A Proposition Game
Pete
proposition game, but I don't recall the name of it.
One side would start in the normal position, and the other side would start
with all 15 men in the air. If I remember it right, the first side gave
about 7 to 1 money odds to the other side (depending on negotiation skills).
The cube was used as normal.
Is the game still played? Anybody recall the name, or other details?
Pete
Larrikin
around in the 70's! Must be closer to 2:1 odds - any taker's at 7:1?
"Pete" <pe...@garlic.com> wrote in message news:<ut13ik7...@corp.supernews.com>...
Michael Sullivan
> "Pete" <pe...@garlic.com> wrote in message
> news:<ut13ik7...@corp.supernews.com>...
> > Back in the 70's when I used to play a lot in various clubs, we had a
> > proposition game, but I don't recall the name of it.
> > One side would start in the normal position, and the other side would start
> > with all 15 men in the air. If I remember it right, the first side gave
> > about 7 to 1 money odds to the other side (depending on negotiation skills).
> > The cube was used as normal.
> > Is the game still played? Anybody recall the name, or other details?
> Sorry, I don't know the details but wow!!! 7:1 odds - I wish I'd been
> around in the 70's! Must be closer to 2:1 odds - any taker's at 7:1?
I'm not a good enough player to judge whether 7:1 is right, rather than
5, 6:1, but I'd give 2:1 odds in a heartbeat for significant money if
you'd agree to 10 games, without knowing who you are, and I'm not all
that good yet (<1500 on FIBS).
Consider that you have at a minimum 4 rolls (and more realistically 6 or
7) in which you can't hit anything outside my home board, which gives me
free reign to maximize my flexibility for making home board points and
reshots on any blots you leave in there during that time.
If you disagree, I'll gladly take you up on this proposition at 2:1 if
we can find an appropriate escrow holder. (Can you do proposition games
on any of the commonly used servers?)
Michael
--
Michael Sullivan
Business Card Express of CT Thermographers to the Trade
Cheshire, CT mic...@bcect.com
Martin Hemming
> on any of the commonly used servers?)
>
>
> Michael
Yes you can play this as a prop on gamesgrid. However you can't begin
with a rolloff for first roll as a normal game should start. One
player must be on roll.
I would offer 4 to 1 to the man on the bar, and give him choice of
whether to roll first or second. Minimum 10 games at $10 a point.
Larrikin
As I indicated (really only half seriousely), I would play it at 7:1
without having any firm idea of the true value of this game but not at
2:1, too risky!
My rough appraisal of the position was that the gammon threat would
probably be strongly one-sided but that the side with 15 in the air
would be able to get 3 or 4 home board points reasonably comfortably
and then play a very well timed backgame with no men out of play.
Bearing in against 3 or 4 points would prove difficult and, I am
guessing of course, but 50% wins for each side might not be
unreasonable. It's only the gammon and backgammon threat that pushes
up the odds. 2:1 might be very optimistic but who knows? Maybe some
of the hotshots from the 70's who have played it have a good feel for
the position.
I wasn't really shaping up for a fight but if we could find somewhere
to play it online I'm game if your game, pending further negotiations
of course.
In the mean time I'll try feeding the position to one of the bots and
see what it thinks.
mic...@bcect.com (Michael Sullivan) wrote in message news:<1fll1x8.ute6zrvlvq2zN%mic...@bcect.com>...
Peter Schneider
> I am
> guessing of course,
> but 50% wins for each side might not be
> unreasonable.
Maybe I show my ignorance, but I bet USD 100 immediately that the player on
the bar wins less than 50%. Come get it ;-).
Regards,
Peter aka the juggler on fibs
Michael Sullivan
> I make no claim to expertise either and I don't play on any server.
> As I indicated (really only half seriousely), I would play it at 7:1
> without having any firm idea of the true value of this game but not at
> 2:1, too risky!
heh. A lot easier to talk than bet, I see. :)
> My rough appraisal of the position was that the gammon threat would
> probably be strongly one-sided but that the side with 15 in the air
> would be able to get 3 or 4 home board points reasonably comfortably
> and then play a very well timed backgame with no men out of play.
> Bearing in against 3 or 4 points would prove difficult and, I am
> guessing of course, but 50% wins for each side might not be
> unreasonable. It's only the gammon and backgammon threat that pushes
> up the odds.
No way. 50% wins? I can't imagine that. Remember there are only 15
checkers! I assume you mean 3 or 4 points in the opponents home board?
If so, your own board is going to be weak. With 6 checkers in my home
board, the most points you can hold in your own homeboard will be 4,
leaving only a 50/50 shot at a dance if you hit me. With 6 checkers
back, I'm going to have to dance more than a couple times for you to
catch up on pip count and abandon your backgame. And with 7 points,
you're going to be crunching fast, so you won't get many shots to hit me
in that good a position.
