Optimal seating order
Henrik Isaksson
At this weekends tournament in Halmstad (report to come in a few days) the
optimal seating order was questioned. Players were unsatisfied that that
there are too few players encountered, and that the variation could be
better.
I've seen the seating order been questioned before, with the standard
answers: "It's optional!"
To bring light to the discussion I will provide two examples:
5 players. Here player 2 is supposed to begin round 2 and 3. Why?
5 Players
Round 1: 1 2 3 4 5
Round 2: 2 5 3 1 4
Round 3: 2 1 5 4 3
13 players (this was the number of players in the tournament)
13 Players
Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
Round 3: 3 12 10 5, 6 4 13 11, 8 1 9 7 2
Round 1 and 2 are not a problem to anyone. Noone sits next to anyone that
they've sat at the same table as before, and the distribution seems even. In
round three though, things gets really messy. Notice players 7,9,1 at table
1 in round 2. Exactly these players sit together at table 3 in round 3, but
with reversed order. Imagine the frustration of player 9 if he plays a vote
deck and players 7 and 1 plays intercept. Also, players 4 and 6 are in
reversed order from round 2. To show that it's possible to make this better
I will provide am alternative round 3: With only a change of players 4 and
9, noone sits next to a player he's sat next to before.
Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
Round 3: 3 12 10 5, 6 9 13 11, 8 1 4 7 2
These reveresed order of two or three players from round two to three is
somehting that occurs very frequently in many charts, and it's my belief
that they all aer unneccesary. It's very boring to sit next to a deck you've
sat next to before, be it as prey or predator. I know of tournament
organizers that don't yse the seating order as they think it's more optimal
to draw the places for each round... (Hi Sten!)
The goal of the Optimal Seating Order (OSO tm) should be:
1. Ensure that noone has the same prey or predator as in a previous round.
(this is covered today)
2. Ensure that each players sees as many decks as possible, and sits next to
as many new decks as possibly. (not covered today)
The questions I would like to have answered are:
Is there an algoritm used to make the optimal seating order, and in that
case, what is it?
Am I correct in my assumption that the charts can be improved a lot?
Perhaps the "randomization" of this chart could be analyzed by some students
as some sort of exam-project?
Best Regards
Henrik Isaksson
Derek Ray
"Henrik Isaksson" <hen...@nissamedia.net> mumbled something about:
>The goal of the Optimal Seating Order (OSO tm) should be:
>1. Ensure that noone has the same prey or predator as in a previous round.
>(this is covered today)
It should also be the first priority, which it is. This is good.
>2. Ensure that each players sees as many decks as possible, and sits next to
>as many new decks as possibly. (not covered today)
Actually, it IS covered as well as possible today. The minimum number
of players required to make sure that no player sits with any other
player more than once is 16 -- because until you reach 16 players, there
are more players sitting at each table than there are tables for them to
sit at. This leads to impossible seating situations; in a 13-player
tournament, for example, you have players A through M, and tables of 5,
4, and 4 players respectively each round.
Assume seating for round 1 is A-B-C-D, E-F-G-H, I-J-K-L-M.
Now try to seat round 2. It fails immediately:
A-E-I, and the 4th player at that table MUST have played with either A,
E, or I in round 1. It simply isn't possible for less than 16 players.
>The questions I would like to have answered are:
>Is there an algoritm used to make the optimal seating order, and in that
>case, what is it?
I don't know. I started to duplicate it by hand once, and found that
they were already optimized for the under-16 players, and above 16 it
becomes easy to do. It may have been done by hand originally for all I
know -- the Archon shows some earmarks of exactly that.
>Am I correct in my assumption that the charts can be improved a lot?
Nope.
>Perhaps the "randomization" of this chart could be analyzed by some students
>as some sort of exam-project?
I assume that you are volunteering?
--
"There's no gray. There's just white that's got grubby." -- T.P.
LSJ
>
> Hi,
>
> At this weekends tournament in Halmstad (report to come in a few days) the
> optimal seating order was questioned. Players were unsatisfied that that
> there are too few players encountered, and that the variation could be
> better.
>
> I've seen the seating order been questioned before, with the standard
> answers: "It's optional!"
The "standard" answer should be more like: "You may be right - there
is no way to prove that a seating chart is optimal in most cases
(perfect cases like 16 and multiples of 25 notwithstanding). But
the only way to show that a seating chart is not optimal is to
provide a chart that is better."
> To bring light to the discussion I will provide two examples:
>
> 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> 5 Players
> Round 1: 1 2 3 4 5
> Round 2: 2 5 3 1 4
> Round 3: 2 1 5 4 3
To avoid repeating the same relative seating with another player.
If you've got a seating chart in mind that is better, please share.
> 13 players (this was the number of players in the tournament)
> 13 Players
> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
> Round 3: 3 12 10 5, 6 4 13 11, 8 1 9 7 2
>
> Round 1 and 2 are not a problem to anyone. Noone sits next to anyone that
> they've sat at the same table as before, and the distribution seems even. In
> round three though, things gets really messy. Notice players 7,9,1 at table
> 1 in round 2. Exactly these players sit together at table 3 in round 3, but
> with reversed order. Imagine the frustration of player 9 if he plays a vote
> deck and players 7 and 1 plays intercept. Also, players 4 and 6 are in
> reversed order from round 2. To show that it's possible to make this better
> I will provide am alternative round 3: With only a change of players 4 and
> 9, noone sits next to a player he's sat next to before.
> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
> Round 3: 3 12 10 5, 6 9 13 11, 8 1 4 7 2
"Next to" isn't one of the criteria. "Same relationship" is a criterion,
however.
So from the stated criteria, the alternate proposal isn't any better.
The alternate proposal has the same number of "repeat pairings at a
table" (seeing the same player twice) and of "continual pairings"
(seeing the same player in all three rounds), so no improvement there.
Any the alternate seating repeats seating position (which is one
of the stated criteria) - player 4 plays 3rd twice and player 9 plays
2nd twice. So, from the perspective of the criteria used to design
the tables, the alternate seating is less optimal than the current.
> These reveresed order of two or three players from round two to three is
> somehting that occurs very frequently in many charts, and it's my belief
> that they all aer unneccesary. It's very boring to sit next to a deck you've
> sat next to before, be it as prey or predator. I know of tournament
> organizers that don't yse the seating order as they think it's more optimal
> to draw the places for each round... (Hi Sten!)
Which is perfectly fine, so long as the VEKN rules are followed.
> The goal of the Optimal Seating Order (OSO tm) should be:
> 1. Ensure that noone has the same prey or predator as in a previous round.
> (this is covered today)
... demanded by the VEKN rules, and provided by the charts already.
> 2. Ensure that each players sees as many decks as possible, and sits next to
> as many new decks as possibly. (not covered today)
This is also one of the goals on the current charts. It has been met, TTBOMK.
> The questions I would like to have answered are:
> Is there an algoritm used to make the optimal seating order, and in that
> case, what is it?
Not that I know of.
> Am I correct in my assumption that the charts can be improved a lot?
I would guess not. But my perspective may be flawed by the amount of effort that
went in to producing them in the first place.
For some small values of "a lot", there may be improvements to be had on a
few tables, though.
--
LSJ (vte...@white-wolf.com) V:TES Net.Rep for White Wolf, Inc.
Links to revised rulebook, rulings, errata, and tournament rules:
http://www.white-wolf.com/vtes/
Derek Ray
Derek Ray <lor...@yahoo.com> mumbled something about:
>I don't know. I started to duplicate it by hand once, and found that
>they were already optimized for the under-16 players, and above 16 it
>becomes easy to do. It may have been done by hand originally for all I
>know -- the Archon shows some earmarks of exactly that.
It also occurs to me, after some thought, that 17- through 19- player
tournaments also would be impossible to optimize perfectly. Again
taking an example situation: a 17-player tournament.
Round 1: A-B-C-D-E, F-G-H-I, J-K-L-M, N-O-P-Q.
Attempting to seat round 2 fails again:
A-F-J-N, and you can't put anyone else on the 5-player table.
So I would modify my previous conclusion to say that in order to have
players never actually repeat sitting at the same table with each other,
you must have either 16 players exactly, or 20 or more players.
Also, given that optimization for that particular requirement is
impossible, I would expect that the Archon then endeavours to even out
the distribution of who sits at the 5-player tables as much as possible;
something which was slightly more important back when total VPs was the
determinant of who got to the finals, but still significant today.
Halcyan 2
>1. Ensure that noone has the same prey or predator as in a previous round.
>(this is covered today)
>2. Ensure that each players sees as many decks as possible, and sits next to
>as many new decks as possibly. (not covered today)
>I know of tournament
>organizers that don't yse the seating order as they think it's more optimal
>to draw the places for each round... (Hi Sten!)
In cases where there are 4 and 5-player tables, isn't there some sort of
optimality to balance things out (some one player doesn't sit at 4-player
tables all three rounds while a second player ends up at 5-player tables all
three rounds?).
At one of the recent tournaments I went to, it didn't actually become a
problem, but it was something I was concerned about. The organizer preferred
the random drawing of places for each round (there were three 5-player tables
and one 4-player table but only two rounds). So I thought it would be kinda
unfair if someone got stuck at the 4-player table both rounds (less possible
VP's) but fortunately that didn't happen (luck of the draw).
Halcyan 2
LSJ
> Also, given that optimization for that particular requirement is
> impossible, I would expect that the Archon then endeavours to even out
> the distribution of who sits at the 5-player tables as much as possible;
> something which was slightly more important back when total VPs was the
> determinant of who got to the finals, but still significant today.
Correct. In all cases, no one ever sits at a 4-player table more than
once more than anyone else.
Frederic Genest
> >
> > 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> > 5 Players
> > Round 1: 1 2 3 4 5
> > Round 2: 2 5 3 1 4
> > Round 3: 2 1 5 4 3
>
> To avoid repeating the same relative seating with another player.
> If you've got a seating chart in mind that is better, please share.
Mmmm. Wouldn't rotating play order for round 3 (keeping the same
seating) until player 4 begins the round simple to do and solve the
problem (i.e. Round 3: 4 3 2 1 5)?
(Maybe I'm just being intrusive; I'll return to lurk mode :-).
Frederic.
LSJ
>
> Frederic Genest wrote:
> >
> > > > To bring light to the discussion I will provide two examples:
> > > >
> > > > 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> > > > 5 Players
> > > > Round 1: 1 2 3 4 5
> > > > Round 2: 2 5 3 1 4
> > > > Round 3: 2 1 5 4 3
> > >
> > > To avoid repeating the same relative seating with another player.
> > > If you've got a seating chart in mind that is better, please share.
> >
> > Mmmm. Wouldn't rotating play order for round 3 (keeping the same
> > seating) until player 4 begins the round simple to do and solve the
> > problem (i.e. Round 3: 4 3 2 1 5)?
