?  Maximizing my chances for
> a sweep?  Or just making sure I get the most victory points.  

These seem to be the same consideration - unless you're talking about
an odds gamble? After all, if you maximise your chances of getting a
table - you are maximising your chances at getting the most VPs.

If you are talking about an odds gamble - either option is fine - you
are playing towards getting the most VPs - maybe you might think the
risk to get 5VPs is far too high, and you're better off playing to get
only 3 VPs which you have a much higher certainty for. It doesn't
matter, so long as your goal is to get the MOST victory points at the
end of the game.

I grant that, in a tournament, you might want to have something that
still makes players continue playing at a table once somebody has
gotten more VPs than you could possibly get, but let's leave that
debate for now. I acknowledge that the current play-to-win clause
fulfills that function - to some extent - but I don't think it's the
best solution.

> Also, since
> the tournament structure has its own definition of a what a "game win" is
> and it has the concept of a game timing out and separately defines how many
> victory points are awarded when a game times out.  Because the rulebook
> rules don't recognize any of these points, the PTW rule needs to make some
> allowance for them or players might be in jeopardy of having to play towards
> truly counterproductive goals (as "playing for tournament standings").  For
> instance, if I have two victory points, I don't need to do anything else
> except to make sure no single player gains a second one to get a game win.
> The PTW rule should recognize this point.  The rulebook rules by themselves
> do not.  On the other hand, if I have only one victory point and no one else
> has any and the game is in danger of timing out, I would fail to get a game
> win by tournament rules but not by rulebook rules.  It should not be OK for
> me to play for a timeout in this situation.

Agreed, and all very valid points. The concept of 'game win' in
tournament rules certainly warrant some sort of extra definition in
the play-to-win rule. That's why I agree that solely abandoning it
isn't the best option (which I've done after looking at the situation
carefully, and getting good feedback), but I still feel it could be
written better.

> This is an important clause.  It creates no double standard that didn't
> already exist.  I'm completely confused why it bothers you.

It does because it creates 2 goals - getting GWs and getting VPs -
whereas the rulebook only has 1 goal - get more VPs than everyone
else. By giving one goal priority over the other (to the point that
the other goal is excluded until the first goal is unattainable),
along with the decision of how reasonable it is to expect people to go
after the two different goals creates a double standard.

Until a player can no longer get a GW - the pursuit of VPs is not even
a goal of that player (except that it generally helps him get the GW,
although it's not always so) - at least, according to the tournament
rules as they currently read.

The Name Forgotten

unread,
Feb 20, 2009, 3:13:14 PM2/20/09
to
>
> > Well - this is the rulebook goal with which the other goals must no
> > conflict (it's one of the first things in the rulebook):
>
> > Object of the Game
> > Your goal is to accumulate the most victory points by destroying the
> > influence held by rival Methuselahs. You accomplish this goal by using
> > your influence to gain control of younger vampires and using those
> > vampires to take actions to reduce the influence of rival Methuselahs.
> > Influence is represented by pool counters (see Equipment Needed, sec.
> > 1.2), the main currency of the game. When a Methuselah runs out of
> > pool counters, she* is bereft of influence and ousted from the game.
> > As Methuselahs are ousted, players earn victory points; the winner is
> > the player with the most victory points at the end of the game (see
> > Victory Points, sec. 9.1).
>
> Right.  You'll notice there are two separate goals listed.  Because
> there are two, it's ambiguous from the clause above, which is actually
> the more important goal.  

There is? Please explain? I only see one - get the most victory
points.

> So removing the clause in the PTW rule which


> sorts this out would return this situation to its ambiguous state.

If the rulebook is ambiguous then the rulebook should change.

> Furthermore, the game win rule in VEKN tournaments was create for some
> very good reasons.  But they don't exist in the rulebook rules.  So,
> left to itself, it could be considered to be an "out-of-game" consideration
> under PTW to play for a tournament game win without this clause.  That
> would be bad.  The clause exists, in part, to give recognition to the
> point that players are supposed to be playing for a tournament game win
> and that goal should considered an in-game goal.

Yes, I've noticed this to, which is why I've said the rule needs to be
amended, and not solely discarded.

> > > It's defined in that section! Why
> > > are we getting rid of this crucial piece?
>
> > Because the Game Win section seems to create a double standard. (I'll
> > try explain this at the end).
>
> > Note that going for a Game Win is not an out-of-game consideration,
>
> Ah, but without this clause, it is!

No - it's still an in-game consideration. Although it might not be
spoken of in the rulebook - it is spoken of in the tournament rules.
It relates directly to the specific game. Out-of-game considerations
are not the same thing as out-of-rulebook considerations. You don't
have to go to far to find other in-game considerations that are not
dealt with in the rulebook, such as deal-making, that are dealt with
in the tournament rules and considered perfectly legal.

> > > so, if anything, the Play-to-Win rule should include an exception
> > > clause that allows for in-game considerations, such as it's already
> > > present deal-making exception, for going for Game Wins as opposed to
> > > VPs - which may be a problem in some corner-case examples.
>
> I'm not sure what you're saying here.  The clause you want to eliminate
> is the clause that creates the exception that allows players to go for
> a game win - and that's a _good_ thing.

Yes, but the rule demands that you go for the game win as a priority
and in ignorance of victory points. That is a _bad_ thing.

> > > Nothing you added fixes the problem you just caused by removing the
> > > section about VPs and GWs. What you added talks about the potentially
> > > conflicting goals, not the ones which must not be conflicted.
>
> > Sorry, I assumed that people were already familiar with the Object of
> > the Game, as stated in the rulebook.
>
> I'm sorry, that response makes no sense.  You added more examples to
> what can be considered out-of-game goals.  That's all you did AFAICS.

Not only examples - but I also included that you cannot play solely to
_benefit_ another player, just as you cannot play solely for the
detriment of another player (for out of game reasons of course).

> My familiarity with the rulebook OotG has nothing to do with this.

I wasn't questioning _your_ familiarity. Your point was that there was
nothing else relevant that provided the goal for the game. My point is
that I assumed that players must be already familiar with the Object
of the Game, as stated in the rulebook - if it wasn't in the rulebook,
then no-one would know what the point of the game is until they played
in a tournament - in fact, there would be no point, no way to win, in
a casual game if it wasn't already in the rulebook. Hence it's
something that is obviously familiar to all players.

That's exactly my point with the colluding case. You'd spot something
strange - you'd say "hey, that looks an awful lot like colluding",
and

> If
> he continually does this, you nab him with a serious of escalating warnings
> and decreasing tolerance for "accidental" markings.  

Or acts that appear to be colluding.

> One way or another,
> you'll stop the cheating.

Agreed.

Look, I admit that marked cards are far easier to spot and far easier
to sanction than colluding - but that's because of the nature of
colluding - it's done secretively and revolves a lot around players
intentions. As a judge, you're asked to spot collusion through the
ways in which players act and play - it's usually your only way of
determining collusion. You'd have to be incredibly careful about how
you do it, and the decisions you make, but if the appearance of
colluding is strong enough - whether or not it's only coming from a
single player or a group - you need to act upon it.

> > The judge looks at the likelihood of something
> > like that happening by accident. If it seems highly unlikely that
> > something like that will happen by accident, then the judge penalizes
> > the player because the judge has strong reason to believe that it was
> > intentional (even though there is a one-in-a-billion chance that the
> > player had not intended this). If he doesn't - then a player could
> > continuously try to cheat, and fall back to the 'I didn't mean to
> > cheat' defense to get away with a lesser penalty or no penalty at all.
>
> See above.  Common sense usually provides an answer.

That's what I was trying to say - your common sense should guide the
issue - if you give me an example where a player looks like he's
colluding, but really he's not, of course I'm going to watch that
player. Yes, some judges might do it in secret, but I prefer telling
players my concern - it often leads far more quickly to better results
- because they can sometimes provide with a simple explanation that
makes sense. Somehow some people seem to think that doing this makes
me a bad judge. Well, tough shit. There are plenty of people that are
more than satisfied with my level of judging, and judges are not there
to please everyone (especially since everyone includes cheaters).

> > The same applies in collusion -
>
> Nope.  To collude, both players have to be involved in attempting to
> collude.  It's really very hard to prove that with the kind of
> collusion you're talking about.  But that's OK.  It's relatively easy to
> prove that Player B was not playing to win (in the situation you
> described) - so just enforce the PTW rule and you'll have effectively
> stopped the collusion...if it was collusion.

True. But even with the current PTW rule - if the person breaks it AND
you strongly suspect he was colluding - the person should be called up
on both charges. Agreed?

Agreed, which is why I said something would need to take it's place.

> > While I agree that the judge *should* hold up the same standard in
> > both situations, given the way that the PTW clause is written at the
> > moment - it seems counter-intuitive.
>
> I don't know what to say.  If other judges understand the meaning of
> the rule, then they, too, should uphold the same standard in both
> situations (if they were truly the same situations).  It doesn't
> make sense to change a rule out of fear that judges will rule it
> wrongly.  