To play the backgame, you've got to get 5-6 points in your home board,
so you can only leave one anchor in mine. And it's going to take you a
*long* time to get 4-5 home board points. I might win before you get
there. Again, timing becomes critical, you end up with 6-7 points for a
playable backgame, which means you'll only get 3-4 rolls before you
crunch away a playable backgame position. Even if I crunch and leave
you a shot in that span, you've got to make the rolls, and you may only
have one or two chances with a less than 50% shot. And *that* may not
even be enough. If I get 2/3 of my guys off before you hit me, I might
be able to get in and get back around to beat you, even if I dance once
or twice. For a sure backgame win, you need to hit 2 or 3 of my guys
while you have a strong board.
In general, the game only plays well for you when you get the right
rolls in a pretty small window. Backgames are always a last resort,
with good reason. Normally, if your only shot is a backgame, your
opponent should *not* double, because the market is already gone --
you're as likely to get gammoned as win.
Seriously, do a rollout, I'll bet it gives you less than a 25% shot to
win, with significant negative gammon & backgammon chances.
Think about this. If somehow the position discussed could be arrived at
on the board in a money game, and your opponent doubled, would you take
the cube?
If you wouldn't take, you're saying you need at least 3-1 odds, since
you only drop a double with <=-.5 game equity (25% you / .75% opp).
> I wasn't really shaping up for a fight but if we could find somewhere
> to play it online I'm game if your game, pending further negotiations
> of course.
> In the mean time I'll try feeding the position to one of the bots and
> see what it thinks.
I don't have a bot (mac user -- I'm waiting to build gnubg when I
install X) but I'd be interested to know what one thought.
Douglas Zare
Pete wrote:
> If I remember it right, the first side gave
> about 7 to 1 money odds to the other side (depending on negotiation skills).
> The cube was used as normal.
I've seen n:1 odds given one some props, such as winning while closed out, or
winning against 14 off. I've seen points per game offered for a prop with the
cube in play. However, what do you mean by 7 to 1 money odds with the cube used
as normal? 3/4 of a point paid with a centered cube?
Douglas Zare
Julian Hayward
Hemming <mhem...@lineone.net> writes
I get 4/1 a little mean for just winning. However, for money it's more
like 8/1, as the normal player averages more than 2 points per game when
he wins...
(Normal plays first: 16% loss, 18% single win, 38% gammon, 28% bgm)
( " " second: 19% loss, 20% single win, 34% gammon, 27% bgm)
--
Julian Hayward 'Booles' on FIBS jul...@ratbag.demon.co.uk
+44-1480-210097 http://www.ratbag.demon.co.uk/
---------------------------------------------------------------------------
"A witty saying proves nothing" - Voltaire
---------------------------------------------------------------------------
Nobody
> However, what do you mean by 7 to 1 money odds with the cube used
> as normal? 3/4 of a point paid with a centered cube?
If I win the game, you pay me $1/point. If you win, I pay you
$7/point. But I will insist that the Jacoby rule not be used. :-)
This prop was played by various people at the Novi tournament this year
(at 6:1 or 7:1; I forget). There are a few things to learn about the
checkerplay in this position, but it mostly seems to be about handling
the cube properly, which is nontrivial at the payoff odds. The prop is
good practice for playing the snake.
Douglas Zare
I don't think the finer points of rolling home an outside prime should
matter much. Most people would have no idea what to do with the doubling
cube when playing at odds. (Should you take if you win 50% and your
opponent is paid 7:1? Yes, even if it were 7:0. You should pass at 7:1 if
you only win 40% single games, though.) That said, I'm sure there is a lot
of skill involved in just the checker play, and I think this would be an
interesting prop by ordinary rules.
Douglas Zare
ihessel
> Pete wrote:
>
> > If I remember it right, the first side gave
> > about 7 to 1 money odds to the other side (depending on negotiation skills).
> > The cube was used as normal.
We varied the sides and it was most helpful in learning about 1/Backames
and 2/ Walking a prime
David Montgomery
There was a chouette playing this prop (at 6-1) at the last Michigan
Summer Championships so the prop lives on. It's a fun one. That
crowd liked giving the 6-1.
Michael Sullivan
> > > However, what do you mean by 7 to 1 money odds with the cube used
> > > as normal? 3/4 of a point paid with a centered cube?
This wouldn't work at all, (and this is true for any situation where the
odds are greater than 3:1. If you get your odds by paying, then the
person with the better position starts off with a double, and the plaer
with everybody on the bar has a drop.