>
To help the "self-testing" procedure (before posting "better" 5-player
seating charts), check an alternative seating chart to see if it does
better in the following than the current chart:
1) #relative seating positions repeated: 0
2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
LSJ
>
> > > To bring light to the discussion I will provide two examples:
> > >
> > > 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> > > 5 Players
> > > Round 1: 1 2 3 4 5
> > > Round 2: 2 5 3 1 4
> > > Round 3: 2 1 5 4 3
> >
> > To avoid repeating the same relative seating with another player.
> > If you've got a seating chart in mind that is better, please share.
>
> Mmmm. Wouldn't rotating play order for round 3 (keeping the same
> seating) until player 4 begins the round simple to do and solve the
> problem (i.e. Round 3: 4 3 2 1 5)?
Then you've got player #1 seating in chair 4 for rounds 2 and 3.
--
Joshua Duffin
"Henrik Isaksson" <hen...@nissamedia.net> wrote in message
news:a73f5j$i0vgp$1...@ID-135400.news.dfncis.de...
> To bring light to the discussion I will provide two examples:
>
> 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> 5 Players
> Round 1: 1 2 3 4 5
> Round 2: 2 5 3 1 4
> Round 3: 2 1 5 4 3
btw, what LSJ hasn't spelled out for this question is that
"repeating going first" is considered exactly equivalent
to "repeating going second" in considering seating. If
player 2 didn't go first twice (eg 4 3 2 1 5 for round 3),
then someone else would repeat a play-order instead (eg
player 5 would play 5th twice *and* player 1 would play
4th twice - note that in the current one, player 4 is
already playing fourth twice; with only five players a
certain amount of play-order-repetition is inevitable).
(I've actually looked at these charts quite a bit,
especially for the smaller numbers of players, since we
use them sometimes at weekly play nights to try and cut
down next-to-each-otherness across games. There's not
much you can do to improve them, as far as I can tell.)
> 13 players (this was the number of players in the tournament)
> 13 Players
> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
> Round 3: 3 12 10 5, 6 4 13 11, 8 1 9 7 2
>
>
> Round 1 and 2 are not a problem to anyone. Noone sits next to anyone that
> they've sat at the same table as before, and the distribution seems even.
In
> round three though, things gets really messy. Notice players 7,9,1 at
table
> 1 in round 2. Exactly these players sit together at table 3 in round 3,
but
> with reversed order. Imagine the frustration of player 9 if he plays a
vote
> deck and players 7 and 1 plays intercept. Also, players 4 and 6 are in
> reversed order from round 2. To show that it's possible to make this
better
> I will provide am alternative round 3: With only a change of players 4 and
> 9, noone sits next to a player he's sat next to before.
> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
> Round 3: 3 12 10 5, 6 9 13 11, 8 1 4 7 2
I think you've missed, in round 3, that your change makes
player 9 sit next to 13 (which he did in round 1) and
player 4 sit next to player 1 (which he did in round 1).
This may be slightly better than having the 1&9, 9&7, 6&4
pairs in the original order, but not that much...
You may or may not be repeating "going first/second/etc"
more than in the original; it would take too long for
me to figure out. :-)
> These reveresed order of two or three players from round two to three is
> somehting that occurs very frequently in many charts, and it's my belief
> that they all aer unneccesary. It's very boring to sit next to a deck
you've
> sat next to before, be it as prey or predator. I know of tournament
> organizers that don't yse the seating order as they think it's more
optimal
> to draw the places for each round... (Hi Sten!)
Derek and LSJ are right: they can't be avoided, for most
tournament sizes. Randomizing seating each round (eg with
die-rolls or card-drawing) is unlikely to do better than
optimal-seating in avoiding "next-to" repeats, and will
definitely do worse at avoiding "predator/prey" repeats
(though you can always redraw when those come up). It's
also not guaranteed to equalize "sitting on the five-player
table" or play-order.
> Perhaps the "randomization" of this chart could be analyzed by some
students
> as some sort of exam-project?
Go for it. I hope you have students to assign this to,
cause I don't. ;-)
Josh
not a teacher
spinney99
> In message <a73f5j$i0vgp$1...@ID-135400.news.dfncis.de>,
> "Henrik Isaksson" <hen...@nissamedia.net> mumbled something about:
>
> >The goal of the Optimal Seating Order (OSO tm) should be:
> >1. Ensure that noone has the same prey or predator as in a previous round.
> >(this is covered today)
> >Am I correct in my assumption that the charts can be improved a lot?
>
> Nope.
>
> >Perhaps the "randomization" of this chart could be analyzed by some students
> >as some sort of exam-project?
here's a valid point. who exactly is player A, and who is player K?
In repeat tournaments with the same # of entrants, though the seating
chart may be as optimal as we can get it, Player A still sits in the
same 3-round seating rotation with respect to Player B, Player K, or
whoever. Is there a guideline covering _randomization_ of who becomes
player A on the seating chart? I hope it isn't first-to-sign-in is
player A. This doesn't achieve as much fairness as randomizing the
pool of entrants before assigning Chart position.
Worse, come to think of it, the Seating chart can become a Cheating
chart too easily, if it's not randomized. As (crooked) tourney
organizer, I want to be sure I have my best pal in round 3 with me so
we can help one or both of us achieve finals, after seeing our
relative positions. I become player A, and manipulate my pal to
become player F (eg) for a 3rd-rd rendezvous. Or I don't want to see
my pal, because we're both running !?malks! and don't want to contest.
I stay Player A, but move pally to Player H, so we never meet in
prelims. This Cheating chart scenario is possible even without a
repeat on # of entrants.
The best thing is that this whole (hypothetical) scenario is easily
prevented by randomizing the pool of entrants _before_ assigning Chart
position. Please regard this note as a safeguard suggestion, and
incorporate something to the floor rules that deals with these
possibilities, whether it's this or not. (Another positive
side-effect is that an organizer's responsibility for the mandatory
imperfections of the seating chart is completely alleviated.)
john sporemage
"A to the muthaf***in' K, homeboy. A to the muthaf***in' K."-Cypress
Hill, who also like the Assault Rifle.
(ps. if there already is such a safeguard in place, please point it
out to me, because i didn't find it.)
LSJ
>
> Derek Ray <lor...@yahoo.com> wrote in message news:<6uia9u0tfh6gts2vm...@4ax.com>...
> > In message <a73f5j$i0vgp$1...@ID-135400.news.dfncis.de>,
> > "Henrik Isaksson" <hen...@nissamedia.net> mumbled something about:
> >
> > >The goal of the Optimal Seating Order (OSO tm) should be:
> > >1. Ensure that noone has the same prey or predator as in a previous round.
> > >(this is covered today)
> (snip explanation, other q&a)
> > >Am I correct in my assumption that the charts can be improved a lot?
> >
> > Nope.
> >
> > >Perhaps the "randomization" of this chart could be analyzed by some students
> > >as some sort of exam-project?
> ===
> here's a valid point. who exactly is player A, and who is player K?
> In repeat tournaments with the same # of entrants, though the seating
> chart may be as optimal as we can get it, Player A still sits in the
> same 3-round seating rotation with respect to Player B, Player K, or
> whoever. Is there a guideline covering _randomization_ of who becomes
> player A on the seating chart?
Yes - assign the numbers (or, in this case, letters) randomly.
> I hope it isn't first-to-sign-in is
> player A. This doesn't achieve as much fairness as randomizing the
> pool of entrants before assigning Chart position.
No. It is non-random, and therfore would not conform to VEKN rules.
> [...]
> The best thing is that this whole (hypothetical) scenario is easily
> prevented by randomizing the pool of entrants _before_ assigning Chart
> position. Please regard this note as a safeguard suggestion, and
> incorporate something to the floor rules that deals with these
> possibilities, whether it's this or not. (Another positive
It's there already. [VEKN 3.1.2]
It's also stated explicitly in the Instructions sheet on the Archon
spreadsheet.
Joshua Duffin
"spinney99" <spore...@hotmail.com> wrote in message
news:27f52a52.02031...@posting.google.com...
> here's a valid point. who exactly is player A, and who is player K?
> In repeat tournaments with the same # of entrants, though the seating
> chart may be as optimal as we can get it, Player A still sits in the
> same 3-round seating rotation with respect to Player B, Player K, or
> whoever. Is there a guideline covering _randomization_ of who becomes
> player A on the seating chart?
V:EKN Tournament Rules 3.1.2: "The judge will randomly assign
players in groups of five. In the event that the number of
players is not evenly divisible by five, the judge will
assign players in groups of four and five, such that as many
groups as possible contain five players.
The judge will group players randomly, using prepared index
cards, a computerized system for generating random groupings,
or another randomizing method."
It should perhaps be spelled out in this section that the
Optimal Seating Charts aren't random (for the first round)
if they're "seeded" with some non-random ordering of players
(eg "in the order they signed up to play").
> I hope it isn't first-to-sign-in is
> player A.
Yeah, I've played in tournaments that did number players
based on "first to sign in" but it's not very cool. Even
if you do randomize, you're liable to play with the people
you came with at least once, but hopefully not all of them
at once. :-)
For this very reason, I always shuffle people's sign-in
index cards to randomize them before giving them "player
numbers" for the optimal seating charts. Alternatively
you can enter people's data (name/city/VEKN number) into
the computer as they sign in but wait to assign their
player numbers until everyone's arrived and they can be
randomized.
(Note that - and this is an objection I've actually heard
of, though not seen in action - if you know what everyone's
"player number" is, you can figure out after the first
round's seated who you'll be playing against in later
rounds. This can be avoided by not telling people what
their numbers are, if necessary...)
> This Cheating chart scenario is possible even without a
> repeat on # of entrants.
Yes, but that would be cheating. :-) Any tournament that
the organizer plays in must be run under the multi-judge
system, and allegations of cheating can be brought to the
appropriate judge (ie one who's not in the game in question).
If all the judges are in it together, then you're screwed,
but that's a whole other problem. (If all the judges are
crooked and conspiring, a non-playing organizer might well
want to manipulate the tournament too...)
Josh
v:ekn prince of washington dc and sometime tournament organizer
Noal McDonald
>>>>> Round 1: 1 2 3 4 5
>>>>> Round 2: 2 5 3 1 4
>>>>> Round 3: 2 1 5 4 3
>
> seating charts), check an alternative seating chart to see if it does
> better in the following than the current chart:
>
> 1) #relative seating positions repeated: 0
> 2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
Okay. Seems easy enough:
Round 1: 1 2 3 4 5
Round 2: 3 5 4 2 1
Round 3: 2 4 1 5 3
1) #relative seating positions repeated: 0
2) #players who sit in same seat twice: 0
The only problem is starting position is skewed slightly in player 5's
favor. Given the small number of participants, this can't be helped.