No, my concern is that either you follow your intuition and the spirit
of play to win and rule differently in the two situations or you
discard your intuitions and uphold the same standard - that seems to
be against the spirit of PTW.

> If that's your fear, then you ask that the same rule be
> maintained but rewritten in a way that's more clear.  Your proposal
> actually changes the rule, so wouldn't be the correct resposne to
> problem you're citing.

Moot.

> > Going for the safety of a VP as
> > opposed to the chance of a GW seems like someone who is, in fact,
> > playing to win (because the certain VP would lead him to tie with
> > someone else and get to coin-flip against them - a 50% chance as
> > opposed to a 33% chance- or lead to a better tournament position).
>
> The safe-VP-vs.-going-for-more judgement is always going to be in the
> game.  This is a judgement for a player to make and the game win rule
> should never be invoked to sanction a player when he made a judgement
> that the judge doesn't agree with.  Your proposed change does not alter
> this point, AFAICS.

Yeah, that is my issue - I've put up several more proposed changes up
- but I'm not sure how solid they are either. Hopefully I'll arrive to
something that's more sensible.

> > That's my problem in a nutshell, and that's why I'm exploring
> > alternatives to the way the PTW clause is written.
>
> > Hopefully, it's a little more clear?
>
> I guess that's more clear, if I understood your explanation.  But I think
> you may misunderstand the application of the PTW rule if you think it's
> causing a problem in this sense.  It shouldn't be.

I might be - in fact I hope it is just me and my understanding of the
rule. If it is, at least this debate will clear the matter up for
others as well.

Thanks, Fred. I know most people would've given up debating rationally
by now. I'm really glad you stuck it through and are helping me out
with this issue (despite it getting frustrating at times).

The Name Forgotten

unread,
Feb 20, 2009, 3:32:54 PM2/20/09
to
On Feb 20, 5:35 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> > On Feb 20, 4:16 pm, LSJ <vtes...@white-wolf.com> wrote:
> >> The Name Forgotten wrote:
> >>> A person should always go for VPs, although a person should be given
> >>> the option to go for a GW even if it conflicts with getting VPs. That
> >>> would close the double standard quite neatly, IMO.
> >> Going for the GW you don't yet have doesn't conflict with getting VPs.
>
> >> How do you go for a GW without going for VPs (unless you've already got the GW
> >> sewn up)?
>
> > I think John provided a good example in this thread:
>
> >http://groups.google.com/group/rec.games.trading-cards.jyhad/msg/57fd...

>
> > There are some bizarre situations that make it better for you to not
> > pursue VPs in order to make sure you get a GW. Of course these
> > generally involve either you having VP's already, or you getting VPs
> > (although you could get more by taking a riskier action).
>
> > Basic situation:
> > A (with 2 VPs) bleeding B bleeding C.
>
> > A has the chance to oust himself and C with a Domain Challenge, thus
> > giving himself 2VPs & the GW and everyone else 1 VP. He can take a
> > riskier action to try oust B (thus gaining a VP), but it's highly
> > unlikely to work, at which point B will oust both C and A with
> > certainty, giving B 3VPs and a GW instead of A.
>
> > It shouldn't be an either/or situation - a person should be free to
> > choose to go for VPs or a GW in any situation. However, with such a
> > low-tolerance towards people who want to self-oust because they cannot
> > reasonably get a VP, you'd need something to stop an equally low
> > tolerance from affecting people's decisions when they need to choose
> > between VPs or GW, especially since those decisions will be based on
> > the spirit of playing to win.
>
> OK. So you allow for risk analysis in determining whether or not to go for the 3
> VP game win over the 2VP game win.

Just in case you misread my example - I allow for the risk analysis in
going for a 2VP game win over
(A) the chance at 3 or more VPs (and a GW) or
(B) the much greater chance of 2VP and no GW.

> So does the current PTW rule.

Agreed.

> So how does
> your rule differ from the current PTW rule?

It differs in the reverse case - when you allow for risk analysis in
going for 2VP over
(A) the chance at a GW or
(B) the much greater chance of only 1 VP.

> Your proposed rule:
> A person should always go for VPs, although a person should be given
> the option to go for a GW even if it conflicts with getting VPs. That
> would close the double standard quite neatly, IMO.
>
> sounds like a restatement of the current rule (except in cases of already having
>   the GW, which your rule seems to mandate the player not be allowed to roll

> over after getting the game win.

Well, if thats the worst part, then I'd be fine with that. However, a
player does not get a Game Win until the game is finished - according
to the Tournament Rules:

"A player receives a Game Win **** at the end of a game **** in which
they have received at least two (2) Victory Points and have more
Victory Points than any other player in the game. (No game win is
awarded in the case of ties.)"

So, once a player is secure with 3 VPs, the player still doesn't have
a Game Win until the end of the game. The player should play out the
rest of the game ensuring that the player, at least, doesn't lose any
VPs (which is possible with Gambit Accepted). That is to say - once
the player has reached the 3VP stage, it's safe to assume that the
player is playing to win - as the player need only play to the
conclusion of the game, or to when the player is ousted.

The Name Forgotten

unread,
Feb 20, 2009, 3:39:39 PM2/20/09
to
On Feb 20, 8:17 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "LSJ" <vtes...@white-wolf.com> wrote in message

>
> news:4Xynl.17326$yr3....@nlpi068.nbdc.sbc.com...
>
> > The Name Forgotten wrote:
> >> A person should always go for VPs, although a person should be given
> >> the option to go for a GW even if it conflicts with getting VPs. That
> >> would close the double standard quite neatly, IMO.
>
> > Going for the GW you don't yet have doesn't conflict with getting VPs.
>
> > How do you go for a GW without going for VPs (unless you've already got
> > the GW sewn up)?
>
> I think the essence of what he's getting at is that one sometimes has
> to decide between a surer single victory point vs. forgoing the surer
> victory point to increase your chances of eventually getting a
> game win.  It's not that one can go for a game win without going for
> victory points but that going for a game win may cost you the certainty
> of getting a single victory point.

Ah, not necessarily. As the example I gave to LSJ - if you are on 2VPs
already, and you can force everyone else to get 1VP in a sweeping
motion by ousting yourself and your predator with a single vote, then
you are welcome to do so. Very corner-case, I know, but possible. It
would ensure that you get a GW without you getting any more VPs, when
before you weren't secure in getting the GW.

>
> (I don't think his proposed change fixes this, but this answers the
> question you posed above.)

Please explain.

> In American football, the analogous situation would be fourth and
> goal at the one yardline, your team is down three points, and there
> are two minutes left in overtime (NFL rules - you'll tie the game
> if time expires when the score is tied).  You're asking the equivalent of,
> "How can your team go for a win without getting points?"  But "getting
> points" isn't the issue.  It's what chance do you have of getting how
> many points and how does this affect your long-term chance of winning
> the game or tying the game.  It's not clear which goal is more
> important within the rules of the game itself.

Uh... sorry, I'm not an American or familiar with the rules of
American football, so I can't really follow this point.

LSJ

unread,
Feb 20, 2009, 4:08:31 PM2/20/09
to
The Name Forgotten wrote:
>> Right. You'll notice there are two separate goals listed. Because
>> there are two, it's ambiguous from the clause above, which is actually
>> the more important goal.
>
> There is? Please explain? I only see one - get the most victory
> points.

Correct. And it's listed twice.

The ambiguity stems from reading the two instances differently.
Typically by reading the first "most" as "get as many VPs as you can" and the
second "most" as "get more VPs than anyone else".

But if you read them both as "end with more VPs than anyone else", there's no
discrepancy.

The goal of the game is to have more VPs than anyone else at the end of the game.

The tournament rules reflect that. And they add an additional goal for player
who cannot reach that main goal: get as many VPs as you can.

Frederick Scott

unread,
Feb 20, 2009, 5:16:13 PM2/20/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:d46ef596-5535-446a...@m40g2000yqh.googlegroups.com...

Two responses:

1) I'll admit, it's been a while since I've studied the explanations of cautions,
warnings, removal, and so forth. So I'll confess that I did not know that was true.
I will further venture that I think _that_ is also a bad idea for exactly the
reason I just gave: a judge's guide should just be a guide to judging. It shouldn't
give rules. If you've written your documents in such a way that your judges guide
is required to understand the rules of tournament play, then it isn't a judge's
guide. It's part of the rulebook. Call it that.

2) Even so, this statement does not claim to contain tournament rules outside of
the application of penalties. It just says the judge's guide explains what happens
*WHEN* rules are violated. So my compaint about the lack of clarity of this clause
stands.

> > Frankly, I hate rules that can so easily misinterpreted. That's why I said:
> > this rule needs to be clarified (in the *TOURNAMENT RULES*) or eliminated.
> > Since I really don't see what function it serves, given PTW, I'd say it should
> > be eliminated.
>
> Well, where it's stated isn't important

It sure is. Rules should be understood from what you can see in the rulebook.
If you can't understand them at all, then they're poorly written. This clause
cannot be understood at all from what's written in 5.1.

