Problem 6
Some might be tricked by the score into hitting, but hitting can't be
better than making a four-point board. What I found surprising was that
downgrading the 4pt to the 3pt and keeping the 8pt came out ahead of
making the solid four-prime. Generally, a solid prime is much better
than a broken one. However, here White has a just a single checker back
and we are not really trying to prime it; we're hoping to attack it by
hitting it next turn. The 8pt-6pt-5pt-3pt configuration is better than
the four-prime for this purpose; if we make the solid four-prime, then
White hops out giving us a single direct shot from the midpoint with 65
63 62 61 51 and escapes entirely with 66 64, while if we play 6/5(2)
4/3(2), then White hops out with 62 54 53 52 and escapes entirely only
with 55.
In Variant 1 I've put another White checker back and now the
solid four-prime wins. In Variant 2 I've switched to a money-game;
6/5(2) 4/3(2) still comes out ahead but by a smaller margin (if you
permit EMG comparisons across different scores).
          
 
         
 | | White is Player 1
score: 1
pip: 138 | 2 point match
Crawford | pip: 125
score: 0
Blue is Player 2 |
|
| XGID=-Ba-B-D-B---eC---bBd-ab---:0:0:1:11:0:1:1:2:10 |
| Blue to play 11 |
| 1. | Rollout1 | 6/5(2) 4/3(2) | eq: +0.862 |
| Player:
Opponent: | 64.63% (G:28.47% B:1.03%)
35.37% (G:9.16% B:0.89%) | Conf: ± 0.010 (+0.852...+0.872)
Duration: 1 hour 29 minutes |
|
| 2. | Rollout1 | 8/5 6/5 | eq: +0.751 (-0.112) |
| Player:
Opponent: | 61.08% (G:26.44% B:1.03%)
38.92% (G:8.80% B:0.78%) | Conf: ± 0.010 (+0.741...+0.761)
Duration: 1 hour 36 minutes |
|
| 3. | Rollout1 | 8/7(2) 6/5(2) | eq: +0.750 (-0.112) |
| Player:
Opponent: | 63.27% (G:24.24% B:0.87%)
36.73% (G:7.80% B:0.66%) | Conf: ± 0.010 (+0.740...+0.760)
Duration: 1 hour 28 minutes |
|
| 4. | Rollout1 | 6/2* | eq: +0.713 (-0.149) |
| Player:
Opponent: | 61.52% (G:24.13% B:0.91%)
38.48% (G:10.15% B:1.12%) | Conf: ± 0.010 (+0.703...+0.723)
Duration: 1 hour 50 minutes |
|
| 5. | Rollout1 | 4/2*(2) | eq: +0.644 (-0.218) |
| Player:
Opponent: | 61.55% (G:20.67% B:0.81%)
38.45% (G:7.60% B:0.96%) | Conf: ± 0.009 (+0.635...+0.653)
Duration: 1 hour 37 minutes |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Variant 1:
 
| | White is Player 1
score: 1
pip: 148 | 2 point match
Crawford | pip: 125
score: 0
Blue is Player 2 |
|
| XGID=-Bb-B-D-B---dC---bBd-ab---:0:0:1:11:0:1:1:2:10 |
| Blue to play 11 |
| 1. | Rollout1 | 8/7(2) 6/5(2) | eq: +0.547 |
| Player:
Opponent: | 65.57% (G:11.76% B:1.21%)
34.43% (G:6.97% B:0.58%) | Conf: ± 0.007 (+0.540...+0.554)
Duration: 3 hours 12 minutes |
|
| 2. | Rollout1 | 6/5(2) 4/3(2) | eq: +0.522 (-0.025) |
| Player:
Opponent: | 63.73% (G:12.37% B:1.24%)
36.27% (G:8.55% B:0.74%) | Conf: ± 0.007 (+0.515...+0.529)
Duration: 3 hours 06 minutes |
|
| 3. | Rollout1 | 8/5 6/5 | eq: +0.521 (-0.026) |
| Player:
Opponent: | 62.09% (G:13.96% B:1.33%)
37.91% (G:8.56% B:0.53%) | Conf: ± 0.007 (+0.514...+0.528)
Duration: 3 hours 23 minutes |
|
| |
1 2592 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Variant 2:
| | White is Player 1
score: 0
pip: 138 | Money session
Jacoby Beaver | pip: 125
score: 0
Blue is Player 2 |