However, the parity is much better than the currently official seating
chart. See the tables below that track starting influences:
Current:
1) 1+4+2=7
2) 2+1+1=4
3) 3+3+4=10
4) 4+4+4=12
5) 4+2+3=9
Proposed:
1) 1+4+3=8
2) 2+4+1=7
3) 3+1+4=8
4) 4+3+2=9
5) 4+2+4=10
I should point out that, since I don't own a laptop, I've never used
the Archon and I've always down seating charts by hand after each
round of a tournament. My priorities have always been, in order:
1) Not sitting next to the same person twice
1a) Not sitting in the same relative position to someone twice (when
#1 can't be avoided)
2) Not sitting in the same position twice
3) Mathematical equality of starting transfers (as close as possible)
Yeah, doing a seating chart by hand does make me more frazzled, I find
the practice helps me adjust gracefully when people drop out of a
tournament. As always, YMMV.
Regards,
Noal McDonald
VEKN Prince of Metro Detroit
Noal McDonald
> (I've actually looked at these charts quite a bit,
> especially for the smaller numbers of players, since we
> use them sometimes at weekly play nights to try and cut
> down next-to-each-otherness across games. There's not
> much you can do to improve them, as far as I can tell.)
ooh. I love a challenge. The 5 player chart was easy to improve. (see
other post in this thread) I'll have to look at the others to see if I
can provide better numbers there. Btw, I narrowly avoided being a math
major, but decided I wasn't _quite_ that anal. ;-)
Regards,
Noal
Joshua Duffin
"Noal McDonald" <dhar...@my-deja.com> wrote in message
news:bb705c59.02031...@posting.google.com...
In a lengthy email, LSJ convinced me that positions 1-4
are all equally good. Position 5 is the only one that's
demonstrably worse. Going first more than once is not
actually a disadvantage; you get an extra master phase
and discard, and you get to 5 transfers before anyone
else does.
(For certain decks, certain seats are better than others,
but as a general rule, going first more than once is
probably better than not. Imagine going fifth compared
to going first: you have to go after player 4 but don't
get any more transfers. Or the other way around, going
first is like going fifth, but with an extra transfer,
master phase, and discard.)
That said, I more than welcome any effort you want to
put into improving the seating charts - your revised
5-player is definitely better, as far as I can tell. :-)
Josh
loves being challenging
Noal McDonald
>> 13 Players
>> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
>> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
>> Round 3: 3 12 10 5, 6 4 13 11, 8 1 9 7 2
While it may not be a criteria, equality of beginning transfers should
be. This chart isn't as unfair as the 5 player chart, it could be
better.
The above seating chart results in the following list of beginning
transfers: (notice players 1&9 having 6 xfers in the opening round,
while 2&10 get 10)
1) 1+3+2=6
2) 2+4+4=10
3) 3+4+1=8
4) 4+2+3=9
5) 1+4+2=7
6) 2+4+1=7
7) 3+1+4=8
8) 4+2+1=7
9) 1+2+3=6
10) 2+4+4=10
11) 3+1+4=8
12) 4+3+2=9
13) 4+1+3=8
I would propose the following chart, which adheres to the criteria of
no repeat relative positions and no repeat numerical position, as well
as doing a better job of equalizing opening xfers: (it may be possible
to reduce the number of times someone sits next to another player
twice, but I'll leave that to someone else)
Round 1) 1 2 3 4, 5 6 7 8, 9 10 11 12 13
Round 2) 13 3 10 5, 12 11 6 9, 8 4 2 7 1
Round 3) 7 13 1 10, 4 9 8 11, 6 12 5 3 2
This produces the following number of opening xfers:
1) 1+4+2=7
2) 2+3+4=9
3) 3+2+4=9
4) 4+2+1=7
5) 1+4+3=8
6) 3+4+1=8
7) 3+4+1=8
8) 4+1+3=8
9) 1+4+2=7
10) 2+3+4=9
11) 3+2+4=9
12) 4+1+2=7
13) 4+1+2=7
I'm sure this can be improved further, but it is better than the one
originally listed.
The Nosferatu Stuff
"LSJ" <vte...@white-wolf.com> wrote in message
news:3C95E89B...@white-wolf.com...
> LSJ wrote:
> >
> > Frederic Genest wrote:
> > >
> > > > > To bring light to the discussion I will provide two examples:
> > > > >
> > > > > 5 players. Here player 2 is supposed to begin round 2 and 3. Why?
> > > > > 5 Players
> > > > > Round 1: 1 2 3 4 5
> > > > > Round 2: 2 5 3 1 4
> > > > > Round 3: 2 1 5 4 3
> > > >
> > > > To avoid repeating the same relative seating with another player.
> > > > If you've got a seating chart in mind that is better, please share.
> > >
> > > Mmmm. Wouldn't rotating play order for round 3 (keeping the same
> > > seating) until player 4 begins the round simple to do and solve the
> > > problem (i.e. Round 3: 4 3 2 1 5)?
> >
> > Then you've got player #1 seating in chair 4 for rounds 2 and 3.
>
> To help the "self-testing" procedure (before posting "better" 5-player
> seating charts), check an alternative seating chart to see if it does
> better in the following than the current chart:
>
> 1) #relative seating positions repeated: 0
> 2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
Isn't an obviously overlooked idea to make 2 round tournaments? When
playing with relativly small numbers of players, playing too many rounds
seems to make it more difficult to work out fair seating. Why not just drop
round 3 on tournaments that can't get 20 people(as Derek pointed out, the
number where there is perfect optimization)
--
Aaron
The Nosferatu Stuff
Noal McDonald
> In a lengthy email, LSJ convinced me that positions 1-4
> are all equally good.
Hrm. I'd be interested to see that e-mail. Additional insight into game
mechanics is always a good thing. :-)
> Position 5 is the only one that's demonstrably worse.
Which builds a case that it's fair that player 5 gets more xfers than
anyone else in the opening round in all 3 games combined. Knowing that,
I'd try to put player 5 into the first slot in the second round, but
that would require me to completely rework the chart.
> Going first more than once is not actually a disadvantage;
> you get an extra master phase and discard, and you get to
> 5 transfers before anyone else does.
Hrm. Not sure about the intrinsic values there.
The discard *may* be an advatage, but I doubt it's a statistically
significant one. Focussed decks should be have a lot of redundancy,
meaning discarding a card is likely to get it replaced with the same
card or something else you already have in your hand. Toolbox decks will
have a number of cards available to counter other strategies which could
be discarded. However, it's not likely that you're going to know which
ones to ditch and which ones to keep on the opening draw.
The master card phase can be very useful, I'll grant. Anarch Revolt,
Dreams, Tomb, Info Highway, Jake Washington are all good cards to play
in the opening round for player 1. However, most master cards require a
minion of some kind to play, making the phase useless until you get a
vampire out.
The 5th xfer may be useful...unless of course, your vampires are all
even numbered capacity. :-P
> (For certain decks, certain seats are better than others,
Which, as shown above, is certainly the case. Evening out the starting
xfers certainly ensures that one deck isn't significantly helped, or
hindered, more than another. If you can limit decks to 7-9 starting
xfers after 3 rounds, rather than 6-10, I'd say that's probably a good
thing.
You'll also note that the opening xfers for each player after 2 rounds
is about 5-6. That way, if people drop out after the 2nd round, parity
has already been achieved.
> That said, I more than welcome any effort you want to
> put into improving the seating charts - your revised
> 5-player is definitely better, as far as I can tell. :-)
Thanks. :-) Maybe I will spend some time going through them. The 5- and
13-player tables didn't take much time since I just used the system that
I've been using for years to determine seating at tournaments.
Regards,
Noal
lehrbuch
> seating charts), check an alternative seating chart to see if it does
> better in the following than the current chart:
>
> 1) #relative seating positions repeated: 0
> 2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
How about:
Round 1: 1 2 3 4 5
Round 2: 3 5 4 2 1
Round 3: 2 4 1 5 3
I think it is 0 and 0 for both criteria.
* lehrbuch
lehrbuch
No. I tell a lie. Some grand-prey relationships are doubled up. I
wasn't thinking of those.
I guess you are correct in assuming that grand-prey relationships are
more important than start-up positions.
* lehrbuch
LSJ
>
> LSJ <vte...@white-wolf.com> wrote:
> >>>>> Round 1: 1 2 3 4 5
> >>>>> Round 2: 2 5 3 1 4
> >>>>> Round 3: 2 1 5 4 3
> >
> > To help the "self-testing" procedure (before posting "better" 5-player
> > seating charts), check an alternative seating chart to see if it does
> > better in the following than the current chart:
> >
> > 1) #relative seating positions repeated: 0
> > 2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
>
> Okay. Seems easy enough:
>
> Round 1: 1 2 3 4 5
> Round 2: 3 5 4 2 1
> Round 3: 2 4 1 5 3
>
> 1) #relative seating positions repeated: 0
Actually, that's 3
#1 has #3 as his grand-prey twice.
#3 has #4 as his grand-prey twice.
#5 has #2 as his grand-prey twice.
> 2) #players who sit in same seat twice: 0
> I should point out that, since I don't own a laptop, I've never used
> the Archon and I've always down seating charts by hand after each
> round of a tournament. My priorities have always been, in order:
>
> 1) Not sitting next to the same person twice
> 1a) Not sitting in the same relative position to someone twice (when
> #1 can't be avoided)
> 2) Not sitting in the same position twice
> 3) Mathematical equality of starting transfers (as close as possible)
This is balanced by the game - no need to try to balance it again
here, except for the 5th seat.
LSJ
It has 3 and 0, respectively. See my response to Noal's post with the
same seating.
LSJ
>
> LSJ <vte...@white-wolf.com> wrote
> >> 13 Players
> >> Round 1: 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> >> Round 2: 7 9 1 10, 13 8 12 2, 11 5 4 6 3
> >> Round 3: 3 12 10 5, 6 4 13 11, 8 1 9 7 2
>
> While it may not be a criteria, equality of beginning transfers should
> be.
Any basis for such an assertion?
The ramping system is added to the game as a balancing factor.
No need to re-balance it.
> I would propose the following chart, which adheres to the criteria of
> no repeat relative positions and no repeat numerical position, as well
> as doing a better job of equalizing opening xfers: (it may be possible
> to reduce the number of times someone sits next to another player
> twice, but I'll leave that to someone else)
>
> Round 1) 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> Round 2) 13 3 10 5, 12 11 6 9, 8 4 2 7 1
> Round 3) 7 13 1 10, 4 9 8 11, 6 12 5 3 2
Less optimal than current:
This results in 9 and 11 facing each other in all three rounds.
Also 10 and 13 face each other thrice.
Current: no players thrice.
Chris Berger
> Noal McDonald wrote:
> >
> > LSJ <vte...@white-wolf.com> wrote:
> > >>>>> Round 1: 1 2 3 4 5
> > >>>>> Round 2: 2 5 3 1 4
> > >>>>> Round 3: 2 1 5 4 3
> > >
> > > To help the "self-testing" procedure (before posting "better" 5-player
> > > seating charts), check an alternative seating chart to see if it does
> > > better in the following than the current chart:
> > >
> > > 1) #relative seating positions repeated: 0
> > > 2) #players who sit in same seat twice: 3 (#2-s1, #3-s3, #4-s4)
> >
> > Okay. Seems easy enough:
> >
> > Round 1: 1 2 3 4 5
> > Round 2: 3 5 4 2 1
> > Round 3: 2 4 1 5 3
> >
> > 1) #relative seating positions repeated: 0
>
> Actually, that's 3
> #1 has #3 as his grand-prey twice.