> > I think it should give you pause to consider whether your idea that the
> > clause you want to eliminate actually creates double standards. I mean,
> > to reduce your argument down to it's essence, this is what I've gotten
> > to date: What the Play-to-Win rule actually says is, "It should be
> > considered unsportsmanlike conduct to play a game towards goals NOT written
> > the rulebook when they're in conflict with the goals written in the rulebook."
> > In essence, it says, you can't play toward "X" kinds of goals where they're in
> > conflict with "Y" kinds of goals - "X" being out-of-game game goals, "Y"
> > being the goals written in the rulebook. Then, it gives a couple examples
> > (but not limited to...) of "X" goals: playing for VEKN ratings or playing for
> > tournament standings. Then, it has a clause which better defines what "Y"
> > goals mean: playing for a game win if obtainable or, failing that, playing
> > for the most victory points.
> >
> > So, claiming that the last clause - which more concretely defines what are
> > the "Y" goals - creates double standards, essentially you wish to get rid of
> > it and replace it with more examples of possible "X" goals. What is wrong
> > with this picture? But this is all any reasonable person can get from your
> > reasoning so far.
>
> No, the "Y" goals are the ones that are already in the rulebook. If
> you're a sucker for repetition, you could recopy the whole section
> there if you'd like.

But there are important things missing from the statement of what are the "Y"
goals which are contained in the section you'd like to eliminate.

> The "X" goals just specify conditions that would
> be different for a tournament situation, including additional
> considerations for tournament play that would likely not appear, or be
> excused in casual play.

Agreed, I think. You understand that the things listed as "X" goals were
not intended to be limited to just the specific things mentioned in 4.8,
right?

I'm not sure what point you made, frankly.

> > > The "other out-of-game"
> > > considerations is exactly the change that adds something to the actual
> > > rule. I think I've lost your point here. Sure VEKN ratings and overall
> > > tournament are further examples of out of game considerations.
> >
> > I have lost the point, you're right. The rule already said that. All
> > you did was to add was some more examples of things that might be
> > out-of-game considerations.
>
> Yes, to close the hole that 'a player helping his friend win' creates
> (by adding in helping as well as attacking).

There was no such hole. A player who is helping his friend win is almost
certainly doing something in conflict with the "Y" goals and hence is
running afoul of PTW as it is currently written.

The exception would be times after he'd already secured a game win or times
when he had no real hope of gaining a victory point. If you're worried
about players' game conduct during those times, then I would write the
additional clause much differently, I think, to make it clear that you can't
do these things whether they're in conflict with "Y" goals or not.

But do we need such a change? Are we having problems with players helping
their friends after they've already secured game wins or were almost eliminated?
In any case, this wouldn't appear to me to have anything to with the double
standard issue you raised at all. So once again, you're confusing me about
what end you're actually trying to accomplish.

> > > > > *** (Remove section about VP's & GW's)
> >
> > > > Removing this, on the other, is a positive detriment to the rule. The
> > > > section about VPs and GWs is very important. It defines what, exactly,
> > > > is meant by "playing in goals which are in conflict with the goal of the
> > > > game as stated in the V:TES rulebook".
> >
> > > No, the rules in the V:TES rulebook define what the goals of the game
> > > are.
> >
> > Not in that precise manner, they don't. They define winning the game. They
> > define victory points. But they're ambiguous about how these two goals
> > relate to one another.
>
> Huh? How are these two below sentences ambiguous about how victory
> points and winning relate at all?
>
> "Your goal is to accumulate the most victory points by destroying the
> influence held by rival Methuselahs."
>
> And
>
> "the winner is the player with the most victory points at the end of
> the game"
>
> It seems to inextricably link winning with earning the most VPs.

Ironically, they're ambiguous in the same way you were telling LSJ
that victory points and tournament rule game wins caused ambiguous
goals: that sometimes a player's priority might be to get a single
sure victory point whereas other times, a player might eschew a
a single sure VP to attempt to score lots of VPs and...win the game.
I guess what this boils down to is that the clause of PTW you want
to axe just redefines "winning the game" in view of the tournament
realities, which include timeouts and the tournament definition of
"game win". The only other thing it effectively does is let players
do anything they want once they're assured of a game win. I suppose
it's philosophical and arguable whether it should (John Whelan's
issue about whether player's are required to continue to maximize
VPs after they have a game win locked in) or not but again, this
was not what you were originally complaining about.

> > > What's more important? Maximizing my chances for
> > > a sweep? Or just making sure I get the most victory points.
>
> These seem to be the same consideration - unless you're talking about
> an odds gamble?

I'm talking about an odds gamble. What else is there? What else
bothers you about allowing for the tournament definition of a game
win in rule 4.8?

> If you are talking about an odds gamble - either option is fine - you
> are playing towards getting the most VPs - maybe you might think the
> risk to get 5VPs is far too high, and you're better off playing to get
> only 3 VPs which you have a much higher certainty for. It doesn't
> matter, so long as your goal is to get the MOST victory points at the
> end of the game.

OK. That's what I think, too.

And in the existing rule, it doesn't matter so long as your is to get
a game win or, failing that, to get the MOST victory points at the
end of the game. How is that different - aside from the good thing
that PTW currently recognizes that tournament game wins are what players
are _supposed_ to be playing for?

> > Also, since
> > the tournament structure has its own definition of a what a "game win" is
> > and it has the concept of a game timing out and separately defines how many
> > victory points are awarded when a game times out. Because the rulebook
> > rules don't recognize any of these points, the PTW rule needs to make some
> > allowance for them or players might be in jeopardy of having to play towards
> > truly counterproductive goals (as "playing for tournament standings"). For
> > instance, if I have two victory points, I don't need to do anything else
> > except to make sure no single player gains a second one to get a game win.
> > The PTW rule should recognize this point. The rulebook rules by themselves
> > do not. On the other hand, if I have only one victory point and no one else
> > has any and the game is in danger of timing out, I would fail to get a game
> > win by tournament rules but not by rulebook rules. It should not be OK for
> > me to play for a timeout in this situation.
>
> Agreed, and all very valid points. The concept of 'game win' in
> tournament rules certainly warrant some sort of extra definition in
> the play-to-win rule. That's why I agree that solely abandoning it
> isn't the best option (which I've done after looking at the situation
> carefully, and getting good feedback), but I still feel it could be
> written better.

OK. This is the first I've heard that you've conceded that. We're
getting somewhere...

> > This is an important clause. It creates no double standard that didn't
> > already exist. I'm completely confused why it bothers you.
>
> It does because it creates 2 goals - getting GWs and getting VPs -
> whereas the rulebook only has 1 goal - get more VPs than everyone
> else.

Nope. There's no difference between the two in any significant
way - except allowing for the tournament game win. There's nothing
about getting VPs until you get a game win that's any different from
getting VPs until you've got the most VPs. Either way, short term gain
vs. long term goals are judgement issues but there is no new conflict
created that wasn't already there.

I will point out to you that:
1) There is also a "game win" rule in the original rulebook rules.
(The "game win" rule in the original rules states that you win
the game if you have the most victory points. In a five player
game, this is 5-0-0-0-0, 4-1-0-0-0, 3-2-0-0-0, or 3-1-1-0-0, or
2-1-1-1-0. Nothing else qualifies.)

and
2) There's no definition about which is more important: getting VPs
or getting the game win.

Here's where the two would come into conflict:
* I have one victory point in a 5 player game. My current predator
also has one, my current prey has zero.
* I can attack my prey this turn, which will grant me a 10% chance of
ousting him. If I oust him, I will have a 100% chance of ousting my
predator as well. But if I fail, my prey will have a 100% chance of ousting
both my predator and myself.
* I can hang back this turn, in which case I can oust my prey for certain
next turn. But then my predator will have a 100% chance of ousting me
on his following turn.

So there you have it: 10% of gaining a 4-1-0-0-0 win and a 90% chance
of winding up losing 3-1-1-0-0 vs. a 100% chance of losing 3-2-0-0-0.
Which is more important? Going for the win or maximizing victory
points? If maximizing victory points, I should go for the sure VP
and let the RULEBOOK RULES game win slip away. If only having the
most victory points at the end of game is important, I should go for
the 10% chance of accomplishing that. The rulebook gives no guidance
which is more important. This is the same thing that you're complaining
about PTW.

> By giving one goal priority over the other (to the point that
> the other goal is excluded until the first goal is unattainable),
> along with the decision of how reasonable it is to expect people to go
> after the two different goals creates a double standard.