|
| XGID=-Ba-B-D-B---eC---bBd-ab---:1:-1:1:11:0:0:3:0:10 |
| Blue to play 11 |
| 1. | Rollout1 | 6/5(2) 4/3(2) | eq: +0.362 |
| Player:
Opponent: | 64.74% (G:27.88% B:0.45%)
35.26% (G:8.44% B:0.45%) | Conf: ± 0.008 (+0.354...+0.370)
Duration: 2 hours 39 minutes |
|
| 2. | Rollout1 | 8/7(2) 6/5(2) | eq: +0.312 (-0.049) |
| Player:
Opponent: | 63.77% (G:23.49% B:0.28%)
36.23% (G:6.93% B:0.34%) | Conf: ± 0.008 (+0.304...+0.320)
Duration: 2 hours 45 minutes |
|
| 3. | Rollout2 | 8/5 6/5 | eq: +0.253 (-0.109) |
| Player:
Opponent: | 61.10% (G:26.41% B:0.26%)
38.90% (G:8.04% B:0.41%) | Conf: ± 0.019 (+0.234...+0.272)
Duration: 38 minutes 52 seconds |
|
| 4. | Rollout2 | 6/2* | eq: +0.244 (-0.117) |
| Player:
Opponent: | 61.08% (G:24.91% B:0.35%)
38.92% (G:8.66% B:0.32%) | Conf: ± 0.020 (+0.224...+0.264)
Duration: 45 minutes 17 seconds |
|
| 5. | Rollout2 | 4/2*(2) | eq: +0.227 (-0.134) |
| Player:
Opponent: | 61.18% (G:22.29% B:0.25%)
38.82% (G:8.04% B:0.33%) | Conf: ± 0.014 (+0.213...+0.241)
Duration: 39 minutes 18 seconds |
|
| |
1 2592 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
2 571 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21
Problem 7
At some point during a blitz, the "local ammunition" starts to run
low and we face a tricky decision. Do we keep rushing the nearest
available ammunition forwards to try to continue attacking, at the risk
of overextending ourselves? Or do we start consolidating by moving our
rear checkers into a better position so that we retain a strong
advantage even if the blitz fails?
Here it is tempting to keep trying to close out the checker by
playing 13/9, but according to the rollout, anything but 20/16 18/16 is a
whopper. I can't say I completely understand why 20/16 18/16 rolls out
so much better than the alternatives. Every plausible variant I tried
yielded more-or-less the same story, so it's not a subtle feature of
this particular arrangement of checkers that is to blame.
 
 
| | White is Player 1
score: 0
pip: 135 | Money session
Jacoby Beaver | pip: 112
score: 0
Blue is Player 2 |
|
| XGID=aBBB-BBA----bAb--dAdB-b---:1:-1:1:42:0:0:3:0:10 |
| Blue to play 42 |
| 1. | Rollout1 | 20/16 18/16 | eq: +0.860 |
| Player:
Opponent: | 77.88% (G:42.81% B:0.28%)
22.12% (G:3.94% B:0.25%) | Conf: ± 0.007 (+0.853...+0.867)
Duration: 1 hour 50 minutes |
|
| 2. | Rollout1 | 20/16 7/5 | eq: +0.761 (-0.099) |
| Player:
Opponent: | 74.91% (G:41.13% B:0.31%)
25.09% (G:5.25% B:0.35%) | Conf: ± 0.007 (+0.754...+0.768)
Duration: 1 hour 26 minutes |
|
| 3. | Rollout1 | 20/18 20/16 | eq: +0.736 (-0.124) |
| Player:
Opponent: | 74.34% (G:39.89% B:0.20%)
25.66% (G:4.92% B:0.31%) | Conf: ± 0.007 (+0.729...+0.743)
Duration: 1 hour 21 minutes |
|
| 4. | Rollout1 | 20/16 13/11 | eq: +0.725 (-0.135) |
| Player:
Opponent: | 73.94% (G:40.96% B:0.27%)
26.06% (G:5.87% B:0.54%) | Conf: ± 0.007 (+0.718...+0.732)
Duration: 1 hour 13 minutes |
|
| 5. | Rollout1 | 18/16 13/9 | eq: +0.708 (-0.152) |
| Player:
Opponent: | 72.92% (G:40.21% B:0.21%)
27.08% (G:3.95% B:0.26%) | Conf: ± 0.007 (+0.701...+0.715)
Duration: 1 hour 35 minutes |
|
| 6. | Rollout1 | 13/9 7/5 | eq: +0.610 (-0.250) |
| Player:
Opponent: | 70.55% (G:36.24% B:0.24%)