> #3 has #4 as his grand-prey twice.
> #5 has #2 as his grand-prey twice.
>
for optimality. However, when applying more rigorous constraints that actually
take into account the game that's going to be played with the seating charts,
Noal's chart seems better (at least, without studying it for *too* long). No
player has the same two players on either side of them in two different games (a
huge concern among most players), no predator/prey relationships are repeated
(much more important than grand-predator/grand-prey relationships), and
first-round transfers are evened out. I'm not sure if I believe that each
position is demonstrably balanced against all others, but even if one takes that
as a given, certain decks are stronger from certain positions, so, if possible,
care should be taken to even out the benefit that could be gained by a lucky
seating placement (of, say, a weenie deck sitting 2nd all the time, or a
Zillah's Valley-heavy deck sitting 4th).
As an aside, I don't think you could convince me that seat 5 is the weakest. If
all other seats were even, then seat 5 would be weakest, however player 1 is
extremely unlikely to have a minion first turn, and even somewhat unlikely to
get one second turn, which leaves 5 in something of the catbird seat, as he has
a good chance to beat up on his lagging prey. Sure, player 1 gets 5 transfers
first, but player 4 gets 8 transfers first. No big difference, except of course
that a lot of the new 8 capacity vampires are very good indeed.
Derek Ray
"Chris Berger" <ark...@ugcs.caltech.edu> mumbled something about:
>As an aside, I don't think you could convince me that seat 5 is the weakest. If
Seat 5 sucks, especially when Seat 4 is a strong deck. You get no speed
advantage, and player 1 is always one turn ahead of you.
>all other seats were even, then seat 5 would be weakest, however player 1 is
>extremely unlikely to have a minion first turn, and even somewhat unlikely to
>get one second turn, which leaves 5 in something of the catbird seat, as he has
??! You play THAT many decks without 5-caps or cheaper?
>a good chance to beat up on his lagging prey. Sure, player 1 gets 5 transfers
??!
Turn 1, I put one on vampire "X", who is probably a 5-cap or cheaper.
Player 5 brings out Gloria Giovanni or Ingrid Russo.
Turn 2, I mumble something about "stupid weenie Dominate" and get ANY
vampire I can post-haste, even if it means NOT influencing to my 9-cap
Quentin and instead getting my 3-cap Juan Cali, who I was going to get
later along with my 5-cap Vincent Day ... and who I was hoping to Govern
out with Quentin in the first place. But with DOM behind me, it's
probably a good time for Plan B.
The only way I'm screwed here is if I'm playing a deck that only has
6-caps and up... which is pretty rare for most decks, and if I only have
6-caps or higher, then I KNOW I'm likely to be screwed by someone
getting out weenies behind me, and have planned for it.
>first, but player 4 gets 8 transfers first. No big difference, except of course
>that a lot of the new 8 capacity vampires are very good indeed.
Which can pretty much bufu player 5. Player 1 doesn't care (if there's
no player 5); he gets, say, Quentin and makes "thpthfbfbft" noises at
Anson, Anson's mom, and Anson's Awe.
Chris Berger
> In message <yOxl8.66859$ZR2....@rwcrnsc52.ops.asp.att.net>,
> "Chris Berger" <ark...@ugcs.caltech.edu> mumbled something about:
>
> >As an aside, I don't think you could convince me that seat 5 is the weakest.
If
>
> Seat 5 sucks, especially when Seat 4 is a strong deck. You get no speed
> advantage, and player 1 is always one turn ahead of you.
>
have a vampire before player 1 in nearly every circumstance, and that's what's
important.
> >all other seats were even, then seat 5 would be weakest, however player 1 is
> >extremely unlikely to have a minion first turn, and even somewhat unlikely to
> >get one second turn, which leaves 5 in something of the catbird seat, as he
has
>
> ??! You play THAT many decks without 5-caps or cheaper?
>
No. But generally 1/3rd of my decks are based on larger than 5 cap vampires.
Which usually means no more than 3 vamps are 5 cap or cheaper (my Lasombra deck
threw in Cameron recently, giving it a 2, 4, 4, 5, but is still very close to
the line where you can/can't expect to have a vamp 2nd turn in the 1st seat).
Add in those bad draws for spread decks (about another 1/3rd of my decks), which
have 1/2 vamps 5 cap or cheaper and the other half bigger than that, but for
some reason insist on drawing 4 of the 6 "Thumbs" when I'm in a bad position.
It seems to me like (pulling numbers out of my ass here) approximately a 5%
chance that a given deck will get a vampire first turn in the 1 hole, and maybe
a 70% chance of having a vampire by second turn (depending on the deck). And
30% is what I qualify as a "decent" chance, i.e. something that happens just
often enough to screw ya.
> >a good chance to beat up on his lagging prey. Sure, player 1 gets 5
transfers
>
> ??!
>
> Turn 1, I put one on vampire "X", who is probably a 5-cap or cheaper.
> Player 5 brings out Gloria Giovanni or Ingrid Russo.
>
> Turn 2, I mumble something about "stupid weenie Dominate" and get ANY
> vampire I can post-haste, even if it means NOT influencing to my 9-cap
> Quentin and instead getting my 3-cap Juan Cali, who I was going to get
> later along with my 5-cap Vincent Day ... and who I was hoping to Govern
> out with Quentin in the first place. But with DOM behind me, it's
> probably a good time for Plan B.
>
You get your one vamp out that gets Seduced. Yay! In 2nd-4th position, you can
have two vampires the turn after your predator brings out Gloria or Ingrid, or
else one big one, and make them pay for their insolence.
> >first, but player 4 gets 8 transfers first. No big difference, except of
course
> >that a lot of the new 8 capacity vampires are very good indeed.
>
> Which can pretty much bufu player 5. Player 1 doesn't care (if there's
> no player 5); he gets, say, Quentin and makes "thpthfbfbft" noises at
> Anson, Anson's mom, and Anson's Awe.
>
Hehe. Gotta love Quentin. Still, the first person to get 8 transfers has, IMO,
a decided advantage. Similarly, the first person to get 4 transfers, although
that's, obviously, the same person. It's a small consolation that the first
person to 5 transfers can get Vittorio out before anyone else.
Noal McDonald
>>1) #relative seating positions repeated: 0
> Actually, that's 3
> #1 has #3 as his grand-prey twice.
> #3 has #4 as his grand-prey twice.
> #5 has #2 as his grand-prey twice.
Ah. Didn't know you were drilling down that far. I'll draft up another
proposal after when I'm done playing a schizo bad-cop, good-cop in the
Unix infrastructure at work tomorrow by heavy-handed, draconian
enforcement of data security policies and helpful troubleshooting of
issues that are over the heads of the on-site Unix admins.
"You changed your root account password in a unapproved manner. It is
now revoked until you fill out this form and wait a month for management
approval. Now about that workstation that can't connect to the
server..." I love my job. :-D
> This is balanced by the game - no need to try to balance it again
> here, except for the 5th seat.
We'll see...
Regards,
Noal
Noal McDonald
>>While it may not be a criteria, equality of beginning transfers should
>>be.
>
> Any basis for such an assertion?
Sure. When possible, it is a criteria to minimize the instances that a
player occupies the same seat. That goal takes it one step further.
Some decks benefit, or suffer, more from certain positions at the table
than others. After ensuring that players don't occupy the same seat
twice, I try to load balance it further to minimize the impact seating
has on the success on a deck. As a supporting theory, while
mathematically no seat (except 5th) may be worse than another, player
impressions may be otherwise. As you know, psychological effects can
certainly have an effect on a game.
> The ramping system is added to the game as a balancing factor.
How so. Would you care to explain this to me?
>>Round 1) 1 2 3 4, 5 6 7 8, 9 10 11 12 13
>>Round 2) 13 3 10 5, 12 11 6 9, 8 4 2 7 1
>>Round 3) 7 13 1 10, 4 9 8 11, 6 12 5 3 2
> This results in 9 and 11 facing each other in all three rounds.
> Also 10 and 13 face each other thrice.
oops. Switch 13 and 9 in the 3rd round and that solves that.
Regards,
Noal
Xian
"Chris Berger" <ark...@ugcs.caltech.edu> wrote in message
news:bRyl8.72416$uA5....@rwcrnsc51.ops.asp.att.net...
> Player one is always screwed by you. He's a turn *behind*, not ahead.
You will
> have a vampire before player 1 in nearly every circumstance, and that's
what's
> important.
I don't mind going first. Assuming you have a 5-cap or less in your crypt,
you'll be able to get a vampire out at approximately the time you'll need
it. Just before your prey gets his first action (if any). Not to mention,
that first master phase *can* be useful. First turn Dreams of the Sphinx?
Yes, please!
> Hehe. Gotta love Quentin. Still, the first person to get 8 transfers
has, IMO,
> a decided advantage. Similarly, the first person to get 4 transfers,
although
> that's, obviously, the same person. It's a small consolation that the
first
Barring 2nd seat playing Info Highway? :)
Xian
Derek Ray
>"Derek Ray" <lor...@yahoo.com> wrote in message
>news:02cd9u0ibs51b3mta...@4ax.com...
>
>> Seat 5 sucks, especially when Seat 4 is a strong deck. You get no speed
>> advantage, and player 1 is always one turn ahead of you.
>>
>Player one is always screwed by you. He's a turn *behind*, not ahead. You will
>have a vampire before player 1 in nearly every circumstance, and that's what's
>important.
But you won't ever be able to use that vampire before player 1 gets his
own out, meaning that player 1 will almost always have a defense
available against you. He gets 5-caps before you can ever take an
action, and 9-caps before you can ever take an action with a Big Fucker.
AND, player 1 can get an Info Highway, Dreams, Zillah's, Creepshow,
Hungry Coyote, or any other hideously-convenient-first-turn-master out
first, meaning the rest of the table has to contest or discard --
something that only player 5 is in a position to do. But he always gets
to use it first -- and he gets one free discard before anyone gets to do
anything, something which we all established long ago is an underrated
resource.
>> ??! You play THAT many decks without 5-caps or cheaper?
>>
>No. But generally 1/3rd of my decks are based on larger than 5 cap vampires.
So twice as often as not, this whole thing won't matter. =)
>Which usually means no more than 3 vamps are 5 cap or cheaper (my Lasombra deck
>threw in Cameron recently, giving it a 2, 4, 4, 5, but is still very close to
>the line where you can/can't expect to have a vamp 2nd turn in the 1st seat).
4 vampires of any given type is 86%, which is, in practice, a
near-certainty. 3 is 75%, which is still pretty damned likely. The
other-thread mentioned Omaya/Gunther deck with 3 of each drew Omaya
every round, and Gunther as well in 2 of those rounds.
