🚀 go-pugleaf

Looks like a close call to me...

hm..

In article <C0D3oC.KD4.1@cs.cmu.edu> Marc.Ringuette@GS80.SP.CS.CMU.EDU writes:
||  24 23 22 21 20 19 | | 18 17 16 15 14 13
||   O  O  O  O  O    | |
||   O  O  O  O  O    | |
||   O           O    | |
||               O    | |
||                    | |
||                    |X|
||   X  X             | |
||   X  X             | |
||   X  X  X  X  X    | |                       +-+
||   X  X  X  X  X    | |                       |2|
||   1  2  3  4  5  6 | |  7  8  9 10 11 12     +-+
||
||  Money game:  O to play 1-1
||
||  O (a good player) played 24/off 24/off 24/off 20/21, which I think
||  is a horrible play.  He couldn't be right, could he?

Some thoughts:

 1: O is going for gammon, and bearing off as many men as possible. If
    X enters on 24, he may not be able to escape immediately.  Still, I
    the odds aren't enough in favor of gammon for O to risk an unlucky
    next roll (6-5) and be hit.

 2: O wants protection from 6-6 as next roll from X, which otherwise
    would bring X only two men behind in bearing off.  But then again,
    only one roll out of 36 is a 6-6, and two men behind is too much to
    stand a reasonable chance of winning.  This is even worse than
    option 1.

 3: The money game is really a *virtual* money game, and O is not
    paying attention, or O is not such a good player after all.  :-)

No, I can't find a real reason either to bear off three men from 21.
I'd safely play 20/22(2), or, if I didn't feel 5-5 or 6-6 for my next
roll as a threat, 20/21 20/22 24/off.                          Vincent.

> In article <C0D3oC.KD4.1@cs.cmu.edu> Marc.Ringuette@GS80.SP.CS.CMU.EDU writes:
> ||  24 23 22 21 20 19 | | 18 17 16 15 14 13
> ||   O  O  O  O  O    | |
> ||   O  O  O  O  O    | |
> ||   O           O    | |
> ||               O    | |
> ||                    | |
> ||                    |X|
> ||   X  X             | |
> ||   X  X             | |
> ||   X  X  X  X  X    | |                       +-+
> ||   X  X  X  X  X    | |                       |2|
> ||   1  2  3  4  5  6 | |  7  8  9 10 11 12     +-+
> ||
> ||  Money game:  O to play 1-1
> ||
> ||  O (a good player) played 24/off 24/off 24/off 20/21, which I think
> ||  is a horrible play.  He couldn't be right, could he?
> 

In article <1993Jan5.084523.28443@rulway.LeidenUniv.nl>, zweije@slc9 (Vincent Zweije) writes:

> Some thoughts:
> 
>  1: O is going for gammon, and bearing off as many men as possible. If
>     X enters on 24, he may not be able to escape immediately.  Still, I
>     the odds aren't enough in favor of gammon for O to risk an unlucky
>     next roll (6-5) and be hit.
> 
I think it may be worth playing for gammon, because O doesn't risk very much
taking 3 men off

if X has to come out on 6 O has at least protected his lead in race. If X comes
in on 1 he has to jump over with 15 and 16, so same applies as above. if X roll
11, 12, 13 or 14 he has to break one point in his 5 point board. If X stays 
out O's gammon chances has raised. If X stays out or stays on 1, it is true
that O may leave a blot with 65 next roll (and maybe other shots later) but at
that point X has already at least 7 men off and even if hit O has still to roll
very well to turn the game (note that to become to this point it is much 
less prabable than X rolling 66 immediatelly). And if not hit immediatelly i 
think there wont be many future shots since X is forced to run soon (and if he
doesn't run soon gammon chances will be higher).

so i think taking 3 off can't be a big mistake, it may be the inferior move,
but definitely not by wide margin.