> > If I win the game, you pay me $1/point. If you win, I pay you
> > $7/point. But I will insist that the Jacoby rule not be used. :-)
So you have to play this way, but it introduces some interesting
complexity. For instance player A (the one with checkers on the bar)
can reasonably double before 50% winning chances. With minimal gm/bgm,
45% chance represents a 2.6 equity, which is roughly equivalent to a
double at 68% chances on even odds. At only a 57% winning chance, A has
lost the market and B will *drop* a double. On the other side, player B
can't reasonably double until the game is almost sewn up, and it will be
*very* difficult to find a spot with a good double that is not a drop.
market losers will abound in any position with a reasonable double for
B.
Which makes me think 6:1 might be fairer. Some thinking about this has
made me realize that using the cube at odds *heavily* favors the player
getting the odds. Take the cube out of the match, and 8 to 1 might be
fairer, but with the cube, 7:1 might be a little high.
> I don't think the finer points of rolling home an outside prime should
> matter much. Most people would have no idea what to do with the doubling
> cube when playing at odds.
Yup. I'm realizing propositions like this would be incredible hustler's
games. If you're competent and have thought a lot about how to use the
cube at odds, you'd have a big advantage over someone who has not, even
if they are a better player in general.
> (Should you take if you win 50% and your
> opponent is paid 7:1? Yes, even if it were 7:0.
At 7:0, it's a hard decision. Opponent has 3.5 equity. Either taking
or dropping gives them 7. Since I don't stand to win *anything*, I'm
not sure it's right to give away the chance of paying out more than 7
rather than closing out the game. Of course, the real problem is that
you gave unwinnable odds in the first place. But that means any time
you can close the game out without giving anything up, I would think
that you should, no? So 50% would be a drop.
> You should pass at 7:1 if
> you only win 40% single games, though.)
It looks to me like 42-43% is the drop point at 7:1.
> That said, I'm sure there is a lot
> of skill involved in just the checker play, and I think this would be an
> interesting prop by ordinary rules.
It would certainly be good practice for backgames and what to do when
way behind.
Michael
r. van tiggelen
15 on the bar: right
opponent does not have opening position but 3 checkers on 6,5,4,3, and 2-pt.
Here comes the catch: side on bar starts AND does not roll but can CALL his
roll.
With this catch the seemingly hopeless side was thought to have 7 to 1
winning chances indeed.
I suppose if this is the game you were referring to your (like with normal
people) memory changed some details over time.
In that case you may be gratefull that some big-headed people remember
details better than normal.
This prop was played only against not-knowing people. Any shot will be hit
and allmost allways rolled home.
Hope this helped.
Ronald van Tiggelen. akaYETI on GG
Pete <pe...@garlic.com> schreef in berichtnieuws
ut13ik7...@corp.supernews.com...
Larrikin
You wrote.
>
> heh. A lot easier to talk than bet, I see. :)
>
After doing some rollouts and playing the position against Snowie 4 a
few times and I am willing to concede that your judgement was better
than mine concerning my flippant 2:1 odds statement. However, I only
ever "offered" to play it at 7:1 and you are welcome to try your luck
against me at these odds if you wish :)
Anyway, putting machismo aside for the moment, I did some SW4 2-ply
and 3-ply rollouts getting pretty similar results. These were done
3-ply full rollouts, 108 games, cubeless:
0.1%__0.2%___20.3%___79.7%___68.6%____51.3% Equity -1.792 +- 0.180
From memory, GNUBG 2-ply full rollouts, 432 games, cubeless, gave the
player on the bar about 30% winning chances and something like -1.300
equity (I did not save the rollout as it's often too much trouble with
GNUBG. As far as I know the only way to "save" the statistics is to
copy, paste and edit a bitmap image, or have I missed something
obvious?).
How well the bots play each side position is a big factor in trusting
these results. Is the skill needed by each player in checker-play
decisions equal or quite lopsided? You can argue that until the
player on the bar (A) gets all his men into action all the mistakes
are made by the opponent (B). However, assuming that (A) gets a shot
and hits he will have lots of difficult decisions to make and he will
potentionally be the one making the bulk of the mistakes.
Previousely, the bots have not had a good reputation for playing
backgames and containment positions. SW 4 is supposed to be a lot
better at both, still, this position is very unusual so Snowie's
rollout might not mean that much. Despite this, I'll also concede
defeat on my winning 50% of games estimate. I think the main problem
is, judging from the very few games I've played with Snowie, that
getting 3 or 4 points in (B)'s homeboard is far to optimistic :(
You also wrote:
> To play the backgame, you've got to get 5-6 points in your home board,
> so you can only leave one anchor in mine. And it's going to take you a
> *long* time to get 4-5 home board points. I might win before you get
> there. Again, timing becomes critical, you end up with 6-7 points for a
> playable backgame, which means you'll only get 3-4 rolls before you
> crunch away a playable backgame position. Even if I crunch and leave
> you a shot in that span, you've got to make the rolls, and you may only
> have one or two chances with a less than 50% shot. And *that* may not
> even be enough. If I get 2/3 of my guys off before you hit me, I might
> be able to get in and get back around to beat you, even if I dance once
> or twice. For a sure backgame win, you need to hit 2 or 3 of my guys
> while you have a strong board.
