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 Historical content remains viewable.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

play 52

16 views
Skip to first unread message

badgolferman

unread,
May 21, 2013, 2:26:50 PM5/21/13
to
GNU Backgammon Position ID: ze0GCAS32QwADA
Match ID : UYnqADAAIAAE
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
| O | | O O O X O O | 3 points
| O | | O O O X O |
| | | O O |
| | | |
| | | |
v| |BAR| | 7 point match
| | | |
| | | |
| | | X |
| X | | X X X X X | Rolled 52
| O X | | X X O X X X | 4 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: badgolferman (Cube:
2)

Pips X=95 O=95

Walt

unread,
May 21, 2013, 3:11:47 PM5/21/13
to
I don't like the idea of sitting on the 3 point waiting to lose, and O's
homeboard blot gives me an excuse to run. Is this sound? After 22/15,
any 5 is bad, so is 11, 12, 13, 14, 34. (doubles are good no matter
what). That's 24 rolls that play very well for O. The thing is, X
will be forced off the anchor fairly soon anyway, so might as well do it
now while there's a blot to shoot at and X's board is strong.

22/15 for me.

--
//Walt

Freeven

unread,
May 21, 2013, 4:07:14 PM5/21/13
to
On 5/21/2013 2:26 PM, badgolferman wrote:
My initial impression is that X can't afford to play 8/6 8/3. This takes
all the pressure off of O, as X is out of time and O can simply park in
the outfield and wait. I'm inclined to try to put pressure on O to
perform. Just how to do that is the question.

22/15?

22/17 6/4*?

6/4* 6/1?

I've no clue. At dmp I'd play 22/17 6/4*. This forces O to perform and
directly addresses X's problems of timing and outfield control. With
gammons costing the match, however, giving up the anchor and leaving 4
blots is a scary proposition. The thing is that the other plays leave X
plenty vulnerable as well -- which is leading me back to 8/6 8/3. I
don't like it, but that's my play.

Bradley K. Sherman

unread,
May 21, 2013, 6:22:35 PM5/21/13
to
badgolferman <REMOVETHISb...@gmail.com> wrote:
> +13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
> | O | | O O O X O O | 3 points
> | O | | O O O X O |
> | | | O O |
> | | | |
> | | | |
>v| |BAR| | 7 point match
> | | | |
> | | | |
> | | | X |
> | X | | X X X X X | Rolled 52
> | O X | | X X O X X X | 4 points
> +12-11-10--9--8--7-------6--5--4--3--2--1-+ X: bgm (Cube:2)
>Pips X=95 O=95

Good question. I would play
8/3,8/6
but I wouldn't be happy.

--bks

Tim Chow

unread,
May 21, 2013, 7:49:48 PM5/21/13
to
On May 21, 2:26 pm, "badgolferman" <REMOVETHISbadgolfer...@gmail.com>
wrote:

The score indicates that X needs to exercise some caution, but the
safe play of 8/6 8/3 seems to give up too much in the way of winning
chances. I would try running with 22/15 while O still has a blot in
her board and we still have a five-point board.

---
Tim Chow

badgolferman

unread,
May 22, 2013, 2:47:48 PM5/22/13
to
It seems we all agreed either running or hiding would be our play,
admittedly holding our noses while choosing those plays. My choice was
8/6, 8/3 but the bot called that a whopper even though it's the second
best play. It seems hitting is the proper play in this case. If X
runs, O is sure to hit either one of those blots and probably close one
of its home points as well. Then X is in a world of hurt. Might as
well force the situation now while X has an anchor and has the chance
to close the 4-point if O misses. It's a bad choice but better than
the alternatives according to the bot. Still, I'm not sure I could do
that play OTB, still opting for 8/6, 8/3 and taking my chances on the
next roll.




1. Rollout 6/4* 6/1 Eq.: -0.523
0.338 0.083 0.002 - 0.662 0.167 0.015 CL -0.600 CF -0.523
{0.001 0.004 0.001 - 0.001 0.002 0.002 CL 0.004 CF 0.005]

2. Rollout 8/6 8/3 Eq.: -0.624 ( -0.101)
0.273 0.036 0.001 - 0.727 0.135 0.011 CL -0.693 CF -0.624
{0.001 0.003 0.000 - 0.001 0.002 0.002 CL 0.004 CF 0.004]

3. Rollout 8/1 Eq.: -0.735 ( -0.212)
0.232 0.024 0.001 - 0.768 0.141 0.013 CL -0.796 CF -0.735
[0.001 0.002 0.001 - 0.001 0.002 0.002 CL 0.004 CF 0.004]