29.45% (G:4.27% B:0.21%) | Conf: ± 0.007 (+0.603...+0.617)
Duration: 1 hour 27 minutes |
|
| 7. | Rollout1 | 13/7 | eq: +0.588 (-0.272) |
| Player:
Opponent: | 70.44% (G:34.68% B:0.22%)
29.56% (G:4.74% B:0.27%) | Conf: ± 0.008 (+0.580...+0.596)
Duration: 1 hour 26 minutes |
|
| 8. | Rollout1 | 20/18 13/9 | eq: +0.581 (-0.279) |
| Player:
Opponent: | 70.00% (G:35.60% B:0.18%)
30.00% (G:5.29% B:0.31%) | Conf: ± 0.007 (+0.574...+0.588)
Duration: 1 hour 11 minutes |
|
| |
1 2592 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21
Problem 8
I threw this one in for anti-QF purposes. It's gammon save so the
natural thing to do is to play safe, and the top play of the rollout is
to play safe. (Clearing the 7pt is better than clearing the 8pt because
we might still want to attack next turn.) However, see the variant
position, where the bold hit comes out on top. Since hitting wins a lot
more in both positions, one still has to weigh wins versus gammons lost
to find the right play.
| | White is Player 1
score: 0
pip: 83 | 2 point match
Crawford | pip: 91
score: 1
Blue is Player 2 |
|
| XGID=-A-BaBDBB-B----a---dbbbab-:0:0:1:65:1:0:1:2:10 |
| Blue to play 65 |
| 1. | Rollout1 | 7/2 7/1 | eq: +0.119 |
| Player:
Opponent: | 56.87% (G:1.24% B:0.00%)
43.13% (G:0.93% B:0.02%) | Conf: ± 0.003 (+0.116...+0.122)
Duration: 6 minutes 27 seconds |
|
| 2. | Rollout1 | 10/4* 6/1 | eq: +0.064 (-0.055) |
| Player:
Opponent: | 66.46% (G:5.26% B:0.07%)
33.54% (G:13.28% B:0.75%) | Conf: ± 0.006 (+0.058...+0.070)
Duration: 15 minutes 21 seconds |
|
| 3. | Rollout1 | 8/3 8/2 | eq: +0.030 (-0.089) |
| Player:
Opponent: | 52.59% (G:0.86% B:0.01%)
47.41% (G:1.10% B:0.04%) | Conf: ± 0.003 (+0.027...+0.033)
Duration: 5 minutes 24 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Variant:
| | White is Player 1
score: 0
pip: 83 | 2 point match
Crawford | pip: 91
score: 1
Blue is Player 2 |
|
| XGID=-A-BaBDBB-B----a---dbbbab-:0:0:1:25:1:0:1:2:10 |
| Blue to play 25 |
| 1. | Rollout1 | 6/1 6/4* | eq: +0.215 |
| Player:
Opponent: | 68.00% (G:5.27% B:0.10%)
32.00% (G:7.25% B:0.21%) | Conf: ± 0.006 (+0.209...+0.221)
Duration: 19 minutes 13 seconds |
|
| 2. | Rollout1 | 10/8 10/5 | eq: -0.004 (-0.219) |
| Player:
Opponent: | 50.33% (G:1.80% B:0.07%)
49.67% (G:0.53% B:0.01%) | Conf: ± 0.003 (-0.007...-0.001)
Duration: 7 minutes 11 seconds |
|
| 3. | Rollout1 | 6/1 3/1 | eq: -0.012 (-0.227) |
| Player:
Opponent: | 51.05% (G:1.27% B:0.01%)
48.95% (G:1.65% B:0.05%) | Conf: ± 0.004 (-0.016...-0.008)
Duration: 11 minutes 20 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Problem 9
This is a ticklish decision since we haven't gained enough pips to
just win the race, and we're forced to play a lot of pips and can't hold
on to all our defensive assets. The 6-away point is usually very
important for containment (recall Problem 6 above), but our rearmost
checker is also valuable, and there's no good way to hold on to both. I
decided to hold on to the 6-away point, but that is apparently a
whopper. Part of what's going on is that uprooting the 6-away point
still gives us the 4pt as compensation, which will be very valuable if
we subsequently hit White's blot.