Technically, 75% or above chance when gambling is the point where you're
supposed to open your wallet, dump it on the table, and start betting
huge chunks, because your wins don't just make up for your losses, they
show a ridiculous profit besides. It can be closely related to stuff
like this in V:TES (4 Wolf Valentine in a deck, for example, mean that I
can count on having him available, and therefore all my !Salubri cards
are playable).
>Add in those bad draws for spread decks (about another 1/3rd of my decks), which
>have 1/2 vamps 5 cap or cheaper and the other half bigger than that, but for
>some reason insist on drawing 4 of the 6 "Thumbs" when I'm in a bad position.
3% chance of not getting the cheap vamps. Statistical anomaly? =)
>It seems to me like (pulling numbers out of my ass here) approximately a 5%
>chance that a given deck will get a vampire first turn in the 1 hole, and maybe
This depends entirely on how many decks total include 1-cap vampires.
The number for any given tourney can be found by .2C*(V/T), where V is
the total number of decks in the tourney which contain a 1-cap, T is the
total number of decks in the tourney, .2 is the chance of sitting first,
and C is the average chance of the "V" decks getting a 1-cap in their
opening crypt (probably no more than 50% or .5 since you don't often put
more than 2 1-caps in any deck). I think all that math comes out to a
number markedly resembling "diddly/squat" since once you take .25 of a
number that's going to be fairly small anyway (V/T),... you get the
picture.
>a 70% chance of having a vampire by second turn (depending on the deck). And
(shrug) Using your own numbers, 66% of your decks almost certainly have
a 5-cap available to be "got" second turn -- let's assume five 5-caps or
less, so 93% of that is 61%. The remaining 33% can be touted at 60%
(although 3-in-a-crypt is 75%, we'll slide away from it a bit to allow
for the "all-8-caps-and-up" decks), getting 20%.
Stir these numbers together slightly, and you get a 31% chance spread
across ALL your decks of not having a vamp second turn... if you're in
the 1-hole, which only happens 20% of the time, so we're back down to
6.2%, or slightly more than diddly/squat.
>30% is what I qualify as a "decent" chance, i.e. something that happens just
>often enough to screw ya.
Not quite to the point where you should open your wallet and dump it on
the table, but definitely well past the point where you should say
things like "let's double the bet... just to make it interesting."
>You get your one vamp out that gets Seduced. Yay! In 2nd-4th position, you can
>have two vampires the turn after your predator brings out Gloria or Ingrid, or
>else one big one, and make them pay for their insolence.
In 2nd slot, I need a 2-cap and 4-cap, or two 3-caps. Rare.
In 3rd slot, I need a 3-cap and 4-cap, or a 2-cap and 5-cap. Better,
but still not THAT common. In 3rd slot, I also generally go for my
7-cap as best bang-for-the-buck.
In 4th slot, I have almost certainly just gone ahead and gotten one big
vamp, who gets Seduced anyway in this scenario. Boooo.
In 5th slot, I need a 4-cap or less, not a 5-cap. Ouch! Screwed!
>> Which can pretty much bufu player 5. Player 1 doesn't care (if there's
>> no player 5); he gets, say, Quentin and makes "thpthfbfbft" noises at
>> Anson, Anson's mom, and Anson's Awe.
>>
>Hehe. Gotta love Quentin. Still, the first person to get 8 transfers has, IMO,
I'm on a big Quentin kick recently. I was on one a few years ago, and
then Sabbat War came out and I got on one again, and I just got on
another one recently. He's just got such a big "Fuck You, Arika" sign
on his forehead for opposing vote decks, especially after Demonstration
and/or your Hunting Ground hits the table.
Now, this latest kick could have been provoked by Meddling of Semsith,
but I'll never tell... although I did play a !Ventrue bleed deck for the
Storyline tournament with Pulling Strings as vote defense =)
>a decided advantage. Similarly, the first person to get 4 transfers, although
>that's, obviously, the same person. It's a small consolation that the first
>person to 5 transfers can get Vittorio out before anyone else.
The vampire isn't the big deal. It's the extra discard and Master phase
that can really matter -- and also having a predator who's always one
action behind you.
What really screws the poor bugger in the 1-hole is when his whole deck
is based around 6-through-8 caps, and getting one of those first.
Ventrue Law Firms do notoriously poorly in the 1-hole (three turns to
Bridges, three turns to Nash, don't want to get Ranjan first because
he's really only there to be Governed out, two turns to Gideon who is
supposed to be free or out first turn... bah), as do Harbingers decks
(three turns to Agaitas or Egothha).
Chris Berger
> In message <bRyl8.72416$uA5....@rwcrnsc51.ops.asp.att.net>,
> "Chris Berger" <ark...@ugcs.caltech.edu> mumbled something about:
>
which
> >have 1/2 vamps 5 cap or cheaper and the other half bigger than that, but for
> >some reason insist on drawing 4 of the 6 "Thumbs" when I'm in a bad position.
>
> 3% chance of not getting the cheap vamps. Statistical anomaly? =)
>
Assamite/Presence deck, I am 50% likely to draw all 4 and 50% likely to draw
none of them. =P
Okay, okay, so maybe it's not entirely true. But 4 of a vamp in a deck means
that there's a very good chance that at least once during a tournament you will
not draw the vampire. 4 small vampires means a very good chance that you'll
have no small vampire to bring out during at least one of your three rounds. I
could, of course, run the numbers to try and back up my point. But I just so
don't feel like it. I've been doing pathfinding algorithms all day, and I'm
just about fed up with any kind of math at all.
>
> >a 70% chance of having a vampire by second turn (depending on the deck). And
>
> (shrug) Using your own numbers, 66% of your decks almost certainly have
> a 5-cap available to be "got" second turn -- let's assume five 5-caps or
> less, so 93% of that is 61%. The remaining 33% can be touted at 60%
> (although 3-in-a-crypt is 75%, we'll slide away from it a bit to allow
> for the "all-8-caps-and-up" decks), getting 20%.
>
> Stir these numbers together slightly, and you get a 31% chance spread
> across ALL your decks of not having a vamp second turn... if you're in
> the 1-hole, which only happens 20% of the time, so we're back down to
> 6.2%, or slightly more than diddly/squat.
>
But wait.... you can't assume that you'll only be in the 1-hole 20% of the
time, because LSJ's seating algorithms don't take into account total 1st-turn
transfers. If you happen to play a deck that will always get screwed in the
1-hole, then one random chance labelling you as player #1 could screw you over
good. If this was unavoidable, then you just write it off to bad luck and be
done with it, but Noal's seating arrangement avoids it just fine, without being
any *worse* in LSJ's stated criteria.
> In 2nd slot, I need a 2-cap and 4-cap, or two 3-caps. Rare.
>
> In 3rd slot, I need a 3-cap and 4-cap, or a 2-cap and 5-cap. Better,
> but still not THAT common. In 3rd slot, I also generally go for my
> 7-cap as best bang-for-the-buck.
>
True. I guess I'll concede that point.
> >a decided advantage. Similarly, the first person to get 4 transfers,
although
> >that's, obviously, the same person. It's a small consolation that the first
> >person to 5 transfers can get Vittorio out before anyone else.
>
> The vampire isn't the big deal. It's the extra discard and Master phase
> that can really matter -- and also having a predator who's always one
> action behind you.
>
I guess that's where we disagree. You've got one extra discard on me when it's
my turn, so it's true that you're one card ahead in your deck. but that's a
small advantage. The extra Master phase helps with master-heavy decks that can
be sure of getting a non-clan master card on their first turn. There are
certainly some good ones, but when I have 3 Info Highways or 2 Dreams of the
Sphinx in a deck, I'm not really expecting to see 'em first turn. I *am*
expecting to see a vampire first turn if I'm number 5, or a big fucker second
turn. You can be ahead of me on master phases, if I can be ahead of you on
actions.
> What really screws the poor bugger in the 1-hole is when his whole deck
> is based around 6-through-8 caps, and getting one of those first.
> Ventrue Law Firms do notoriously poorly in the 1-hole (three turns to
> Bridges, three turns to Nash, don't want to get Ranjan first because
> he's really only there to be Governed out, two turns to Gideon who is
> supposed to be free or out first turn... bah), as do Harbingers decks
> (three turns to Agaitas or Egothha).
>
So, well... that's really all I'm saying, is that position is a huge deal to
many decks. Personally, I feel that the 1-hole is a bad deal (dare I say,
unfair - no, I dare not, but I wanted to label the post "MOMMY") for anyone but
the weeniest of weenies, but I can agree to disagree on that point if necessary.
But to completely disregard 1st-turn transfers as a seating criteria is
something which I think should be avoided, even at the cost of repeating
grand-prey/grand-predator relationships.
BTW, what do [BW] and [JC] stand for? Blood Wars and Jesus Christ???
spinney99
> spinney99 wrote:
(snip scenario by me and rational answers from lsj)
The best thing is that this whole (hypothetical) scenario is easily
> > prevented by randomizing the pool of entrants _before_ assigning Chart
> > position. Please regard this note as a safeguard suggestion, and
> > incorporate something to the floor rules that deals with these
> > possibilities, whether it's this or not.
>
> It's also stated explicitly in the Instructions sheet on the Archon
> spreadsheet.
It's re-assuring to know this is already dealt with in the official
rules. "I love this game:)" Though not a newbie to jyhad/vtes, I'm a
Slowbie on the net, not to mention _reading_ some days.
spinn sporb
"stupid is as stupid does..."-gump's mama
LSJ
>
> LSJ wrote:
>
> >>While it may not be a criteria, equality of beginning transfers should
> >>be.
> >
> > Any basis for such an assertion?
>
> Sure. When possible, it is a criteria to minimize the instances that a
> player occupies the same seat. That goal takes it one step further.
>
> Some decks benefit, or suffer, more from certain positions at the table
> than others. After ensuring that players don't occupy the same seat
> twice, I try to load balance it further to minimize the impact seating
> has on the success on a deck. As a supporting theory, while
> mathematically no seat (except 5th) may be worse than another, player
> impressions may be otherwise. As you know, psychological effects can
> certainly have an effect on a game.
>
> > The ramping system is added to the game as a balancing factor.
>
> How so. Would you care to explain this to me?
?
If no ramping system (player 1 gets 4 transfers on her first turn), then
player 1 has a distinct advantage.
> >>Round 1) 1 2 3 4, 5 6 7 8, 9 10 11 12 13
> >>Round 2) 13 3 10 5, 12 11 6 9, 8 4 2 7 1
> >>Round 3) 7 13 1 10, 4 9 8 11, 6 12 5 3 2
> > This results in 9 and 11 facing each other in all three rounds.
> > Also 10 and 13 face each other thrice.
>
> oops. Switch 13 and 9 in the 3rd round and that solves that.
Round 1) 1 2 3 4, 5 6 7 8, 9 10 11 12 13
Round 2) 13 3 10 5, 12 11 6 9, 8 4 2 7 1
Round 3) 7 9 1 10, 4 13 8 11, 6 12 5 3 2
OK. That looks no less optimal than the current.