People don't need to go after two different goals; the "most victory
point" thing is only enforced in situations where they no longer can
hope to get a game win. How is this one iota different from following
the rulebook goals which, likewise, effectively state that your goal
is to get five victory points (in a five player game) but failing that
to get four victory points but failing that to get three victory points
but failing that to get two victory points but failing that to get
one victory point? See, the rulebook rules actually give players
*five* different goals! Of course, if you break it down that way, the
PTW rule gives players a number of different goal levels as well (which
is intended), but it's all the same thing. I don't see why you're
hung up on the tournament game win thing. And it needs to be there
anyway so it I don't see how you could improve on whatever it is that's
bothering you.

Fred


Frederick Scott

unread,
Feb 20, 2009, 6:05:09 PM2/20/09
to
"The Name Forgotten" <tex...@telkomsa.net> wrote in message
news:5d4c6d79-c8a5-4f99...@h5g2000yqh.googlegroups.com...

> > > Well - this is the rulebook goal with which the other goals must no
> > > conflict (it's one of the first things in the rulebook):
> >
> > > Object of the Game
> > > Your goal is to accumulate the most victory points by destroying the
> > > influence held by rival Methuselahs. You accomplish this goal by using
> > > your influence to gain control of younger vampires and using those
> > > vampires to take actions to reduce the influence of rival Methuselahs.
> > > Influence is represented by pool counters (see Equipment Needed, sec.
> > > 1.2), the main currency of the game. When a Methuselah runs out of
> > > pool counters, she* is bereft of influence and ousted from the game.
> > > As Methuselahs are ousted, players earn victory points; the winner is
> > > the player with the most victory points at the end of the game (see
> > > Victory Points, sec. 9.1).
> >
> > Right. You'll notice there are two separate goals listed. Because
> > there are two, it's ambiguous from the clause above, which is actually
> > the more important goal.
>
> There is? Please explain? I only see one - get the most victory
> points.

No, there's also a game win rule there. It says that you win the game if
you have more victory points than anyone else. This is every bit as much
a game win rule as the tournament game win rule and it conflicts with the
goal of gaining victory points in exactly the same way (meaning it doesn't
really conflict that much but to whatever exent it can be seen to conflict,
they both conflict the same way).

> > So removing the clause in the PTW rule which
> > sorts this out would return this situation to its ambiguous state.
>
> If the rulebook is ambiguous then the rulebook should change.

It's not really the point. I don't see the problem either way.

> > Furthermore, the game win rule in VEKN tournaments was create for some
> > very good reasons. But they don't exist in the rulebook rules. So,
> > left to itself, it could be considered to be an "out-of-game" consideration
> > under PTW to play for a tournament game win without this clause. That
> > would be bad. The clause exists, in part, to give recognition to the
> > point that players are supposed to be playing for a tournament game win
> > and that goal should considered an in-game goal.
>
> Yes, I've noticed this to, which is why I've said the rule needs to be
> amended, and not solely discarded.

Eliminating the clause which refers to the tournament game win certainly doesn't
satisfy this requirement. I don't know what you're proposing that would
simultaneously recognize that players are playing for a tournament game win
yet doesn't allow them to do so in conflict with the goals of getting the
most victory points they can get.

> > > > It's defined in that section! Why
> > > are we getting rid of this crucial piece?
> >
> > > Because the Game Win section seems to create a double standard. (I'll
> > > try explain this at the end).
> >
> > > Note that going for a Game Win is not an out-of-game consideration,
> >
> > Ah, but without this clause, it is!
>
> No - it's still an in-game consideration. Although it might not be
> spoken of in the rulebook - it is spoken of in the tournament rules.

Well yes. It would be considered "playing for tournament standing" - which
is not allowed by rule 4.8.

> It relates directly to the specific game. Out-of-game considerations
> are not the same thing as out-of-rulebook considerations.

OK, never mind my "playing for tournament standing" comment above, trying
to claim that playing for a tournament game win is "playing for an
in-game consideration" because it "relates directly to the specific
game" doesn't work. Everything one does in a game "relates to the
specific game" - so nothing would be illegal by such an interpretation.
If I'm trying to get my friend in the finals, I'm doing it by helping
him get victory points in this specific game, so THAT "relates to the
specific game" as well.

> You don't
> have to go to far to find other in-game considerations that are not
> dealt with in the rulebook, such as deal-making, that are dealt with
> in the tournament rules and considered perfectly legal.

Deal-making is a game tactic, not a goal. You mean, *keeping* a deal
made in such a way that is in conflict with PTW - and that's specifically
accounted for in the PTW rule. Pick another example, if you can find one.

> > > > so, if anything, the Play-to-Win rule should include an exception
> > > > clause that allows for in-game considerations, such as it's already
> > > > present deal-making exception, for going for Game Wins as opposed to
> > > > VPs - which may be a problem in some corner-case examples.
> >
> > I'm not sure what you're saying here. The clause you want to eliminate
> > is the clause that creates the exception that allows players to go for
> > a game win - and that's a _good_ thing.
>
> Yes, but the rule demands that you go for the game win as a priority
> and in ignorance of victory points. That is a _bad_ thing.

No, it isn't. If you didn't allow players to go for game wins in
preference to maximizing victory points, this would damage the original
reason tournament "game wins" were defined. It is *intended* that
players should - first and foremost - be playing for game wins. That's
why tournaments use game wins as the most important determiner of who
gets into the final and the second most important determiner of who wins
the final. Tournament finals were once resolved *only* using victory
points with no reference to game wins at all. This caused problems.
That's why the rule was changed. What you propose to do is worse than
what we had originally: creating tournament standings which recognize that
tournament game wins are the most important thing but - by rules - not
allow them to pursue game wins in preference to maximizing victory
points. Now *that* would cause problems!

> > > > Nothing you added fixes the problem you just caused by removing the
> > > > section about VPs and GWs. What you added talks about the potentially
> > > > conflicting goals, not the ones which must not be conflicted.
> >
> > > Sorry, I assumed that people were already familiar with the Object of
> > > the Game, as stated in the rulebook.
> >
> > I'm sorry, that response makes no sense. You added more examples to
> > what can be considered out-of-game goals. That's all you did AFAICS.
>
> Not only examples - but I also included that you cannot play solely to
> _benefit_ another player, just as you cannot play solely for the
> detriment of another player (for out of game reasons of course).

This was already so.it

> > > Welcome to my point - how does a judge decide what an obvious
> > > deliberate marking is?
> >
> > Huh?!? I don't think that's a valid question. I don't see it being
> > very hard to determine what is a deliberate, intentional mark and what
> > appears to be an accident. I guess if it could reasonably be an accident,
> > like a dog-eared card sleeve, you have to assume it was an accident. If
> > it couldn't be an accident, you assume it was deliberate. Players won't
> > usually be very successful at cheating by marking their cards in ways
> > that could be mistaken as accidental because such markings are very easy
> > to detect. If that's the cheater's game, then you may have to assume it
> > was an accident the first time you see it but keep an eye on him.
>
> That's exactly my point with the colluding case. You'd spot something
> strange - you'd say "hey, that looks an awful lot like colluding",
> and

The two cases are different. In the 'marked cards' case, the offending
player actually _did_ something bad - have cards that were marked. Whether
it was an innocent accident or on purpose, he still actually did something.
If you think it's an accident, you can treat it lightly - telling him to fix
the problem and that's the end of it. If he was trying to cheat, he will
do it again and you will catch him again, at which point it's perfectly
reasonable to escalate the penalties.

On the other hand, in your so-called "collsion" example, Player A did
*nothing, *zippo* *nada* *not-a-damned-thing*. So you can do nothing
to him no matter how long this keeps up.

If the other player, Player B, continues to play to make Player A
win, then he is still guilty of violating PTW and you can certainly
escalate the penalties on him. But you still can't do anything to
Player A because passively receiving aid from Player B isn't colluding.

> > > The same applies in collusion -
> >
> > Nope. To collude, both players have to be involved in attempting to
> > collude. It's really very hard to prove that with the kind of
> > collusion you're talking about. But that's OK. It's relatively easy to
> > prove that Player B was not playing to win (in the situation you
> > described) - so just enforce the PTW rule and you'll have effectively
> > stopped the collusion...if it was collusion.
>
> True. But even with the current PTW rule - if the person breaks it AND
> you strongly suspect he was colluding - the person should be called up
> on both charges. Agreed?

Nope. You really don't know whether it was collusion or not, and it doesn't
matter. So no. Don't worry about collusion. If players play to win,
that form of collusion doesn't matter anyway so don't worry about it.
I sure don't.

> > > While I agree that the judge *should* hold up the same standard in
> > > both situations, given the way that the PTW clause is written at the
> > > moment - it seems counter-intuitive.
> >
> > I don't know what to say. If other judges understand the meaning of
> > the rule, then they, too, should uphold the same standard in both
> > situations (if they were truly the same situations). It doesn't
> > make sense to change a rule out of fear that judges will rule it
> > wrongly.
>
> No, my concern is that either you follow your intuition and the spirit
> of play to win and rule differently in the two situations or you
> discard your intuitions and uphold the same standard - that seems to
> be against the spirit of PTW.