In the variant position, I've moved the dilly builder from the
2pt to the 1pt. Now playing 6/1 gains in value and the balance among the
plays shifts.
  
| | White is Player 2
score: 0
pip: 72 | | 1 point match | pip: 104
score: 0
Blue is Player 1 |
|
| XGID=--Ca-BCB-BA-A---A--bbdbbb-:0:0:1:55:0:0:0:1:10 |
| Blue to play 55 |
| 1. | Rollout1 | 12/7 10/5 9/4(2) | eq: -0.058 |
| Player:
Opponent: | 47.10% (G:1.98% B:0.03%)
52.90% (G:1.59% B:0.01%) | Conf: ± 0.003 (-0.061...-0.055)
Duration: 1 hour 46 minutes |
|
| 2. | Rollout1 | 10/5 9/4(2) 6/1 | eq: -0.118 (-0.060) |
| Player:
Opponent: | 44.08% (G:1.57% B:0.01%)
55.92% (G:7.59% B:0.06%) | Conf: ± 0.003 (-0.121...-0.115)
Duration: 1 hour 31 minutes |
|
| 3. | Rollout1 | 16/1 10/5 | eq: -0.125 (-0.067) |
| Player:
Opponent: | 43.73% (G:1.22% B:0.07%)
56.27% (G:5.44% B:0.06%) | Conf: ± 0.003 (-0.128...-0.122)
Duration: 1 hour 21 minutes |
|
| 4. | Rollout1 | 16/6 12/7 10/5 | eq: -0.127 (-0.069) |
| Player:
Opponent: | 43.64% (G:1.33% B:0.02%)
56.36% (G:1.61% B:0.01%) | Conf: ± 0.003 (-0.130...-0.124)
Duration: 1 hour 06 minutes |
|
| 5. | Rollout1 | 16/11 12/2 10/5 | eq: -0.154 (-0.096) |
| Player:
Opponent: | 42.28% (G:1.16% B:0.01%)
57.72% (G:3.56% B:0.01%) | Conf: ± 0.003 (-0.157...-0.151)
Duration: 1 hour 19 minutes |
|
| 6. | Rollout1 | 12/2 9/4(2) | eq: -0.158 (-0.100) |
| Player:
Opponent: | 42.09% (G:1.22% B:0.01%)
57.91% (G:4.93% B:0.03%) | Conf: ± 0.003 (-0.161...-0.155)
Duration: 1 hour 30 minutes |
|
| 7. | Rollout1 | 16/11 10/5 9/4(2) | eq: -0.159 (-0.101) |
| Player:
Opponent: | 42.04% (G:1.67% B:0.03%)
57.96% (G:4.31% B:0.01%) | Conf: ± 0.003 (-0.162...-0.156)
Duration: 56 minutes 47 seconds |
|
| 8. | Rollout1 | 12/7 9/4(2) 6/1 | eq: -0.196 (-0.138) |
| Player:
Opponent: | 40.22% (G:1.28% B:0.02%)
59.78% (G:6.85% B:0.05%) | Conf: ± 0.003 (-0.199...-0.193)
Duration: 1 hour 11 minutes |
|
| 9. | Rollout1 | 16/1 6/1 | eq: -0.200 (-0.142) |
| Player:
Opponent: | 39.98% (G:1.14% B:0.01%)
60.02% (G:8.87% B:0.05%) | Conf: ± 0.003 (-0.203...-0.197)
Duration: 1 hour 24 minutes |
|
| 10. | Rollout1 | 12/2 10/5 6/1 | eq: -0.203 (-0.145) |
| Player:
Opponent: | 39.86% (G:0.84% B:0.00%)
60.14% (G:4.74% B:0.04%) | Conf: ± 0.004 (-0.207...-0.199)
Duration: 1 hour 59 minutes |
|
| 11. | Rollout1 | 16/11 12/7 10/5 6/1 | eq: -0.204 (-0.146) |
| Player:
Opponent: | 39.79% (G:1.61% B:0.01%)
60.21% (G:5.13% B:0.02%) | Conf: ± 0.003 (-0.207...-0.201)
Duration: 1 hour 12 minutes |
|
| 12. | Rollout1 | 16/1 12/7 | eq: -0.215 (-0.157) |
| Player:
Opponent: | 39.23% (G:1.31% B:0.01%)
60.77% (G:5.99% B:0.01%) | Conf: ± 0.003 (-0.218...-0.212)
Duration: 55 minutes 47 seconds |
|
| 13. | Rollout1 | 16/6 9/4(2) | eq: -0.236 (-0.178) |
| Player:
Opponent: | 38.21% (G:1.33% B:0.02%)
61.79% (G:4.59% B:0.02%) | Conf: ± 0.003 (-0.239...-0.233)
Duration: 59 minutes 00 second |
|
| 14. | Rollout1 | 16/6 12/2 | eq: -0.242 (-0.184) |
| Player:
Opponent: | 37.92% (G:0.80% B:0.01%)
62.08% (G:4.63% B:0.01%) | Conf: ± 0.003 (-0.245...-0.239)
Duration: 1 hour 04 minutes |
|