Certainly evenening out "initial transfers" cannot hurt.
I'll update the 13-player chart in the next release of the Archon.
Thanks
LSJ
> Okay, so it's "as optimal" as your chart when you apply your loose constraits
> for optimality. However, when applying more rigorous constraints that actually
> take into account the game that's going to be played with the seating charts,
> Noal's chart seems better (at least, without studying it for *too* long). No
Current so-called "loose" goals, from the ancient (defunct) web page from the
days when the seating charts were first introduced (when the DCI was the
governing body for VTES tournaments), derived from lengthy study:
Tournament seating in V:TES is a bit more involved than "normal"
(two-player game) tournaments. The DCI rules include guidelines for
seating players: randomly, such that no two players are assigned to share
a table more than once, such that no two players have the same
predator-prey relationship to each other if they do share a table more
than once (sections 2.3.1 and 2.3.2).
For determining "optimal" arrangements, however, the following goals (a
superset of the DCI guidelines) are used:
1. No pair of players repeat their predator-prey relationship. This is
mandatory, by the DCI rules.
2. No pair of players share a table through all three rounds. An extremal
version of an option from the DCI rules.
3. Available VPs are equitably distributed. That is, if one player plays
at three five-player tables during the tournament, she'll have 15
"available" VPs. In this case, no other player should have less than
14 "available" VPs (that is, the difference in available VPs should be
no more than one and should be zero if possible).
4. No pair of players share a table more often than necessary. This is an
option from the DCI rules.
5. A player doesn't play in the same position. If a player plays second
at one table, she shouldn't be seated "second" in any other round.
If there are any mistakes in the arrangements listed (failing to meet one
of the above goal when it could have been met), please send me some
e-mail.
As an addendum to #5, when a player must repeat a seating position, repeating
seats 1-4 is preferable to repeating seat 5.
It shouldn't matter, but since it also won't hurt, a 7th criterion [see
my comments below for the 6th criterion] can be added (one that the current
charts do not use, but which could be used to illustrate that a new alternative
seating chart is more optimal than the current ones so long as the other 5
criteria are not sacrificed) - Noal's criterion:
6. Starting transfers are equitably distributed. [NOAL]
> player has the same two players on either side of them in two different games (a
> huge concern among most players),
"most"?
I wasn't aware of any such data. If you're speaking of "most players
you know" or under the usual (and usually justified) opinion that
"most players think as I do", then please say so. If you actually
mean "most", please cite your source(s). It's not that big of a
concern among the players I've judged (with the understanding that a
"huge concern" would be vocalized when the situation occurs).
At any rate, whenever possible, an unspoken extension to 4 was used
after 5 (the sixth criterion) to do exactly that. This criterion could
be moved to be after the 7th criterion above, depending on tastes.
> no predator/prey relationships are repeated
> (much more important than grand-predator/grand-prey relationships), and
Yes. The current charts follow this. Indeed, it is mandatory (save
when not possible, as in a 4-player 3+F tournament), by the VEKN rules.
[VEKN 3.1.2]
> first-round transfers are evened out. I'm not sure if I believe that each
> position is demonstrably balanced against all others, but even if one takes that
> as a given, certain decks are stronger from certain positions, so, if possible,
> care should be taken to even out the benefit that could be gained by a lucky
> seating placement (of, say, a weenie deck sitting 2nd all the time, or a
> Zillah's Valley-heavy deck sitting 4th).
Yes. The current charts do not have the same player repeating the same
position any more than necessary.
LSJ
> It shouldn't matter, but since it also won't hurt, a 7th criterion [see
> my comments below for the 6th criterion] can be added (one that the current
> charts do not use, but which could be used to illustrate that a new alternative
> seating chart is more optimal than the current ones so long as the other 5
> criteria are not sacrificed) - Noal's criterion:
>
> 6. Starting transfers are equitably distributed. [NOAL]
That should be "7."
> > no predator/prey relationships are repeated
> > (much more important than grand-predator/grand-prey relationships), and
>
> Yes. The current charts follow this. Indeed, it is mandatory (save
> when not possible, as in a 4-player 3+F tournament), by the VEKN rules.
> [VEKN 3.1.2]
Nevermind about that "when possible" bit - the Prey/Predator repeat
is always avoidable (and is always avoided), even in 3+F 4-player.
Joshua Duffin
"Noal McDonald" <dhar...@none-of-your-fscking-business.com> wrote in
message news:3C968EB9...@none-of-your-fscking-business.com...
> Joshua Duffin wrote:
>
> > In a lengthy email, LSJ convinced me that positions 1-4
> > are all equally good.
>
> Hrm. I'd be interested to see that e-mail. Additional insight into game
> mechanics is always a good thing. :-)
I miswrote there actually - it wasn't a lengthy single
email but rather a lengthy email argument. :-) That
said, I'd be happy to post excerpts here if Scott doesn't
mind.
> > Going first more than once is not actually a disadvantage;
> > you get an extra master phase and discard, and you get to
> > 5 transfers before anyone else does.
>
> Hrm. Not sure about the intrinsic values there.
>
> The discard *may* be an advatage, but I doubt it's a statistically
> significant one.
Maybe not, but if it's a good thing to have in general,
it's an advantage of having an "extra" turn.
> Focussed decks should be have a lot of redundancy,
> meaning discarding a card is likely to get it replaced with the same
> card or something else you already have in your hand. Toolbox decks will
> have a number of cards available to counter other strategies which could
> be discarded. However, it's not likely that you're going to know which
> ones to ditch and which ones to keep on the opening draw.
But if a focused deck is missing something it knows it
has a lot of (and instead has too much of something else)
the discard is useful. The toolbox may not know what
needs discarding in terms of counter-strategy-cards it's
holding, but it may be able to decide that something in
hand won't be useful soon enough to be worth hanging on
to, or at the least gets a chance to tune the hand
toward a more-desirable balance of masters, actions,
reactions, etc.
> The master card phase can be very useful, I'll grant. Anarch Revolt,
> Dreams, Tomb, Info Highway, Jake Washington are all good cards to play
> in the opening round for player 1. However, most master cards require a
> minion of some kind to play, making the phase useless until you get a
> vampire out.
Whether or not you can make use of it in a particular game,
though, having the opportunity to play your second master
card before anyone else plays their second is an advantage.
> The 5th xfer may be useful...unless of course, your vampires are all
> even numbered capacity. :-P
Right. Certainly possible. But not an a priori argument
that going first is bad.
> > (For certain decks, certain seats are better than others,
>
> Which, as shown above, is certainly the case. Evening out the starting
> xfers certainly ensures that one deck isn't significantly helped, or
> hindered, more than another. If you can limit decks to 7-9 starting
> xfers after 3 rounds, rather than 6-10, I'd say that's probably a good
> thing.
I agree that starting transfers should be about as even
as possible given other constraints - the more even, the
closer people are to having played the same "table spots"
as each other, and as you say, that means that each deck
should get about the same advantages/disadvantages from
starting positions over the tournament.
> Thanks. :-) Maybe I will spend some time going through them. The 5- and
> 13-player tables didn't take much time since I just used the system that
> I've been using for years to determine seating at tournaments.
Do *you* have an algorithm? Or do you just start writing
stuff down and look at it? (That's pretty much what I've
done when I've tried it, but it seems very time-consuming
and easy to make mistakes.)
Josh
going for the rhythm
Joshua Duffin
>
> I *am*
> expecting to see a vampire first turn if I'm number 5, or a big fucker
second
> turn. You can be ahead of me on master phases, if I can be ahead of you
on
> actions.
You can't assume that player 5 will get out a vampire first
turn, though, if you're assuming that player 1 has some
large chance of *not* getting out a vampire by his *second*
turn. :-)
Maybe it would help to compare being player 1 to being
player 6: if you were 6th, you would still be preyed on
by player 5, who got 4 transfers his first turn. But
you wouldn't have had the extra 1 transfer, master phase,
and discard for going first. That would be worse, right?
Same thing for player 5. His predator got 4 transfers
before he did. His prey got 1 transfer before he did.
And he doesn't get any extra phases to make up for it.
If everybody's vampires are the same size, each player
always gets his first vampire out before his prey. But
no player gets to take an action before his prey has
a vampire out. I think it's a mistake to look at it
as "well, by my 2nd turn I have 8 transfers, and the
1st player only has 5 by his second turn." You should
look at it as a continual flow of time. The 4th player
will have 8 transfers after his second turn. But the
1st player will have gotten to 9 transfers two turns
later (on his third turn, which comes before the 4th
player's third turn).
Sure, if you have no odd-capacity vampires in your crypt,
going first kinda sucks. But so what? If your crypt
is all 3-caps, going 2nd kinda sucks.
> So, well... that's really all I'm saying, is that position is a huge deal
to
> many decks.
It matters, but I doubt that it's a *huge* deal. If your
deck totally sucks when you have to go first, you should
probably change it so it doesn't. That first time around
the table shouldn't have that large an impact on how you
do.
That said, the seating charts *do* take it into account
as far as not having people repeat play-order-spots any
more than necessary, and they can now take it into account
in trying to give people relatively even total starting
transfers over the three rounds.
Josh
transference
Chris Berger
"Joshua Duffin" <jtdu...@yahoo.com> wrote in message
news:a77osp$jc82e$1...@ID-121616.news.dfncis.de...
>
> "Chris Berger" <ark...@ugcs.caltech.edu> wrote in message
> news:OjBl8.66874$af7.47715@rwcrnsc53...
> >
> > I *am*
> > expecting to see a vampire first turn if I'm number 5, or a big fucker
> second
> > turn. You can be ahead of me on master phases, if I can be ahead of you
> on
> > actions.
>
> You can't assume that player 5 will get out a vampire first
> turn, though, if you're assuming that player 1 has some
> large chance of *not* getting out a vampire by his *second*
> turn. :-)
>
fucker second turn.
> Maybe it would help to compare being player 1 to being
> player 6: if you were 6th, you would still be preyed on
> by player 5, who got 4 transfers his first turn. But
> you wouldn't have had the extra 1 transfer, master phase,
> and discard for going first. That would be worse, right?
>
Nope. Because you'd still be able to prey on player 1, who's still in a worse
position than you. 8)
LSJ
>
> Joshua Duffin wrote:
>
> > In a lengthy email, LSJ convinced me that positions 1-4
> > are all equally good.
>
> Hrm. I'd be interested to see that e-mail. Additional insight into game
> mechanics is always a good thing. :-)
I'd be interested in hearing it, too. I don't have copies of
those email handy.
Probably goes something like:
Entries are (activeCap/tranfers/masterphases)
<PRE>
Turn Player 1 Player 2 Player 3 Player 4 Player 5
1 0/1/1 0/0/0 0/0/0 0/0/0 0/0/0
2 0/1/1 0/2/1 0/0/0 0/0/0 0/0/0
3 0/1/1 0/2/1 0/3/1 0/0/0 0/0/0
4 0/1/1 0/2/1 0/3/1 0/4/1 0/0/0
5 0/1/1 0/2/1 0/3/1 0/4/1 0/4/1
6 1/5/2 0/2/1 0/3/1 0/4/1 0/4/1
7 1/5/2 2/6/2 0/3/1 0/4/1 0/4/1
</PRE>
and so on...