How is that different from "fearing that judges will rule it wrongly".

> > If that's your fear, then you ask that the same rule be
> > maintained but rewritten in a way that's more clear. Your proposal
> > actually changes the rule, so wouldn't be the correct resposne to
> > problem you're citing.
>
> Moot.

I really don't think so. You're fearing judges will make incorrect
rulings unless I'm missing something - so clarify the rule, don't change.

> > > That's my problem in a nutshell, and that's why I'm exploring
> > > alternatives to the way the PTW clause is written.
> >
> > > Hopefully, it's a little more clear?
> >
> > I guess that's more clear, if I understood your explanation. But I think
> > you may misunderstand the application of the PTW rule if you think it's
> > causing a problem in this sense. It shouldn't be.
>
> I might be - in fact I hope it is just me and my understanding of the
> rule. If it is, at least this debate will clear the matter up for
> others as well.
>
> Thanks, Fred. I know most people would've given up debating rationally
> by now. I'm really glad you stuck it through and are helping me out
> with this issue (despite it getting frustrating at times).

Sure. When someone seems to be advocating something I don't understand, I
get insanely curious where they're coming from. Is it their misunderstanding?
Is it mine? Both? I usually learn a bunch of stuff, like in this case, that
the judge's guide actually contains important tournament rules that aren't in
the tournament rules. (BAD - IMHO!) I did not know that.

Fred


The Name Forgotten

unread,
Feb 21, 2009, 4:48:36 PM2/21/09
to
On Feb 20, 11:08 pm, LSJ <vtes...@white-wolf.com> wrote:
> The Name Forgotten wrote:
> >> Right.  You'll notice there are two separate goals listed.  Because
> >> there are two, it's ambiguous from the clause above, which is actually
> >> the more important goal.  
>
> > There is? Please explain? I only see one - get the most victory
> > points.
>
> Correct. And it's listed twice.
>
> The ambiguity stems from reading the two instances differently.
> Typically by reading the first "most" as "get as many VPs as you can" and the
> second "most" as "get more VPs than anyone else".

Ah, understood. But given that none of the 'most' specifies "get as
many VPs as you can" and one of the 'most' does specify "get more VPs
than anyone else" it's more logical that the meaning of 'most' refers
to "get more VPs than anyone else".

> But if you read them both as "end with more VPs than anyone else", there's no
> discrepancy.
>
> The goal of the game is to have more VPs than anyone else at the end of the game.
>
> The tournament rules reflect that. And they add an additional goal for player
> who cannot reach that main goal: get as many VPs as you can.

On this point, going over my previous example (which was an example of
attempting to go for a GW without going for VPs as well), I realise my
example was in error, as it neglected to take into consideration the
VP one gets for surviving.

After our discussion about the matter, and the points that you have
raised, wouldn't it be better if we do rephrase the current PTW clause
in the current manner:

For tournaments, playing to win means playing to get a Game Win or to
get as many Victory Points as is reasonably possible.

My reasoning for this is, as I said, for those situations where you
are tied between the choice of attempting to go for the smaller chance
at a GW, or the much greater likelihood of a VP without a GW. The
current clause makes your choice dependent on whether or not you can
reasonably achieve a GW without considering risked VPs (potential or
real) at all, whereas my suggested alternative allows the player to
have the choice to pursue either course, and only once they have
failed both requirements will they be able to self-oust or do whatever
they want (which, perhaps, should also be clarified in the official
tournament rules.)

I agree that the current writing usually leads to the same result, but
this would close down some corner-case situations where a judge could
be called in to say that a person is not taking a reasonable chance at
a GW (because they are going for more-certain VPs instead).

This would also allow a player, once they've 'sewn-up' the GW, so to
speak, to still have the option to distribute VPs to the other players
(which is in line with the current writing of the PTW clause). I'm not
sure that this is a good thing - but I don't find anything wrong with
it either.

LSJ

unread,
Feb 21, 2009, 5:27:00 PM2/21/09
to
The Name Forgotten wrote:
> After our discussion about the matter, and the points that you have
> raised, wouldn't it be better if we do rephrase the current PTW clause
> in the current manner:
>
> For tournaments, playing to win means playing to get a Game Win or to
> get as many Victory Points as is reasonably possible.

That's the current rule: GW is possible, otherwise as many VPs as possible.

> My reasoning for this is, as I said, for those situations where you
> are tied between the choice of attempting to go for the smaller chance
> at a GW, or the much greater likelihood of a VP without a GW. The
> current clause makes your choice dependent on whether or not you can
> reasonably achieve a GW without considering risked VPs (potential or
> real) at all, whereas my suggested alternative allows the player to
> have the choice to pursue either course, and only once they have
> failed both requirements will they be able to self-oust or do whatever
> they want (which, perhaps, should also be clarified in the official
> tournament rules.)

Sure. They can do that now. Risk analysis is allowed.

> I agree that the current writing usually leads to the same result, but
> this would close down some corner-case situations where a judge could
> be called in to say that a person is not taking a reasonable chance at
> a GW (because they are going for more-certain VPs instead).

I suppose there may be some over-reaching judges out there.

But they'll interfere with the game inappropriately no matter how the rule is
written.

The Name Forgotten

unread,
Feb 21, 2009, 6:27:51 PM2/21/09
to
On Feb 21, 12:16 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:d46ef596-5535-446a...@m40g2000yqh.googlegroups.com...

>
> > On Feb 20, 8:07 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > > "The Name Forgotten" <tex...@telkomsa.net> wrote in
> > > messagenews:fad3ba08-d0b4-4c58...@f24g2000vbf.googlegroups.com...
> > > > On Feb 20, 12:02 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> > > > > "The Name Forgotten" <tex...@telkomsa.net> wrote in
> > > > > messagenews:130898ea-7fe0-4c63...@p23g2000prp.googlegroups.com...
>
> > > > > > On Feb 19, 6:47 pm, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
>
> > From the introduction of the V:EKN Tournament Rules:
>
> > "Players who violate sections of the V:EKN Tournament Rules will be
> > subject to the appropriate provisions of the V:EKN Judges' Guide."
>
> > The Judge's Guide is very much part of the tournament rules.
>
> Two responses:
>
> 1) I'll admit, it's been a while since I've studied the explanations of cautions,
> warnings, removal, and so forth. So I'll confess that I did not know that was true.
> I will further venture that I think _that_ is also a bad idea for exactly the
> reason I just gave: a judge's guide should just be a guide to judging. It shouldn't
> give rules. If you've written your documents in such a way that your judges guide
> is required to understand the rules of tournament play, then it isn't a judge's
> guide. It's part of the rulebook. Call it that.

This will lead onto another issue which should best be dealt with on
it's own. I personally, have no problem with the way the Judge's Guide
has been written, and I've no problem with the fact that a judge is
required to read it, and understand it to judge correctly. I don't
really see it as giving rules so much as clarifying them. But I can
see your point as well, and, to be honest, I'm not sure which is the
more valid point.

> 2) Even so, this statement does not claim to contain tournament rules outside of
> the application of penalties. It just says the judge's guide explains what happens
> *WHEN* rules are violated. So my compaint about the lack of clarity of this clause
> stands.

As I've said, it doesn't contain any tournament rules - it simply
clarifies the issue of what is, and is not, to be considered
collusion, i.e. what is considered to be a violation of the rules. The
judge's guide states:

"The V:EKN Judges' Guidelines (formerly called 'Penalty Guidelines')
provide a structure to help judges resolve play errors

and

determine the appropriate penalty for infractions that occur during
the course of a tournament."

In other words, it has two purposes - not only to explain what happens
when rules are violated - but also to explain when the rules are, in
fact, violated, and what the different degrees of violation are - i.e.
the structure that helps judges resolve play errors.

At least, that's how I see it. As I said - I do think your point is
valid, but I don't think the current structure is overly problematic
in that sense.

> > > Frankly, I hate rules that can so easily misinterpreted. That's why I said:
> > > this rule needs to be clarified (in the *TOURNAMENT RULES*) or eliminated.
> > > Since I really don't see what function it serves, given PTW, I'd say it should
> > > be eliminated.
>
> > Well, where it's stated isn't important
>
> It sure is. Rules should be understood from what you can see in the rulebook.
> If you can't understand them at all, then they're poorly written. This clause
> cannot be understood at all from what's written in 5.1.

A valid point.