Again I'll state that I make no claim to expertise but I think you
have the wrong idea on how to play this position for (A). (A)'s
position is very unlikely to ever crunch. After (A) hits he forms the
best prime he can (intially a very broken one in (B)'s outerboard) and
gradually rolls this home. You are right that (A) needs to leave a
couple of anchored checkers behind but these are there to pick up the
second or third man. Of course rolling it home is not easy, nor
getting the second checker, but you do have 15 checkers in play to do
it with.
Ian Dunstan.
Pete
to see it still survives. I don't even know if the club has survived. I am
pretty sure we played it at 7 to 1, and the cube would sometimes get to 8 or
16, so the tension could be cut with a knife, if you're laying the price.
Pete
"David Montgomery" <davidmo...@netzero.net> wrote in message
news:3dd48d1c....@news.sonic.net...
Pete
his checkers. In fact most of his play is pretty automatic, no-brainer
types.
Generally the way these games go is that B hits everything in sight without
any care in the world. He hopes that A rolls a 6 with one of his dice. Both
sides are hoping for small doubles to make points in B's inner board. If A
managed to make the 1 and 2 (or 3 points) then he basically has a won game.
In fact, I don't think B can take a double.
"Larrikin" <larri...@optusnet.com.au> wrote in message
news:8c2ca95f.02111...@posting.google.com...
John R MacLeod
> player on the bar about 30% winning chances and something like -1.300
> equity (I did not save the rollout as it's often too much trouble with
> GNUBG. As far as I know the only way to "save" the statistics is to
> copy, paste and edit a bitmap image, or have I missed something
> obvious?).
I'm a beginner here but I wanted to play one of these games against gnubg
just for fun but I couldn't find anyway to set up the position - how do you
do it?
Larrikin
You wrote:
> I'm a beginner here but I wanted to play one of these games against gnubg
> just for fun but I couldn't find anyway to set up the position - how do you
> do it?
I have not had much experience with with GNUBG either but here is what
I do. Very briefly:
1) File menu: Start new game (you can not edit a position without
starting a game or match)
2) Use Hotkey "edit position" at top of screen
3) Click into the side panel to clear the board of all checkers, set
scores etc if required
4) Set up the position you want to play
5) Set up who is on roll (as far as I know you can not set up a
"nobody on roll" start to the game)
6) Click the dice and away you go
This should get you started, hope I didn't miss out a step or two :)
Ian.
John R MacLeod
gnubg three times running - two gammons and a backgammon and I'm certainly
not a good player. The only real surprise was in the first game when gnubg
still had 9 on the bar, 6 on the 3 point and I had all other points and had
just started to bear off - gnubg doubled me. I can't think why.
"Larrikin" <larri...@optusnet.com.au> wrote in message
news:8c2ca95f.02111...@posting.google.com...
Douglas Zare
Michael Sullivan wrote:
> Douglas Zare <za...@math.columbia.edu> wrote:
> > Nobody wrote:
> > > Douglas Zare wrote:
>
> > > > However, what do you mean by 7 to 1 money odds with the cube used
> > > > as normal? 3/4 of a point paid with a centered cube?
>
> This wouldn't work at all, (and this is true for any situation where the
> odds are greater than 3:1. If you get your odds by paying, then the
> person with the better position starts off with a double, and the plaer
> with everybody on the bar has a drop.
Rather than mess up the payoff system, one could pay the side on the bar to
hold a 2 cube, or pay for the right to hold a 2-cube against the player on the
bar. How much is fair? I don't think experience with the 6:1 version tells
you.
> [...]
>
> It would certainly be good practice for backgames and what to do when
> way behind.
I disagree. Most backgames that arise in practice do not involve 15 checkers
back. If you have 15 checkers back, certain things are very different from if
you have only 5, 8, or even 12 checkers back. You can kill/damage most
deep-anchor backgames by priming them, but if 15 checkers are back that only
threatens to make your opponent release the backwards anchor(s). Quite
different tactics kill backgames with 15 checkers back.
Most backgames do not produce full primes in the outer boards, but it is an
important strategy here. Most backgames produce some chances to run off the
backgammon in case the other side clears points early, but that tends to be
hopeless here. Most backgames allow you to choose to break anchors in order to
produce more shots (say, when your opponent has 12 off and 3 checkers on the 3
point), but here you might easily end up with 8 checkers on a point you would
like to break.
Douglas Zare