It is not difficult to see the balance (except for #5) in
that system.
At any rate, I asked Todd to examine the tournament reports he's
received (as it turns out, a subset - just the most recent ones -
the ones that were readily available on his laptop) to report
statistical results.
The results show seat 5 doing better than seat 4, so either the
analysis above isn't correct, or the sample is not statistically
significant enough to overcome other factors (like deck design
and player ability). The latter seems more likely, so don't put
too much stock in these numbers. Sorry I don't have better numbers
to offer.
Here are the numbers:
4-player tables
(number in sample: 92)
----------------------
VPs (including half-vps at time-out):
#1: 27.9%
#2: 21.1%
#3: 25.6%
#4: 25.4%
GWs (73.9% - 68 of the 92 - tables produced game wins):
#1: 32.4%
#2: 20.6%
#3: 23.5%
#4: 23.5%
5-player tables
(number in sample: 116)
-----------------------
VPs (including half-vps at time-out):
#1: 21.3%
#2: 24.5%
#3: 20.0%
#4: 15.3%
#5: 18.9%
GWs (81.0% - 94 of the 116 - tables produced game wins):
#1: 22.3%
#2: 28.7%
#3: 14.9%
#4: 12.8%
#5: 21.3%
Curevei
>Hungry Coyote, or any other hideously-convenient-first-turn-master out
>first, meaning the rest of the table has to contest or discard --
>something that only player 5 is in a position to do. But he always gets
>to use it first -- and he gets one free discard before anyone gets to do
>anything, something which we all established long ago is an underrated
>resource.
There was a time when I dreaded going first as the first Info Highway always
seemed to get Sudden-ed. Course, this isn't meaningful to real analysis.
What I dispute is the advantage of having the first discard. Some players
discard, like, every turn. I recall Curt playing that way. But, a lot don't.
When you go first, you have precisely zero information on what would be useful
against your opponents. The only reason to discard would be if your opening
hand seemed excessively redundant. Even then, that redundancy might be
precisely what you need against certain decks.
In the fifth seat, which I'm not disputing is the weakest more often than not,
you might get some inkling of what the other decks are trying to do. Even a
second time around the table still puts you in a better position to read the
table than if you went first.
Joshua Duffin
> Noal McDonald wrote:
> > Hrm. I'd be interested to see that e-mail. Additional insight into game
> > mechanics is always a good thing. :-)
>
> I'd be interested in hearing it, too. I don't have copies of
> those email handy.
OK. It may not be entirely obvious just from the text why
I was convinced at the end, but you'll have to take my
word that I am now convinced. :-)
It went something like this...
I asked about the 5-player chart and why player 2 went
first twice.
[Throughout the rest of this message, single > denotes
text quoted from the other person's email. > >, where
used, is the current writer's text from the email before
last. Text within square brackets [] was added in composing
this message.]
[Scott wrote]
"Going first" has no priority over any other position, regardless
of what players are more concerned about. Players who are more
concerned about repeating 1st than repeating other positions
merely haven't thought about it enough.
Since repeating a seating cannot be avoided, all options are equal
(except for having 5 start, which makes 3 3rd all three rounds).
[I wrote]
Really? Isn't that only true if all positions are in fact equal?
Most players (myself included) consider going first less good than
any other position (unless you're playing a deck with lots of 1-cap
vampires). Fourth is clearly better than fifth, but except for
that, the less-transfers disadvantage has at least the potential of
outweighing the earlier-turn advantage.
[Scott wrote]
Going first = going last plus an extra master phase, discard phase,
and an "extra" transfer that you get at the start of the game
(quotes because your prey gets an extra as well, and so on).
[I wrote]
If you go last, you get to take your first action with a 4-cap
before the person who went first does. Yes, the person who went
first can have out a 5-cap vampire at that time, but she only gets
an action on her second turn if she drew a 1-cap vampire; the person
who goes last can get an action on her second turn if she draws a
vampire of capacity 4 or less.
IOW, you certainly get the extra master and discard, but you don't
get a time-advantage on actions unless the 1 transfer makes a
difference in being able to get out a vampire vs not.
Am I missing something here?
Fourth is definitely better than fifth; therefore, not all positions
are equal. That doesn't prove that first is worse than second, but
it does suggest that it could be, and as far as I can tell it will
be worse in some (fairly common) situations.
Here's a chart of "possible capacity-points of vampire actions" for
a five-player table:
0 0 0 0 0
1 2 3 4 4
5 6 7 8 8
...
2 capacity-points is significantly better than 1; likewise, 3 better
than 2, etc. The first player does get to 6 [NB 3-19-2002: here
I'm referring to "cumulative capacity-action points"] before anyone
else goes over 4, but then the second player gets to 8 before anyone
goes over 6, and so on, so it seems to me that that's not enough of
an advantage to outweigh having only 1 on the second time around the
table.
[Scott wrote]
Narrow thinking.
The first player can take her first action with a 5-cap before anyone
else does. etc.
"Action on second turn" is not relevant. You can treat the first player's
second turn as her first (as if she went last) with the bonus MPA and
discard (and transfer) if you like.
The ramp up system ensures that everything is equal down past the resolution
of, say, luck of the draw etc.
>IOW, you certainly get the extra master and discard, but you don't
>get a time-advantage on actions unless the 1 transfer makes a
>difference in being able to get out a vampire vs not.
Extendable to any position 1-4.
>Am I missing something here?
Just that the arguments apply equally to all four positions.
>Fourth is definitely better than fifth; therefore, not all positions
>are equal. That doesn't prove that first is worse than second, but
>it does suggest that it could be, and as far as I can tell it will
>be worse in some (fairly common) situations.
And will be better in some (fairly common) situtions.
[I wrote]
I'm still not convinced. Maybe I'm just stubborn. :-)
(On the first player vs last player point: You can treat the
first player as if she went "sixth", but going last *is* a
disadvantage, except for the additional MPA, discard, transfer.
Whether those outweigh the going-last disadvantage is a
question, but, well, that's the issue I guess.)
Action on turn X may not be exactly relevant in that form, but
actions by time T is certainly relevant. Number of actions
taken is very important to winning at VTES, which is why weenie
decks are as strong as they are.
The earlier you have N capacity of vampires able to act, the
better you're doing. When N=1, the actions you can take are
fairly limited. You have more options (and effectiveness) at
N=2, and so on.
What the ramp-up system ensures is that it's not a big advantage
to go first: if you got 4 transfers the first turn, it'd be
obviously the best position to have. Starting at 1 transfer is,
I think, undeniably a disadvantage. Whether the additional MPA
and discard makes up for it is arguable, but there's no way that
you wouldn't be better off getting more transfers on your first
turn.
You do agree that fourth is better than fifth, right? In that
case, isn't it clear that it *does* matter which positions are
repeated?
> Extendable to any position 1-4.
Yes. But the additional transfers make more of a difference
in "got a vampire out" vs "didn't" as they increase, because
not every deck has vampires of 1-cap or lower, but many decks
have vampires of 4-cap or lower.
[first worse than second?]
> And will be better in some (fairly common) situtions.
Better if you want to get out vampires with capacities adding up
to 1, 5, 9, etc. Worse for 2, 6, 10, etc. If it's worse as often
as it's better, doesn't that imply that it matters? Or is your
point of view that, if the probabilities are equal, it doesn't
matter? Even then, the probabilities are going to be predictably
unequal for some decks, which would then be "hurt" by going first,
making it reasonable for their players to not want to repeat the
first position.
[Scott wrote]
> Starting at 1 transfer is,
>I think, undeniably a disadvantage.
I deny it.
It's a function of the capacities found in your starting uncontrolled
region, mostly.
>You do agree that fourth is better than fifth, right?
Yes. I've said so already.
>case, isn't it clear that it *does* matter which positions are
>repeated?
I've always restricted my statements to positions 1-4.
No one repeats 5th, ever.
>Better if you want to get out vampires with capacities adding up
>to 1, 5, 9, etc. Worse for 2, 6, 10, etc. If it's worse as often
>as it's better, doesn't that imply that it matters? Or is your
No. That implies that it doesn't.
>point of view that, if the probabilities are equal, it doesn't
>matter? Even then, the probabilities are going to be predictably
That's "my view" (and standard view, statistically speaking).
>unequal for some decks, which would then be "hurt" by going first,
>making it reasonable for their players to not want to repeat the
>first position.
Then it's a function of deck, not of seating charts.
Some decks may not want to repeat 2nd position. Or 3rd. etc.
[I wrote]
> I deny it.
> It's a function of the capacities found in your starting uncontrolled
> region, mostly.
OK, that I pretty much agree with. I just think that, for almost
all decks in tournaments, those capacities make first worse than
fourth. :-)
[NB 3-19-2002: This may be true; I'm not sure if I still believe
it. But even if it is true, the seating charts already have
people not repeat positions any more than necessary, and they
will soon, apparently, try to "spread out" positions so that
starting transfers are as equal as possible. That means that
everyone should be, over a tournament, about equally affected by
starting-transfer issues.]
> >if the probabilities are equal, it doesn't
> >matter?
>
> That's "my view" (and standard view, statistically speaking).
Hmm, it is, isn't it. I should've argued that it's worse more often
than it's better. Which I think it is, but it does depend on what
people choose for their vampire capacities, so I guess it's beyond
the scope of optimal seating.
> Then it's a function of deck, not of seating charts.
> Some decks may not want to repeat 2nd position. Or 3rd. etc.
Right. I suppose they just have to get screwed sometimes, because
someone has to get screwed sometimes. :-)
------
[end of transcription]
Josh
recapped
Noal McDonald
>>> The ramping system is added to the game as a balancing factor.
>>
>> How so. Would you care to explain this to me?
>
> player 1 has a distinct advantage.
Ah. Wasn't sure what you meant by "ramping system." Evening out
initial transfers is an attempt to equalize the players' overall
placement in the ramping system.
> OK. That looks no less optimal than the current.
> Certainly evenening out "initial transfers" cannot hurt.
> I'll update the 13-player chart in the next release of the Archon.
>
> Thanks
Thank you. :-)
Regards,
Noal
Joshua Duffin
>
> "Joshua Duffin" <jtdu...@yahoo.com> wrote in message
> news:a77osp$jc82e$1...@ID-121616.news.dfncis.de...
> > You can't assume that player 5 will get out a vampire first
> > turn, though, if you're assuming that player 1 has some
> > large chance of *not* getting out a vampire by his *second*
> > turn. :-)
> >
> I know, but I can assume that I'll either have a first turn vamp, or a big
> fucker second turn.
Same for the first player, except substitute "second" for
"first" and "third" for "second". You still don't get
your first vamp (or big fucker) out more than one player-
turn later than the next slowest person, and you may well
get him out before your prey (who has only had 2 transfers
when you make your 5th and 6 when you make your 9th), just
like your predator does to you.