> > > I think it should give you pause to consider whether your idea that the
> > > clause you want to eliminate actually creates double standards. I mean,
> > > to reduce your argument down to it's essence, this is what I've gotten
> > > to date: What the Play-to-Win rule actually says is, "It should be
> > > considered unsportsmanlike conduct to play a game towards goals NOT written
> > > the rulebook when they're in conflict with the goals written in the rulebook."
> > > In essence, it says, you can't play toward "X" kinds of goals where they're in
> > > conflict with "Y" kinds of goals - "X" being out-of-game game goals, "Y"
> > > being the goals written in the rulebook. Then, it gives a couple examples
> > > (but not limited to...) of "X" goals: playing for VEKN ratings or playing for
> > > tournament standings. Then, it has a clause which better defines what "Y"
> > > goals mean: playing for a game win if obtainable or, failing that, playing
> > > for the most victory points.
>
> > > So, claiming that the last clause - which more concretely defines what are
> > > the "Y" goals - creates double standards, essentially you wish to get rid of
> > > it and replace it with more examples of possible "X" goals. What is wrong
> > > with this picture? But this is all any reasonable person can get from your
> > > reasoning so far.
>
> > No, the "Y" goals are the ones that are already in the rulebook. If
> > you're a sucker for repetition, you could recopy the whole section
> > there if you'd like.
>
> But there are important things missing from the statement of what are the "Y"
> goals which are contained in the section you'd like to eliminate.

Yes, that's what I've been exploring - i.e. what are the exact things
that are missing? After debating and pulling it apart, I think I've
started to formulate what really is at the heart of the PTW clause and
how it can be rewritten in a *slightly* better way. As I've recently
responded to LSJ on the matter. If you haven't been following the rest
of the thread I can pass on the current link of where I am on that
issue.

> > The "X" goals just specify conditions that would
> > be different for a tournament situation, including additional
> > considerations for tournament play that would likely not appear, or be
> > excused in casual play.
>
> Agreed, I think. You understand that the things listed as "X" goals were
> not intended to be limited to just the specific things mentioned in 4.8,
> right?
>

Yeah - that's sounds about right.

> > > Essentially, "for instance". But my version was more consise and made
> > > better sense than your's. Your's said, "(for instance) race, gender
> > > animosity, or other-of-game considerations..." adding another 'e.g.'
> > > there but essentially means everything else that's not an in-game
> > > consideration. Everything you mentioned was an out-of-game consideration
> > > so...why not just say THAT?!? Basically, you're not suppose to play
> > > for out-of-game considerations when they're in conflict with rulebook
> > > game goals - and here's a lot of possible examples. That's much more
> > > straightforward than what you wrote.
>
> > True, my version could be more refined. It was written quickly and
> > indelicately. It was written like that because I wanted to keep as
> > much of the original ruling intact without inserting bit's and pieces
> > in several different areas. I think the relevant point was made
> > clearly enough to be understood.
>
> I'm not sure what point you made, frankly.

The point is that, in the spirit of PTW, a player shouldn't play to
_benefit_ another player for out-of-game considerations, as much as a
player shouldn't play to _attack_ another player for out-of-game
considerations. Is that a bit more clear? Have I got the point across?

> > > > The "other out-of-game"
> > > > considerations is exactly the change that adds something to the actual
> > > > rule. I think I've lost your point here. Sure VEKN ratings and overall
> > > > tournament are further examples of out of game considerations.
>
> > > I have lost the point, you're right. The rule already said that. All
> > > you did was to add was some more examples of things that might be
> > > out-of-game considerations.
>
> > Yes, to close the hole that 'a player helping his friend win' creates
> > (by adding in helping as well as attacking).
>
> There was no such hole. A player who is helping his friend win is almost
> certainly doing something in conflict with the "Y" goals and hence is
> running afoul of PTW as it is currently written.

True, but the same could be said for attacking players due to their
VEKN ratings etc. It conflicts with the "Y" goals as well, in that it
makes the pursuit of GWs or VPs harder.

> The exception would be times after he'd already secured a game win or times
> when he had no real hope of gaining a victory point. If you're worried
> about players' game conduct during those times, then I would write the
> additional clause much differently, I think, to make it clear that you can't
> do these things whether they're in conflict with "Y" goals or not.
>
> But do we need such a change? Are we having problems with players helping
> their friends after they've already secured game wins or were almost eliminated?

Very good point. To be honest, I didn't even think about that - but it
could be worth considering for that reason alone.

> In any case, this wouldn't appear to me to have anything to with the double
> standard issue you raised at all. So once again, you're confusing me about
> what end you're actually trying to accomplish.

Well, it only has something to do with the double standard issue in so
far as the double standard issue asks for the removal of the "Y" goal
set out in that PTW clause (i.e. go for GW, then go for most VPs
possible). It, really, does have very little to do with my initial
point though, which is kind of why I wanted to avoid going to far down
that debate.

I've agreed though, that once it is removed, however, there is a big
gaping whole left of what to do for players who cannot get the most
VPs at the table, so something needs to be put in it's place.

>
> > > Not in that precise manner, they don't. They define winning the game. They
> > > define victory points. But they're ambiguous about how these two goals
> > > relate to one another.
>
> > Huh? How are these two below sentences ambiguous about how victory
> > points and winning relate at all?
>
> > "Your goal is to accumulate the most victory points by destroying the
> > influence held by rival Methuselahs."
>
> > And
>
> > "the winner is the player with the most victory points at the end of
> > the game"
>
> > It seems to inextricably link winning with earning the most VPs.
>
> Ironically, they're ambiguous in the same way you were telling LSJ
> that victory points and tournament rule game wins caused ambiguous
> goals: that sometimes a player's priority might be to get a single
> sure victory point whereas other times, a player might eschew a
> a single sure VP to attempt to score lots of VPs and...win the game.

Yes, but with only one clear thing to acquire, i.e. VPs, you can
include the acquisition of less VPs with a better certainty into your
risk-analysis. The moment you introduce GWs as your only goal to the
exclusion of VPs until such point as going for the GW is no longer
reasonable, then what you are saying is that risking a better chance
at getting VPs is not even a consideration to be factored into your
risk-analysis in going after the GW. That is what the current writing
of the PTW tournament clause does. It says:

"and when a Game Win is not reasonably possible, THEN" (emphasis mine,
of course)

That to me, reads as: You must pursue the GW in the same manner that
others, who can no longer possibly get the GW, must pursue VPs. Until
the GW is no longer reasonably possible, VPs have no value beyond
getting you to that goal.

> I guess what this boils down to is that the clause of PTW you want
> to axe just redefines "winning the game" in view of the tournament
> realities, which include timeouts and the tournament definition of
> "game win".

Yes, thats true, which is also why I've agreed that something does
need to take it's place to consider tournament realities.

> The only other thing it effectively does is let players
> do anything they want once they're assured of a game win.

Sorry, just to clarify, what is the 'it' in this sentence?

> I suppose
> it's philosophical and arguable whether it should (John Whelan's
> issue about whether player's are required to continue to maximize
> VPs after they have a game win locked in) or not but again, this
> was not what you were originally complaining about.

Yeah, that is a seperate issue - and one that I've said I really don't
care much about, one way or the other. Personally, I think once the
player has achieved the 'most' VPs at the table, i.e. has 3VPs or more
in a 4/5 player - the player has already proven that they are playing
to win - how players proceed from there is really their own business.
If a player really wants to give VPs away to another player, then
fine. Although, I do think that such decisions on who to benefit
shouldn't be based on those out-of-game considerations we were talking
about before. However, if the player thinks that one of the remaining
players really deserves some VPs because of a deal they'd made, or
some favour that player had given before, or whatever in-game
consideration, then that's fine in my books.

> > If you are talking about an odds gamble - either option is fine - you
> > are playing towards getting the most VPs - maybe you might think the
> > risk to get 5VPs is far too high, and you're better off playing to get
> > only 3 VPs which you have a much higher certainty for. It doesn't
> > matter, so long as your goal is to get the MOST victory points at the
> > end of the game.
>
> OK. That's what I think, too.
>
> And in the existing rule, it doesn't matter so long as your is to get
> a game win or, failing that, to get the MOST victory points at the
> end of the game. How is that different - aside from the good thing
> that PTW currently recognizes that tournament game wins are what players
> are _supposed_ to be playing for?

I think I've already given my response to this above.

> > > This is an important clause. It creates no double standard that didn't
> > > already exist. I'm completely confused why it bothers you.
>
> > It does because it creates 2 goals - getting GWs and getting VPs -
> > whereas the rulebook only has 1 goal - get more VPs than everyone
> > else.
>
> Nope. There's no difference between the two in any significant
> way - except allowing for the tournament game win. There's nothing
> about getting VPs until you get a game win that's any different from
> getting VPs until you've got the most VPs. Either way, short term gain
> vs. long term goals are judgement issues but there is no new conflict
> created that wasn't already there.

There is one significant difference - in a tournament situation, you
still want players to go for victory points, even though they are no
longer eligible to get the most VPs on the table, or a GW. To
accomodate this PTW-spirit, the tournament rules have put in a
seperate clause about this. The problem with the clause, as it is
written, is that players need to go for the GW, to the exclusion of
the pursuit of VPs, with the same amount of risk-analysis and
determination as a player who cannot possibly get a GW, but could
still go for a VP. I've made this clearer above, so I won't repeat the
whole point again.