4. Rollout 22/17 6/4* Eq.: -0.786 ( -0.263)
0.376 0.126 0.007 - 0.624 0.361 0.079 CL -0.865 CF -0.786
[0.001 0.004 0.001 - 0.001 0.002 0.005 CL 0.005 CF 0.005]

5. Rollout 8/3 6/4* Eq.: -0.967 ( -0.444)
0.268 0.076 0.002 - 0.732 0.312 0.047 CL -1.027 CF -0.967
[0.001 0.004 0.001 - 0.001 0.002 0.004 CL 0.005 CF 0.005]

Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 682943018 and
quasi-random dice
Play: supremo 2-ply cubeful prune [world class]
keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
Skip pruning for 1-ply moves.
Cube: 2-ply cubeful prune [world class]

Tim Chow

unread,
May 23, 2013, 10:29:11 PM5/23/13
to
On May 22, 2:47 pm, "badgolferman" <REMOVETHISbadgolfer...@gmail.com>
wrote:
> badgolferman wrote:
> > GNU Backgammon  Position ID: ze0GCAS32QwADA
> >                 Match ID   : UYnqADAAIAAE
> > +13-14-15-16-17-18------19-20-21-22-23-24-+     O: gnubg
> > |                O |   | O  O  O  X  O  O |     3 points
> > |                O |   | O  O  O  X  O    |
> > |                  |   | O     O          |
> > |                  |   |                  |
> > |                  |   |                  |
> >v|                  |BAR|                  |     7 point match
> > |                  |   |                  |
> > |                  |   |                  |
> > |                  |   |                X |
> > |             X    |   | X  X     X  X  X |     Rolled 52
> > |       O     X    |   | X  X  O  X  X  X |     4 points
> > +12-11-10--9--8--7-------6--5--4--3--2--1-+     X: badgolferman
> >(Cube:  2)
>
> >Pips X=95  O=95

This is a totally cool position! The logic of 6/4* 6/1 is clear (hit
to win, safety to reduce gammons) but for psychological reasons I
totally overlooked it as a candidate.

For good measure, I have rolled out the position with XG, after
switching to money play. XG deems my play hellishly bad, with an EMG
equity of -0.666. The DMP play of 22/17 6/4* fares slightly better
than in the original position because gammon losses aren't as painful
for money as they are at 3a4a owning the cube.

XGID=-CBBaBB-B-a-------bcbcBba-:1:1:1:52:0:0:0:0:10
X:Player 1 O:Player 2

Score is X:0 O:0. Money session
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O | | O O O X O O |
| O | | O O O X O |
| | | O O |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X | +---+
| X | | X X X X X | | 2 |
| O X | | X X O X X X | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pipcount X: 95 O: 95 X-O: 0-0
Cube : 2, X own cube
X to play 52

1. Rollout: 6/1 6/4* eq:-0.324
Player : 33.72% (G:5.62% B:0.07%)
Opponent: 66.28% (G:16.42% B:0.56%)
Confidence: ±0.008 (-0.332<E<-0.316)

2. Rollout: 22/17 6/4* eq:-0.365 (-0.041)
Player : 37.92% (G:10.99% B:0.24%)
Opponent: 62.08% (G:36.63% B:3.79%)
Confidence: ±0.015 (-0.380<E<-0.350)

3. Rollout: 8/6 8/3 eq:-0.483 (-0.159)
Player : 26.56% (G:2.30% B:0.05%)
Opponent: 73.44% (G:12.24% B:0.34%)
Confidence: ±0.007 (-0.490<E<-0.476)

4. Rollout: 8/1 eq:-0.584 (-0.260)
Player : 23.06% (G:1.89% B:0.03%)
Opponent: 76.94% (G:14.02% B:0.32%)
Confidence: ±0.007 (-0.591<E<-0.577)

5. Rollout: 8/3 6/4* eq:-0.631 (-0.307)
Player : 27.62% (G:6.02% B:0.08%)
Opponent: 72.38% (G:32.17% B:1.49%)
Confidence: ±0.008 (-0.639<E<-0.623)

6. Rollout: 22/15 eq:-0.666 (-0.342)
Player : 23.89% (G:7.20% B:0.35%)
Opponent: 76.11% (G:27.90% B:0.62%)
Confidence: ±0.010 (-0.676<E<-0.656)

1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply

eXtreme Gammon Version: 1.21

---
Tim Chow
0 new messages