> > Maybe it would help to compare being player 1 to being
> > player 6: if you were 6th, you would still be preyed on
> > by player 5, who got 4 transfers his first turn. But
> > you wouldn't have had the extra 1 transfer, master phase,
> > and discard for going first. That would be worse, right?
> >
> Nope. Because you'd still be able to prey on player 1, who's still in a
worse
> position than you. 8)
You may be kidding, but if not: player 1 *isn't* in a worse
position than you; his predator doesn't get to go before
him with the same number of transfers.
Player 1's prey gets 1 more transfer than him, one player-
turn later. Player 1's predator gets 3 more transfers than
him, four player-turns later. Player 5's prey gets 1 more
transfer than him, one player-turn later. Player 5's
predator gets 0 more transfers than him, one player-turn
*earlier*. (or 4 more transfers, four player-turns later.)
That's worse.
Josh
not quite clear on this whole good-bad thing
Noal McDonald
> Here are the numbers:
>
> 4-player tables
> (number in sample: 92)
> ----------------------
>
> VPs (including half-vps at time-out):
> #1: 27.9%
> #2: 21.1%
> #3: 25.6%
> #4: 25.4%
>
> GWs (73.9% - 68 of the 92 - tables produced game wins):
> #1: 32.4%
> #2: 20.6%
> #3: 23.5%
> #4: 23.5%
These charts show that on a 4 player table, seat 1 has a clear
statistical advantage, reinforcing LSJs claims.
> 5-player tables
> (number in sample: 116)
> -----------------------
>
> VPs (including half-vps at time-out):
> #1: 21.3%
> #2: 24.5%
> #3: 20.0%
> #4: 15.3%
> #5: 18.9%
>
> GWs (81.0% - 94 of the 116 - tables produced game wins):
> #1: 22.3%
> #2: 28.7%
> #3: 14.9%
> #4: 12.8%
> #5: 21.3%
However, on a 5 player table, the posit that seat 4 is better than
seat 5 is contradicted by the stats. Player 1 still has an advantage
over both of those positions, again backing up LSJ's theories, but
seat 2 has gone from worst to a significant first.
Granted, these stats don't show all the factors that led to those
results...and VTES has a lot of factors.
Regards,
Noal
LSJ
> > VPs (including half-vps at time-out):
> > #5: 18.9%
> >
> > GWs (81.0% - 94 of the 116 - tables produced game wins):
> > #5: 21.3%
>
> However, on a 5 player table, the posit that seat 4 is better than
> seat 5 is contradicted by the stats. Player 1 still has an advantage
> over both of those positions, again backing up LSJ's theories, but
> seat 2 has gone from worst to a significant first.
>
> Granted, these stats don't show all the factors that led to those
> results...and VTES has a lot of factors.
Right. From the post you're quoting above:
[QUOTE]
The results show seat 5 doing better than seat 4, so either the
analysis above isn't correct, or the sample is not statistically
significant enough to overcome other factors (like deck design
and player ability). The latter seems more likely, so don't put
too much stock in these numbers. Sorry I don't have better numbers
to offer.
[ENDQUOTE]
Noal McDonald
> [QUOTE]
> The results show seat 5 doing better than seat 4, so either the
> analysis above isn't correct, or the sample is not statistically
> significant enough to overcome other factors (like deck design
> and player ability). The latter seems more likely, so don't put
> too much stock in these numbers. Sorry I don't have better numbers
> to offer.
> [ENDQUOTE]
Well, thanks for the clarification! :-)
Seriously though, your quote had more to do with seats 4 and 5. I was
focusing more on the other seats. Either way, there's no reason to
apologize for the numbers. As you note, there's just not enough raw
data to empirically draw any provable conclusions.
You'll note that, statistically, seat 2 does the best and seat 4 does
the worst on 5 player table. I would conclude that if you have to
repeat seatings, as in the current 5 player seating chart, you should
avoid repeating those positions in 5 player tables. However, in 4
player tables, the stats show those are the positions you would want
repeated if you had to.
Regards,
Noal
LSJ
... based on questionable data.
If there's any actual reason to believe that seat 4 is in fact
a detrimental position, I'd be happy to re-examine the charts.
But I can't think of any reason that the stats aren't simply lying. :-)
CurtAdams
>The results show seat 5 doing better than seat 4, so either the
>analysis above isn't correct, or the sample is not statistically
>significant enough to overcome other factors (like deck design
>and player ability). The latter seems more likely, so don't put
>too much stock in these numbers. Sorry I don't have better numbers
>to offer.
And a chi-square in Excel gives a p value of .11, which is larger
than .05, so the standard interpretation is that this does *NOT*
demonstrate difference in win rates between the seats. That
even a > 2:1 ratio between some seats is not significant indicates
we have way too little data to say much. To pick out a difference
of 20% we would need literally thousands of games.
Curt Adams (curt...@aol.com)
"It is better to be wrong than to be vague" - Freeman Dyson
Noal McDonald
> For determining "optimal" arrangements, however, the following goals (a
> superset of the DCI guidelines) are used:
>
> 1. No pair of players repeat their predator-prey relationship. This is
> mandatory, by the DCI rules.
> 2. No pair of players share a table through all three rounds. An extremal
> version of an option from the DCI rules.
> 3. Available VPs are equitably distributed. That is, if one player plays
> at three five-player tables during the tournament, she'll have 15
> "available" VPs. In this case, no other player should have less than
> 14 "available" VPs (that is, the difference in available VPs should be
> no more than one and should be zero if possible).
> 4. No pair of players share a table more often than necessary. This is an
> option from the DCI rules.
> 5. A player doesn't play in the same position. If a player plays second
> at one table, she shouldn't be seated "second" in any other round.
> seats 1-4 is preferable to repeating seat 5.
I noted that none of this prohibits repeating of relative position,
other than predator-prey, which you used to disqualify my suggestion
for the 5 player seating chart. Perfectly understandable, but if you
want that to be a criteria, would you please modify criterium #1 to
include that?
Along those lines, take note of the 9 player seating charts displayed
below. In the current one, player 2 is the grand-predator of player 4
in all three rounds. In the proposed seating chart below, 3 is the
grand-predator of 4 twice and 7 is the grand-predator of 5 twice.
Neither are repeated three times. I'm not certain whether you would
view this as an improvement or not, but I would. I'd like your view on
this before I proceed further.
In both the current chart and my suggestion, three pairs of players
are seated at the same table in all three rounds. Unfortunately, I
don't believe that can be avoided. Also, in the current one, player 5
has 6 starting transfers. In the suggested chart, all players have 7-9
starting transfers.
Current - 9 players
Round 1) 1 2 3 4 | 5 6 7 8 9
Round 2) 3 7 9 6 | 4 5 8 2 1
Round 3) 6 8 5 1 | 2 9 4 7 3
Suggested - 9 players
Round 1) 1 2 3 4 | 5 6 7 8 9
Round 2) 9 7 2 5 | 4 8 6 3 1
Round 3) 8 1 9 6 | 3 5 4 2 7
Regards,
Noal McDonald
VEKN Prince of Metro Detroit
LSJ
As I said - that list is ancient.
I believe somewhere in this thread I indicated that I avoid seating
the same "relative positions".
> Along those lines, take note of the 9 player seating charts displayed
> below. In the current one, player 2 is the grand-predator of player 4
> in all three rounds. In the proposed seating chart below, 3 is the
> grand-predator of 4 twice and 7 is the grand-predator of 5 twice.
> Neither are repeated three times. I'm not certain whether you would
> view this as an improvement or not, but I would. I'd like your view on
> this before I proceed further.
It's an improvement, sure.
For cases where it matters (not this one), I view five different
"relative position" relationship values:
1) prey
2) predator
3) ally (cross-table at a 4-player)
4) grand-prey (at a 5)
5) grand-predator (at a 5)
When possible, I also group 1 and 2 as "neighbors" and 3,4,5 as
"cross-table" (and avoid repeating those designations as well).
> In both the current chart and my suggestion, three pairs of players
> are seated at the same table in all three rounds. Unfortunately, I
> don't believe that can be avoided. Also, in the current one, player 5
> has 6 starting transfers. In the suggested chart, all players have 7-9
> starting transfers.
>
> Current - 9 players
> Round 1) 1 2 3 4 | 5 6 7 8 9
> Round 2) 3 7 9 6 | 4 5 8 2 1
> Round 3) 6 8 5 1 | 2 9 4 7 3
>
> Suggested - 9 players
> Round 1) 1 2 3 4 | 5 6 7 8 9
> Round 2) 9 7 2 5 | 4 8 6 3 1
> Round 3) 8 1 9 6 | 3 5 4 2 7
Quite right.
BTW,
I've produced new seating charts for 14,15,17,17,19,20,23,24, and 26
players which include the #7 criterion (to avoid duplicating effort).
LSJ
Optimal Seating Criteria, with additional previously-unformalized criteria included:
1. No pair of players repeat their predator-prey relationship. This is mandatory, by the VEKN rules.
2. No pair of players share a table through all three rounds, when possible.
3. Available VPs are equitably distributed.
4. No pair of players share a table more often than necessary.
5. A player doesn't sit in the fifth seat more than once.
6. No pair of players repeat the same relative position[*], when possible.
7. A player doesn't play in the same seat position, if possible.
8. Starting transfers are equitably distributed. [NOAL]
9. No pair of players repeat the same relative position group[^], when possible.
[*] "relative position" relationship values:
A) prey
B) predator
C) cross-table at a 4-player
D) grand-prey at a 5
E) grand-predator at a 5
Note that repeating A and repeating B is already handled (prohibited) by criterion #1.
[^] "relative position group" values:
i) Adjacent (prey or predator)
ii) Not adjacent
Noal McDonald
> I believe somewhere in this thread I indicated that I avoid seating
> the same "relative positions".
You did. In fact, I referred to it in the text to which this was a
reply. I noticed your next post which contains the updated list of
criteria. Very nice. :-)
> I've produced new seating charts for 14,15,17,17,19,20,23,24, and 26
> players which include the #7 criterion (to avoid duplicating effort).
Wow. You're pretty damned efficient. I've only had time to go through 4,
5, 8 and 9, of which I could only improve #9. I assume the charts you've
changed (17 twice? must be 18) will be in the next version of the
Archon, so I'll look at them when you release it. I'll go through the
others as time permits.
Regards,
Noal
Henrik Isaksson
diskussionsgruppsmeddelandet:3C9CAE2...@white-wolf.com...
I would like the 9th criteria to be moved up to place 5. The player that I
sit next to affects the play more for me than any grandprey/predator
relationship, and i also think it's less important which seat and how many
transfers I get. It's simply no fun to sit next to, for instance, dedicated
rush or dedicated intercept twice in a row, but if I get a weak ally (or
grandprey/predator) twice it's not the end of the worlds...
My 2 cents
--
Henrik Isaksson