| 15. | Rollout1 | 16/11 12/7 9/4(2) | eq: -0.265 (-0.207) |
| Player:
Opponent: | 36.74% (G:1.40% B:0.02%)
63.26% (G:5.35% B:0.01%) | Conf: ± 0.003 (-0.268...-0.262)
Duration: 43 minutes 51 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
First 4 moves and cube decisions: 4 ply
Remaining moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Variant:
| | White is Player 2
score: 0
pip: 72 | | 1 point match | pip: 103
score: 0
Blue is Player 1 |
|
| XGID=-ABa-BCB-BA-A---A--bbdbbb-:0:0:1:55:0:0:0:1:10 |
| Blue to play 55 |
| 1. | Rollout1 | 10/5 9/4(2) 6/1 | eq: +0.006 |
| Player:
Opponent: | 50.30% (G:1.88% B:0.01%)
49.70% (G:5.06% B:0.04%) | Conf: ± 0.003 (+0.003...+0.009)
Duration: 12 minutes 05 seconds |
|
| 2. | Rollout1 | 16/1 10/5 | eq: +0.005 (-0.001) |
| Player:
Opponent: | 50.25% (G:1.52% B:0.01%)
49.75% (G:3.22% B:0.02%) | Conf: ± 0.003 (+0.002...+0.008)
Duration: 10 minutes 16 seconds |
|
| 3. | Rollout1 | 12/7 10/5 9/4(2) | eq: -0.039 (-0.045) |
| Player:
Opponent: | 48.07% (G:2.24% B:0.01%)
51.93% (G:2.19% B:0.01%) | Conf: ± 0.003 (-0.042...-0.036)
Duration: 9 minutes 58 seconds |
|
| 4. | Rollout1 | 12/2 10/5 6/1 | eq: -0.041 (-0.047) |
| Player:
Opponent: | 47.93% (G:1.18% B:0.01%)
52.07% (G:4.81% B:0.03%) | Conf: ± 0.004 (-0.045...-0.037)
Duration: 11 minutes 55 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Problem 10
I try to choose quiz problems whose rollouts I trust and in which
thematically different plays have significantly different equities.
Here, I made the mistake of including this position in the quiz on the
strength of the eXtreme Gammon 1.21 rollout alone, before studying the
position to make sure I trusted the rollout. If I had to do it over
again, I wouldn't have included this position in the quiz.
But I'm getting ahead of myself. The question here is whether to
play safely or boldly, and if boldly, whether to break anchor with 23/18
or to slot with 10/5. Either bold play will win more games and lose
more gammons. At gammon save, the rollout says one should play safe (see
variant below). At 4a4a with a centered cube, however, the rollout says
to play boldly. O.K., but should we play 23/18 or 10/5? Here's what my
rollout says:
  
| | White is Player 1
score: 0
pip: 89 | | 4 point match | pip: 120
score: 0
Blue is Player 2 |
|
| XGID=--BBB-B-CaBa-------dcbdB--:0:0:1:51:0:0:0:4:10 |
| Blue to play 51 |
| 1. | Rollout1 | 23/18 10/9* | eq: +0.075 |
| Player:
Opponent: | 50.57% (G:15.03% B:0.84%)
49.43% (G:23.50% B:2.04%) | Conf: ± 0.028 (+0.047...+0.103)
Duration: 2 hours 49 minutes |
|
| 2. | Rollout1 | 10/5 10/9* | eq: -0.101 (-0.176) |
| Player:
Opponent: | 48.05% (G:12.26% B:0.51%)
51.95% (G:19.45% B:1.34%) | Conf: ± 0.020 (-0.121...-0.081)
Duration: 2 hours 13 minutes |
|
| 3. | Rollout1 | 10/9* 8/3 | eq: -0.112 (-0.187) |
| Player:
Opponent: | 46.30% (G:10.73% B:0.46%)
53.70% (G:15.35% B:1.12%) | Conf: ± 0.018 (-0.130...-0.094)
Duration: 2 hours 37 minutes |
|
| 4. | Rollout1 | 10/9* 9/4 | eq: -0.184 (-0.259) |
| Player:
Opponent: | 43.37% (G:9.90% B:0.41%)