> I will point out to you that:
> 1) There is also a "game win" rule in the original rulebook rules.
> (The "game win" rule in the original rules states that you win
> the game if you have the most victory points. In a five player
> game, this is 5-0-0-0-0, 4-1-0-0-0, 3-2-0-0-0, or 3-1-1-0-0, or
> 2-1-1-1-0. Nothing else qualifies.)

Ah, ok, I was a bit confused by your point here at first. Ok, I agree.

>
> and
>
> 2) There's no definition about which is more important: getting VPs
> or getting the game win.

Well, in terms of the tournament PTW clause - it clearly states that
the GW is more important - to the point that getting VPs isn't a
direct consideration.

In other words, if there was some other way of getting a GW without
getting any VPs, then players would be allowed to pursue that way,
even if it meant losing VPs.

For example, imagine a card came out, like Gambit Accepted, that said
"Lose 2 Victory Points to get a Game Win" - if a player had the card
in his hand, and had the requisite 2 VPs - the player would be
entitled to play it under the current PTW clause - because the player
must play to get the GW where reasonably possible. Playing that card
would give the player a 100% chance at a GW, so the player should play
it.

> Here's where the two would come into conflict:
> * I have one victory point in a 5 player game. My current predator
> also has one, my current prey has zero.
> * I can attack my prey this turn, which will grant me a 10% chance of
> ousting him. If I oust him, I will have a 100% chance of ousting my
> predator as well. But if I fail, my prey will have a 100% chance of ousting
> both my predator and myself.
> * I can hang back this turn, in which case I can oust my prey for certain
> next turn. But then my predator will have a 100% chance of ousting me
> on his following turn.

Ok, a bit difficult to follow, but I think I understand. I'm not sure
if such a situation could exist, and I'd be very hard pressed to think
up of such a situation, but I'll try to imagine it's possibility.

> So there you have it: 10% of gaining a 4-1-0-0-0 win and a 90% chance
> of winding up losing 3-1-1-0-0 vs. a 100% chance of losing 3-2-0-0-0.
> Which is more important? Going for the win or maximizing victory
> points? If maximizing victory points, I should go for the sure VP
> and let the RULEBOOK RULES game win slip away. If only having the
> most victory points at the end of game is important, I should go for
> the 10% chance of accomplishing that.

If that is what you consider to be a reasonable chance, then yes, you
should. I've already gone over the risk-analysis issue - the problem
arises in the tournament situation because you want people to go for
VPs (if they have a reasonable chance) when they can no longer get the
GW. If that's the case, then you have to use the same level of risk-
analysis to determine when a person can be reasonably said to go for
the GW (or the most victory points) as you do when you determine when
a person can be reasonably said to go for a VP.

> The rulebook gives no guidance
> which is more important.

The rulebook gives the importance to going for the most VP's at the
end of the game.

> This is the same thing that you're complaining
> about PTW.

No. My complaint is that the level (or standard) of risk-analysis
applied to whether or not a person should be bound to go for a GW,
doesn't appear to be the same (or even intuitively feel the same) as
the level (or standard) of risk-analysis applied to whether or not a
person should be bound to go for a VP (when going for a GW is no
longer a reasonable option). While I agree that you could try to
enforce that the standard be kept the same in both situations, that
just seems to be intuitively wrong, as the person who decides not to
go for a risky GW may be doing it for reasons that seem to fall in
line with the spirit of play to win.

> > By giving one goal priority over the other (to the point that
> > the other goal is excluded until the first goal is unattainable),
> > along with the decision of how reasonable it is to expect people to go
> > after the two different goals creates a double standard.
>
> People don't need to go after two different goals; the "most victory
> point" thing is only enforced in situations where they no longer can
> hope to get a game win.

Precisely.

> How is this one iota different from following
> the rulebook goals which, likewise, effectively state that your goal
> is to get five victory points (in a five player game) but failing that
> to get four victory points but failing that to get three victory points
> but failing that to get two victory points but failing that to get
> one victory point?

No. The rulebook says you should get the most VPs of all the players -
not the most VPs you could get.

> See, the rulebook rules actually give players
> *five* different goals! Of course, if you break it down that way, the
> PTW rule gives players a number of different goal levels as well (which
> is intended), but it's all the same thing.

Moot.

> I don't see why you're
> hung up on the tournament game win thing.

Because it would force me, as a judge, to hold the same standard in
the two situations, even though in the one it seems like the person is
following the spirit of play-to-win, but I have to rule against the
person because of the way the clause is currently written and to keep
the same standard for the player who wants to give up when that player
still has a reasonable chance at getting a VP and nothing substantial
to lose by going for it.

> And it needs to be there
> anyway so it I don't see how you could improve on whatever it is that's
> bothering you.

Well, I wouldn't know how either until I try. This is me trying... 8)

The Name Forgotten

unread,
Feb 21, 2009, 7:28:00 PM2/21/09
to
On Feb 21, 1:05 am, "Frederick Scott" <nos...@no.spam.dot.com> wrote:
> "The Name Forgotten" <tex...@telkomsa.net> wrote in messagenews:5d4c6d79-c8a5-4f99...@h5g2000yqh.googlegroups.com...
>
>
> > There is? Please explain? I only see one - get the most victory
> > points.
>
> No, there's also a game win rule there.  It says that you win the game if
> you have more victory points than anyone else.  This is every bit as much
> a game win rule as the tournament game win rule and it conflicts with the
> goal of gaining victory points in exactly the same way (meaning it doesn't
> really conflict that much but to whatever exent it can be seen to conflict,
> they both conflict the same way).

Ah - I see your issue - you are reading 'get the most victory points'
as 'get the most victory points you could possibly get'. However, the
rulebook never says that. It's a misinterpretation on your behalf, I
think (barring some clarification from LSJ).


> > > Furthermore, the game win rule in VEKN tournaments was create for some
> > > very good reasons. But they don't exist in the rulebook rules. So,
> > > left to itself, it could be considered to be an "out-of-game" consideration
> > > under PTW to play for a tournament game win without this clause. That
> > > would be bad. The clause exists, in part, to give recognition to the
> > > point that players are supposed to be playing for a tournament game win
> > > and that goal should considered an in-game goal.
>
> > Yes, I've noticed this to, which is why I've said the rule needs to be
> > amended, and not solely discarded.
>
> Eliminating the clause which refers to the tournament game win certainly doesn't
> satisfy this requirement.  I don't know what you're proposing that would
> simultaneously recognize that players are playing for a tournament game win
> yet doesn't allow them to do so in conflict with the goals of getting the
> most victory points they can get.

Well, my proposal is to let the players have the choice to go for both
or either (in the case that they conflict).

> > > > > It's defined in that section! Why
> > > > are we getting rid of this crucial piece?
>
> > > > Because the Game Win section seems to create a double standard. (I'll
> > > > try explain this at the end).
>
> > > > Note that going for a Game Win is not an out-of-game consideration,
>
> > > Ah, but without this clause, it is!
>
> > No - it's still an in-game consideration. Although it might not be
> > spoken of in the rulebook - it is spoken of in the tournament rules.
>
> Well yes.  It would be considered "playing for tournament standing" - which
> is not allowed by rule 4.8.

Incorrect. What is not allowed by 4.8. is _attacking_ other players
based on tournament standings. It says nothing about playing your own
game to improve your tournament standing.

> > It relates directly to the specific game. Out-of-game considerations
> > are not the same thing as out-of-rulebook considerations.
>
> OK, never mind my "playing for tournament standing" comment above, trying
> to claim that playing for a tournament game win is "playing for an
> in-game consideration" because it "relates directly to the specific
> game" doesn't work.  Everything one does in a game "relates to the
> specific game" - so nothing would be illegal by such an interpretation.
> If I'm trying to get my friend in the finals, I'm doing it by helping
> him get victory points in this specific game, so THAT "relates to the
> specific game" as well.

Obviously I don't mean it in such a broad sense. You could make the
reverse point in that everything you do in-game is playing for
tournament standing in one way or the other as well. My point is that
GWs are part of the tournament rules, hence it makes it a valid in-
game consideration in a tournament game, even though GWs are not dealt
with in the rulebook. Is that a bit more clear?

> > You don't
> > have to go to far to find other in-game considerations that are not
> > dealt with in the rulebook, such as deal-making, that are dealt with
> > in the tournament rules and considered perfectly legal.
>
> Deal-making is a game tactic, not a goal.  You mean, *keeping* a deal
> made in such a way that is in conflict with PTW - and that's specifically
> accounted for in the PTW rule.  Pick another example, if you can find one.

Well, that's exactly my point now, isn't it? Deal-making or keeping a
deal is an in-game consideration that is not dealt with in the
rulebook but is dealt with in the tournament rules. It's a 'rule' that
is dealt with inside the tournament rules, but outside of rulebook. If
you're not happy with deal-making, then what about time-outs? While
it's not a consideration under the normal rulebook, if you feel that
the table is going to time out before your predator is going to have
his/her go, then you might lunge at your prey in the hopes that you
can oust your prey for a VP, even though it would ordinarily stand to
lose you the game (if there was no time-out).


