2010/08/31

Last problem of "100 backgammon puzzles" by Paul Lamford 3

Even 4ply rollout doesn't understand proper cube action here. XG 4ply will take bad cube in later stage so that he states it is too good now.
is Player 2

score: 0
pip: 60
                         
Money session
Jacoby Beaver
                          pip: 330
score: 0

is Player 1
XGID=-------------------eEeEeE-:0:0:-1:00:0:0:3:0:10
on roll, cube action?


Analyzed in Rollout
No Double
  Player Winning Chances: 75.60% (G: 54.91% B: 26.59%)
  Opponent Winning Chances: 24.40% (G: 0.01% B: 0.00%)
Double/Take
  Player Winning Chances: 72.41% (G: 60.97% B: 26.96%)
  Opponent Winning Chances: 27.59% (G: 0.14% B: 0.00%)
 
Cubeless Equities
No Double:+0.512
Double:+2.652
 
Cubeful Equities
No Double:+1.016
Double/Take:+2.415 (+1.399)
Double/Drop:+1.000 (-0.016)
 
Best Cube action: Too Good to Double / Drop
 
Rollout details
1296 Games rolled with Variance Reduction.
Dice Seed: 1341
Moves and cube decisions: 4 ply
 
Confidence No Double: ± 0.012 (+1.004...+1.028)
Confidence Double: ± 0.107 (+2.308...+2.522)
 
Double Decision confidence: 99.5%
Take Decision confidence: 100.0%
 
Duration: 3 hours 31 minutes

eXtreme Gammon Version: 1.21

2010/08/30

Last problem of "100 backgammon puzzles" by Paul Lamford 2

I created a new position from Paul's last problem.
XG thinks it is way too good (with jacoby and centered cube???). This is insane.
is Player 2

score: 0
pip: 60
                         
Money session
Jacoby Beaver
                          pip: 263
score: 0

is Player 1
XGID=-------------BBBBAAeAeBeB-:0:0:-1:00:0:0:3:0:10
on roll, cube action?


Analyzed in Rollout
No Double
  Player Winning Chances: 68.10% (G: 53.43% B: 15.73%)
  Opponent Winning Chances: 31.90% (G: 0.06% B: 0.00%)
Double/Take
  Player Winning Chances: 70.07% (G: 56.74% B: 18.08%)
  Opponent Winning Chances: 29.93% (G: 0.13% B: 0.00%)
 
Cubeless Equities
No Double:+0.362
Double:+2.297
 
Cubeful Equities
No Double:+1.490
Double/Take:+2.056 (+0.566)
Double/Drop:+1.000 (-0.490)
 
Best Cube action: Too Good to Double / Drop
 
Rollout details
1296 Games rolled with Variance Reduction.
Dice Seed: 1341
Moves and cube decisions: 3 ply
 
Confidence No Double: ± 0.092 (+1.398...+1.582)
Confidence Double: ± 0.124 (+1.932...+2.181)
 
Double Decision confidence: 100.0%
Take Decision confidence: 100.0%
 
Duration: 43 minutes 15 seconds

eXtreme Gammon Version: 1.21

Last problem of "100 backgammon puzzles" by Paul Lamford

is Player 2

score: 0
pip: 60
                         
Money session
Jacoby Beaver
                          pip: 330
score: 0

is Player 1
XGID=-------------------eEeEeE-:0:0:-1:D:0:0:3:0:10
double to 2 take ?


Analyzed in Rollout
No Double
  Player Winning Chances: 83.98% (G: 56.17% B: 30.61%)
  Opponent Winning Chances: 16.02% (G: 0.01% B: 0.00%)
Double/Take
  Player Winning Chances: 81.40% (G: 69.01% B: 39.53%)
  Opponent Winning Chances: 18.60% (G: 0.14% B: 0.00%)
 
Cubeless Equities
No Double:+0.680
Double:+3.424
 
Cubeful Equities
No Double:+1.024
Double/Take:+3.257 (+2.233)
Double/Drop:+1.000 (-0.024)
 
Best Cube action: Too Good to Double / Drop
 
Rollout details
1296 Games rolled with Variance Reduction.
Dice Seed: 1341
Moves and cube decisions: 3 ply
 
Confidence No Double: ± 0.017 (+1.007...+1.041)
Confidence Double: ± 0.136 (+3.121...+3.392)
 