-Mika

>  2: O wants protection from 6-6 as next roll from X, which otherwise
>     would bring X only two men behind in bearing off.  But then again,
>     only one roll out of 36 is a 6-6, and two men behind is too much to
>     stand a reasonable chance of winning.  This is even worse than
>     option 1.
> 
>  3: The money game is really a *virtual* money game, and O is not
>     paying attention, or O is not such a good player after all.  :-)
> 
> No, I can't find a real reason either to bear off three men from 21.
> I'd safely play 20/22(2), or, if I didn't feel 5-5 or 6-6 for my next
> roll as a threat, 20/21 20/22 24/off.                          Vincent.

In article <C0D3oC.KD4.1@cs.cmu.edu> Marc.Ringuette@GS80.SP.CS.CMU.EDU writes:
>24 23 22 21 20 19 | | 18 17 16 15 14 13
> O  O  O  O  O    | |
> O  O  O  O  O    | |
> O           O    | |
>             O    | |
>                  | |
>                  |X|
> X  X             | |
> X  X             | |
> X  X  X  X  X    | |                       +-+
> X  X  X  X  X    | |                       |2|
> 1  2  3  4  5  6 | |  7  8  9 10 11 12     +-+
>
>Money game:  O to play 1-1
>
>O (a good player) played 24/off 24/off 24/off 20/21, which I think
>is a horrible play.  He couldn't be right, could he?
>
>
>-- Marc Ringuette (mnr@cs.cmu.edu)


I think he's right.

What risk is he taking?  

  1: X dances or lands on the 1 point and stays there.

  2: O rolls 5-6 3-5 3-6 and blots, taking off 2.

  3: O gets nuked

  4: X wins, despite having no 6 or 5 point and an opponent with only 8
     checkers on the board.


or other smaller risks involving later hits.  So making this play costs perhaps
1/36 of a point from this parley and perhaps another 1/36 or 1/72 from getting
hit later when he has even more men off.


What does he get in return?  A mild increase in the gammon chances; he should
be off in 5 rolls [budgeting one double and two misses via aces] wheras he
would have needed 6 rolls playing safe.  If X dances once it's really close;
figure he increases his gammon chances by this table:

  1: X dances.  8/18.  Gammon chances improve perhaps 1/4 as compared with the
     safe play.  Loss chances are perhaps 1/6[blot]*1/3[hit]*1/8[lose] = 1/144

  2: X comes in and stays on the 1.  1-1 1-2 1-3 1-4.  4/18.  Gammon chances
     are a bit less than they would have been with the safe play [which would
     have made X dance] and loss chances are again 1/144.  This is the worst
     case 

  3: X passes by.  Gammon chances improve somewhat [it's a hard gammon even
     with the agressive play].  No risk.

I think O made a fine play.

-dk

Also from the same session, here's an even more
spectacular play by the same player:

24 23 22 21 20 19 | | 18 17 16 15 14 13     +-+
 X  X  O  O  O  O | |  O  O  O              |2|
 X     O  O  O  O | |  O  O                 +-+
 X           O    | |
 X                | |
                  | |
                  | |
                  | |
       X  X       | |
 X     X  X  X    | |
 X  O  X  X  X    | |
 1  2  3  4  5  6 | |  7  8  9 10 11 12

Money game:  X to play 3-1

X played 5/2* 24/23 !?  Any comments?


-- Gerry Tesauro (tesauro@watson.ibm.com)

In article <C0GG9n.B90.1@cs.cmu.edu>, tesauro@watson.ibm.com (Gerald Tesauro (8-863-7682)) writes:
|> Also from the same session, here's an even more
|> spectacular play by the same player:
|> 
|> 24 23 22 21 20 19 | | 18 17 16 15 14 13     +-+
|>  X  X  O  O  O  O | |  O  O  O              |2|
|>  X     O  O  O  O | |  O  O                 +-+
|>  X           O    | |
|>  X                | |
|>                   | |
|>                   | |
|>                   | |
|>        X  X       | |
|>  X     X  X  X    | |
|>  X  O  X  X  X    | |
|>  1  2  3  4  5  6 | |  7  8  9 10 11 12
|> 
|> Money game:  X to play 3-1
|> 
|> X played 5/2* 24/23 !?  Any comments?