> > Yes, but the rule demands that you go for the game win as a priority
> > and in ignorance of victory points. That is a _bad_ thing.
>
> No, it isn't.  If you didn't allow players to go for game wins in
> preference to maximizing victory points, this would damage the original
> reason tournament "game wins" were defined.  It is *intended* that
> players should - first and foremost - be playing for game wins.  That's
> why tournaments use game wins as the most important determiner of who
> gets into the final and the second most important determiner of who wins
> the final.  Tournament finals were once resolved *only* using victory
> points with no reference to game wins at all.  This caused problems.
> That's why the rule was changed.  

You've missed the boat on this one. The major point I'm making is that
you need to go for GW's _ignoring_ VPs. I'm saying this is bad. You
should go for GWs, but you should also be allowed to go for VPs. When
you're in a situation that these two goals conflict - you should be
able to analyse the risks involved in both situations, and go for the
situation with the least risk involved, if you so choose. Currently,
you can only go for the least risk situation IF a certain threshold
for what can be considered a 'reasonably possibile' chance at going
for a GW is not met. This is a bad thing, especially since that same
threshold is used when you decide whether a player should go for a VP
when they have nothing else to risk in its pursuit.

> What you propose to do is worse than
> what we had originally: creating tournament standings which recognize that
> tournament game wins are the most important thing but - by rules - not
> allow them to pursue game wins in preference to maximizing victory
> points.  Now *that* would cause problems!

Note: this is not what I propose. I propose that the player has a
choice to go for either, without prejudice.

> > > > > Nothing you added fixes the problem you just caused by removing the
> > > > > section about VPs and GWs. What you added talks about the potentially
> > > > > conflicting goals, not the ones which must not be conflicted.
>
> > > > Sorry, I assumed that people were already familiar with the Object of
> > > > the Game, as stated in the rulebook.
>
> > > I'm sorry, that response makes no sense. You added more examples to
> > > what can be considered out-of-game goals. That's all you did AFAICS.
>
> > Not only examples - but I also included that you cannot play solely to
> > _benefit_ another player, just as you cannot play solely for the
> > detriment of another player (for out of game reasons of course).
>
> This was already so.it

Not in any way that I could see. If it is, could you please point it
out to me - other than by refering to the PTW clause that we were
talking about already, because, as I've said, that same clause would
also serve to eliminate the need to talk about _attacking_ players on
out-of-game considerations.

> > > > Welcome to my point - how does a judge decide what an obvious
> > > > deliberate marking is?
>
> > > Huh?!? I don't think that's a valid question. I don't see it being
> > > very hard to determine what is a deliberate, intentional mark and what
> > > appears to be an accident. I guess if it could reasonably be an accident,
> > > like a dog-eared card sleeve, you have to assume it was an accident. If
> > > it couldn't be an accident, you assume it was deliberate. Players won't
> > > usually be very successful at cheating by marking their cards in ways
> > > that could be mistaken as accidental because such markings are very easy
> > > to detect. If that's the cheater's game, then you may have to assume it
> > > was an accident the first time you see it but keep an eye on him.
>
> > That's exactly my point with the colluding case. You'd spot something
> > strange - you'd say "hey, that looks an awful lot like colluding",
> > and
>
> The two cases are different.  In the 'marked cards' case, the offending
> player actually _did_ something bad - have cards that were marked.  Whether
> it was an innocent accident or on purpose, he still actually did something.
> If you think it's an accident, you can treat it lightly - telling him to fix
> the problem and that's the end of it.  If he was trying to cheat, he will
> do it again and you will catch him again, at which point it's perfectly
> reasonable to escalate the penalties.
>
> On the other hand, in your so-called "collsion" example, Player A did
> *nothing,  *zippo*  *nada*  *not-a-damned-thing*.  So you can do nothing
> to him no matter how long this keeps up.

Yes, which is why, in my original collusion case I said accuse player
B. However, somebody said you can't accuse one without the other, and
also went on the tangent that Player A is trying A's utmost to not co-
operate with B's help, including stating loudly that A doesn't want
B's help, blocking bleeds towards A's prey, etc. That strikes me as
being a bit over-protective, and, well, suspicious. If A is involved
in suspicious behaviour, or at least looks like it (intentionally or
not), A should be told by the judge that A's behaviour looks
suspicious and why. A could say that the reason why bleeds are being
blocked, etc, is because A is terrified that A might be accused of
colluding and disqualified - at which point, I, as a judge, would tell
A not to worry about that, and play as normal - if B is the one
involved in suspicious behaviour, then B will be the one punished for
it.

I guess it was a rather snarky comment from me, and not really clear,
but there you have it.

> If the other player, Player B, continues to play to make Player A
> win, then he is still guilty of violating PTW and you can certainly
> escalate the penalties on him.  But you still can't do anything to
> Player A because passively receiving aid from Player B isn't colluding.

Agreed. So long as Player A doesn't continue to act in a very
suspicious manner.

> > > > The same applies in collusion -
>
> > > Nope. To collude, both players have to be involved in attempting to
> > > collude. It's really very hard to prove that with the kind of
> > > collusion you're talking about. But that's OK. It's relatively easy to
> > > prove that Player B was not playing to win (in the situation you
> > > described) - so just enforce the PTW rule and you'll have effectively
> > > stopped the collusion...if it was collusion.
>
> > True. But even with the current PTW rule - if the person breaks it AND
> > you strongly suspect he was colluding - the person should be called up
> > on both charges. Agreed?
>
> Nope.  You really don't know whether it was collusion or not,

You know as much as you'll ever be certain to know.

> and it doesn't
> matter.  So no.  Don't worry about collusion.  If players play to win,
> that form of collusion doesn't matter anyway so don't worry about it.
> I sure don't.

Well, then we need to agree to disagree. As a final point on the
matter, let's use an example:

Player N and Player M have met up in secret before the tournament.
They both decide to take vote decks to the tournament, and say that,
so long as they do not start as predator or prey of each other, they
will always vote in favour of each others pool-gaining political
actions (e.g. consag boons, diversity, honor the elders, exclusion
principle, etc.), and will allow each other to get 1 VP.

On the day of the tournament, one of the judges gets called over, and
witnesses this happening. Time and again, pool-gaining votes are
called and passed, thanks to the cross-table alliance, and they
obviously help each other get 1 VP. While it's not strictly against
play to win, as each of them getting a VP does nothing to stop them
from going after a GW, every bit of evidence shows that the duo are
certainly working together - and they doing so in a manner that alters
the outcome of the game.

Hence, a collusion rule is necessary, even with PTW as it currently
stands. Once there is a collusion rule in the game, as a judge, I feel
it's necessary to enforce it, regardless of whether or not a person
has broken another rule as well or not. If you think it's unnecessary,
then, as I said, that's fine - we'll just have to agree to disagree. I
don't think there's anything forcing a judge to call a person up on
every single infraction of the rules - I'm just meticulous that way.

> > > > While I agree that the judge *should* hold up the same standard in
> > > > both situations, given the way that the PTW clause is written at the
> > > > moment - it seems counter-intuitive.
>
> > > I don't know what to say. If other judges understand the meaning of
> > > the rule, then they, too, should uphold the same standard in both
> > > situations (if they were truly the same situations). It doesn't
> > > make sense to change a rule out of fear that judges will rule it
> > > wrongly.
>
> > No, my concern is that either you follow your intuition and the spirit
> > of play to win and rule differently in the two situations or you
> > discard your intuitions and uphold the same standard - that seems to
> > be against the spirit of PTW.
>
> How is that different from "fearing that judges will rule it wrongly".

Because, I fear that the judges will rule it correctly (by holding the
same standard) as well, and this will lead to counter-intuitive
results - i.e. punishing a player because their real-life intention to
play-to-win is superceeded by a PTW rule that asks them to play in a
way that's not really playing to win, but more playing at a chance of
getting a GW.

> > > If that's your fear, then you ask that the same rule be
> > > maintained but rewritten in a way that's more clear. Your proposal
> > > actually changes the rule, so wouldn't be the correct resposne to
> > > problem you're citing.
>
> > Moot.
>
> I really don't think so.  You're fearing judges will make incorrect
> rulings unless I'm missing something - so clarify the rule, don't change.

Still moot.

> > Thanks, Fred. I know most people would've given up debating rationally
> > by now. I'm really glad you stuck it through and are helping me out
> > with this issue (despite it getting frustrating at times).
>
> Sure.  When someone seems to be advocating something I don't understand, I
> get insanely curious where they're coming from.  Is it their misunderstanding?
> Is it mine?  Both?  I usually learn a bunch of stuff, like in this case, that
> the judge's guide actually contains important tournament rules that aren't in
> the tournament rules.  (BAD - IMHO!)  I did not know that.

Yeah, sure, you've given me plenty to think about as well. It
certainly has been enlightning.

0 new messages