Double Decision confidence: 99.7%
Take Decision confidence: 100.0%
 
Duration: 20 minutes 42 seconds

eXtreme Gammon Version: 1.21

This is a last problem of "100 backgammon puzzles" written by Paul Lamford. Paul says it is very easy take there fore no-doulbe is proper cube action. XG rollout (3ply full) is strongly against it.
However, I coudn't trust bot's recommendation of "too good to double" with Jacoby rule, centerd cube. Who is wrong, bot or Paul?
I set up XG playing each other to see how bot plays. I understand why it was too good. 3ply bot will take the bad cube later, so that doubler can gain by waiting one more roll.
Also, 3ply can't handle the outside prime("snake") properly. As a result, take side simply wasting TOO MUCH equity in the process.
The position below is ending scene of rollout and good sample of how bad bot plays. Human wouldn't stack 7 checkers on the 18point in this game.
is eXtremeGammon

score: 0
pip: 10
                         
Money session
Jacoby Beaver
                          pip: 242
score: 0

is Masayuki Mochizuki
XGID=---------BC-------G-Aa-cB-:1:1:-1:66:0:0:3:0:10
to play 66

* 1.3 ply4/0ff 2/0ff(3)eq: +3.000
Player:
Opponent:
100.00% (G:100.00% B:100.00%)
0.00% (G:0.00% B:0.00%)

eXtreme Gammon Version: 1.21

is ScottyK

score: 0
pip: 168
                         
11 point match
                          pip: 155
score: 0

is mochy
XGID=--a-aBDaB---dE--Ac-e----A-:1:-1:1:51:0:0:0:11:10
to play 51

1.XG Roller+13/7* eq: +0.189
Player:
Opponent:
61.10% (G:18.18% B:0.91%)
38.90% (G:8.70% B:0.37%)
* 2.XG Roller+8/7* 7/2* eq: +0.145 (-0.044)
Player:
Opponent:
59.43% (G:19.18% B:0.95%)
40.57% (G:9.89% B:0.48%)
3.XG Roller+16/15 13/8eq: +0.139 (-0.050)
Player:
Opponent:
60.07% (G:15.45% B:0.60%)
39.93% (G:8.21% B:0.28%)
4.XG Roller+13/8 6/5eq: +0.128 (-0.061)
Player:
Opponent:
59.80% (G:14.73% B:0.58%)
40.20% (G:7.94% B:0.27%)

eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross

I have too few builders for completing blitz. Even though 13/7 has so many return shot, there are 13 bad numbers. (12, 26, 36, 46, 56, 11, 55, 66) I could get 14 bad numbers after the double hit, but the best play has much better shape.

2010/08/27

is Player 2

score: 0
pip: 180
                         
Money session
                          pip: 74
score: 0

is Player 1
XGID=-BBBbbEcD---------bbbb----:1:1:1:00:0:0:0:0:10
on roll, cube action?


Analyzed in XG Roller+
Player Winning Chances: 66.56% (G: 51.37% B: 0.31%)
Opponent Winning Chances: 33.44% (G: 3.21% B: 0.15%)
 
Cubeless Equities
No Redouble:+0.814
Redouble:+1.539
 
Cubeful Equities
No Redouble:+1.004
Redouble/Take:+1.308 (+0.303)
Redouble/Drop:+1.000 (-0.004)
 
Best Cube action: Too Good to Redouble / Drop
 

eXtreme Gammon Version: 1.21

My opponent (known to be strong player) took this cube with out hesitation. It is a good example of how complicated backgame is. Even an experienced player can be screwed up.

2010/08/26

is Player 2

score: 0
pip: 156
                         
1 point match
                          pip: 128
score: 0

is Player 1
XGID=-b-bB-CBDB--b--A-a-c-abAb-:0:0:-1:42:0:0:0:1:10
to play 42

1.XG Roller+24/20 22/20eq: -0.453
Player:
Opponent:
27.34% (G:5.50% B:0.13%)
72.66% (G:28.65% B:2.19%)
2.XG Roller+24/20 4/2* eq: -0.457 (-0.004)
Player:
Opponent:
27.14% (G:7.34% B:0.17%)
72.86% (G:27.74% B:2.18%)
3.XG Roller+6/2* 4/2eq: -0.464 (-0.011)
Player:
Opponent:
26.79% (G:8.91% B:0.15%)
73.21% (G:25.21% B:1.91%)
4.XG Roller+22/20 6/2* eq: -0.500 (-0.047)
Player:
Opponent:
24.98% (G:6.32% B:0.15%)
75.02% (G:35.45% B:5.32%)
5.3 ply24/20 6/4eq: -0.538 (-0.084)
Player:
Opponent:
24.28% (G:5.04% B:0.16%)
75.72% (G:28.25% B:1.82%)

eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross

2010/08/25

51S-42P-62R-61

is Player 2

score: 0
pip: 153
                         
Money session
Jacoby Beaver
                          pip: 161
score: 0

is Player 1
XGID=--a-B-D-Ba--dE---d-e----B-:0:0:1:16:0:0:3:0:10
to play 16

1.Rollout113/7 6/5eq: -0.006
Player:
Opponent:
49.30% (G:15.57% B:0.58%)
50.70% (G:13.54% B:0.88%)
Conf: ± 0.011 (-0.017...+0.005) Duration: 5 hours 32 minutes
2.Rollout124/23 13/7eq: -0.038 (-0.032)
Player:
Opponent:
48.33% (G:15.28% B:0.56%)
51.67% (G:13.02% B:0.75%)
Conf: ± 0.011 (-0.049...-0.027) Duration: 4 hours 48 minutes
3.Rollout113/6eq: -0.043 (-0.037)
Player:
Opponent:
48.15% (G:15.16% B:0.57%)
51.85% (G:12.84% B:0.71%)
Conf: ± 0.012 (-0.055...-0.031) Duration: 4 hours 20 minutes
4.Rollout113/7 8/7eq: -0.045 (-0.039)
Player:
Opponent:
48.06% (G:15.82% B:0.57%)
51.94% (G:12.91% B:0.75%)
Conf: ± 0.011 (-0.056...-0.034) Duration: 4 hours 44 minutes
5.Rollout124/23 24/18eq: -0.049 (-0.044)
Player:
Opponent:
48.57% (G:12.11% B:0.48%)
51.43% (G:12.02% B:0.50%)
Conf: ± 0.010 (-0.059...-0.039) Duration: 4 hours 45 minutes
 
1 2594 Games rolled with Variance Reduction. Dice Seed: 1341 Moves and cube decisions: 4 ply

eXtreme Gammon Version: 1.21

13/7 is conceptually forced. You have a heavy midpoint screaming to unstack.
You are not leading in the race and your opponent has no board.

As for the ace, it would be a difficult choice for me. At least you will be in good shape if your opponent misses the doubl shot.

2010/08/24

is proflyerfan2

score: 0
pip: 122
                         
11 point match
                          pip: 149
score: 1

is mochy
XGID=-a-Ba-D-B--AbD----cbBbbb--:0:0:1:62:1:0:0:11:10
to play 62

1.XG Roller+11/3eq: -0.335
Player:
Opponent:

40.53% (G:7.55% B:0.20%)

59.47% (G:10.80% B:0.27%)
* 2.XG Roller+13/11 13/7eq: -0.549 (-0.213)
Player:
Opponent:

38.38% (G:8.09% B:0.30%)

61.62% (G:19.56% B:0.54%)
3.3 ply11/5 6/4* eq: -0.624 (-0.289)
Player:
Opponent:

38.13% (G:9.15% B:0.41%)

61.87% (G:27.12% B:0.89%)
4.3 ply13/5eq: -0.625 (-0.289)
Player:
Opponent:

36.58% (G:6.98% B:0.25%)

63.42% (G:21.71% B:0.63%)
5.3 ply13/7 6/4* eq: -0.722 (-0.387)
Player:
Opponent:

37.95% (G:8.93% B:0.39%)

62.05% (G:30.71% B:1.34%)

eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross

is Player 2

score: 8
pip: 9
                         
11 point match
                          pip: 77
score: 10

is Player 1
XGID=-BDE--A-A---------A-A--cc-:1:1:1:43:10:8:0:11:10
to play 43

1.XG Roller+20/16 8/5eq: -2.849
Player:
Opponent:

0.00% (G:0.00% B:0.00%)

100.00% (G:92.47% B:0.00%)
2.XG Roller+18/14 8/5eq: -2.863 (-0.013)
Player:
Opponent:

0.00% (G:0.00% B:0.00%)

100.00% (G:93.13% B:3.10%)
3.XG Roller+20/16 18/15eq: -2.870 (-0.021)
Player:
Opponent:

0.00% (G:0.00% B:0.00%)

100.00% (G:93.51% B:0.00%)
4.XG Roller+18/11eq: -2.878 (-0.029)
Player:
Opponent:

0.00% (G:0.00% B:0.00%)

100.00% (G:93.90% B:4.12%)
5.XG Roller+18/15 8/4eq: -2.886 (-0.037)
Player:
Opponent:

0.00% (G:0.00% B:0.00%)

100.00% (G:94.32% B:2.72%)

eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross

Double four will be better with the best play. (13 savers vs 4 savers) Also, I would be dead if I roll an ace with my play.

English correction: Double fours will play better if you make the best play. (13 plays that then save the gammon v. 4 with the second best play)