I would never have looked at this play, but it makes sense.  O is
playing for a gammon, and having extra checkers back certainly doesn't
hurt X at this point.  On the other hand, stacking on the 1-point
doesn't help.  X is unlikely to get off the gammon unless O is hit in
the bear-off and re-hit again and again while X brings the back
checkers home.  

5/2* is likely to recirculate a checker which would otherwise be put
out of play.  If X is lucky, he might even be stuck on the bar for a
few rolls.

- Phil

In article <C0GG9n.B90.1@cs.cmu.edu> tesauro@watson.ibm.com (Gerald Tesauro (8-863-7682)) writes:
>Also from the same session, here's an even more
>spectacular play by the same player:
>
>24 23 22 21 20 19 | | 18 17 16 15 14 13     +-+
> X  X  O  O  O  O | |  O  O  O              |2|
> X     O  O  O  O | |  O  O                 +-+
> X           O    | |
> X                | |
>                  | |
>                  | |
>                  | |
>       X  X       | |
> X     X  X  X    | |
> X  O  X  X  X    | |
> 1  2  3  4  5  6 | |  7  8  9 10 11 12
>
>Money game:  X to play 3-1
>
>X played 5/2* 24/23 !?  Any comments?
>

He needs to make the 23 point for two reasons:

  - the back game is a slim but finite reed

  - if he's gonna get off the gammon he needs the 23 point as his launch pad --
    five checkers on the 24 point is gammon-bait.

He is greatly overtimed [too fast] for a back game; he hopes to do some serious
R&R on the bar.

I agree with the move.

-dk

tesauro@watson.ibm.com (Gerald Tesauro (8-863-7682)) writes:

>24 23 22 21 20 19 | | 18 17 16 15 14 13     +-+
> X  X  O  O  O  O | |  O  O  O              |2|
> X     O  O  O  O | |  O  O                 +-+
> X           O    | |
> X                | |
>                  | |
>                  | |
>                  | |
>       X  X       | |
> X     X  X  X    | |
> X  O  X  X  X    | |
> 1  2  3  4  5  6 | |  7  8  9 10 11 12

>Money game:  X to play 3-1

>X played 5/2* 24/23 !?  Any comments?


Classic back-game tactic: secure 1/3 or 1/2 (i.e., 24/22 or 24/23) points
and delay, delay, delay.... an additional minor factor is that X has the
cube, gives him some (admittedly unlikely) power... worst outcome for X is
if 0 misses him from the bar... then X is still way too advanced.

-- 

Mark Bisgeier              mdb@netcom.com

        ...Let your mind wander... 

tesauro@watson.ibm.com (Gerald Tesauro (8-863-7682)) writes:

>Also from the same session, here's an even more
>spectacular play by the same player:

>24 23 22 21 20 19 | | 18 17 16 15 14 13     +-+
> X  X  O  O  O  O | |  O  O  O              |2|
> X     O  O  O  O | |  O  O                 +-+
> X           O    | |
> X                | |
>                  | |
>                  | |
>                  | |
>       X  X       | |
> X     X  X  X    | |
> X  O  X  X  X    | |
> 1  2  3  4  5  6 | |  7  8  9 10 11 12

>Money game:  X to play 3-1

>X played 5/2* 24/23 !?  Any comments?

Basically I think it is the only move to play.  The 24/23 is a necessity
because even though it is not too likely that O will make the 23 point,
X will come in on a 5/9 chance which is ok but if he doesn't hit the 23
point (11/36 chance of hitting) then O needs to roll a 6 or 7 to make
the point (assuming he hit from the spare piece on 20) or whatever.
Having the prime on the 23-17 points is really really bad for X.  As to
what to do with the 3,  5/2 is also the only alternative.  Getting hit
at this point (on 5 or 2 when O comes back in) is fairly irrelevant and
stacking a third piece on the 1 point is pointless.

Just my thoughts
Billy Shields
alias Gorg