56.63% (G:13.75% B:0.87%) | Conf: ± 0.018 (-0.202...-0.166)
Duration: 2 hours 22 minutes |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
First 4 moves and cube decisions: 4 ply
Remaining moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
23/18 is right by a mile, apparently. But hold on, is this really
right? I looked a bit further to see what the cube action would be after
10/9* 10/5, dance. It's intuitively clear that this should be a massive
cube. However, XG1 doesn't bother looking further than its 1-ply
evaluation, which says it's a massive ND! I'm pretty sure that even
though the rollout parameters are telling XG to use "4-ply cube" that
it's not actually doing a 4-ply evaluation of that cube decision—in
which case the rollout results are rubbish. I asked a friend of mine who
owns XG2 to roll this position out, and he sent me the following
results:
| 1. | Rollout1 | 23/18 10/9* | eq: +0.088 |
| Player:
Opponent: | 50.83% (G:15.09% B:0.84%)
49.17% (G:23.69% B:1.70%) | Conf.: ± 0.008 (+0.080...+0.096) - [100.0%]
Duration: 19 minutes 24 seconds |
|
| 2. | Rollout1 | 10/5 10/9* | eq: +0.053 (-0.035) |
| Player:
Opponent: | 48.54% (G:11.41% B:0.37%)
51.46% (G:19.43% B:1.26%) | Conf.: ± 0.005 (+0.048...+0.057) - [0.0%]
Duration: 14 minutes 31 seconds |
|
| 3. | Rollout2 | 10/9* 8/3 | eq: -0.112 (-0.201) |
| Player:
Opponent: | 46.30% (G:10.73% B:0.46%)
53.70% (G:15.35% B:1.12%) | Conf.: ± 0.018 (-0.130...-0.095) - [0.0%]
Duration: 2 hours 37 minutes |
|
| 4. | Rollout2 | 10/9* 9/4 | eq: -0.184 (-0.272) |
| Player:
Opponent: | 43.37% (G:9.90% B:0.41%)
56.63% (G:13.75% B:0.87%) | Conf.: ± 0.018 (-0.202...-0.166) - [0.0%]
Duration: 2 hours 22 minutes |
|
| |
1 5184 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller
2 1296 Games rolled with Variance Reduction.
Dice Seed: 2
First 4 moves and cube decisions: 4-ply
Remaining moves and cube decisions: 3-ply
Using version 1
|
eXtreme Gammon Version: 2.00, MET: Kazaross XG2
This now looks more believable. I decided to roll this out with GNU as well (with an older MET, g11 I think):
| • |
1 | R | 10/9* 10/5 | +0.118 |
| | | 0.491 | 0.134 | 0.011 | - | 0.509 | 0.203 | 0.021 | -0.102 | +0.118 |
| 0.001 | 0.002 | 0.001 | - | 0.001 | 0.002 | 0.002 | 0.004 | 0.009 | | |
| | Full cubeful rollout with var.redn. | |
| | 2592 games, Mersenne Twister dice gen. with seed 927483677 and quasi-random dice | |
| | Play: world class 2-ply cubeful prune [world class] | |
| | keep the first 0 0-ply moves and up to 8 more moves within equity 0.16 | |
| | Skip pruning for 1-ply moves. | |
| | Cube: 2-ply cubeful prune [world class] | |
| | 2 | R | 23/18 10/9* | +0.100 ( -0.018) |
| | | 0.510 | 0.152 | 0.015 | - | 0.490 | 0.235 | 0.019 | -0.085 | +0.100 |
| 0.001 | 0.001 | 0.002 | - | 0.001 | 0.002 | 0.001 | 0.004 | 0.008 | | |
| | Full cubeful rollout with var.redn. | |
| | 2592 games, Mersenne Twister dice gen. with seed 927483677 and quasi-random dice | |
| | Play: world class 2-ply cubeful prune [world class] | |
| | keep the first 0 0-ply moves and up to 8 more moves within equity 0.16 | |
| | Skip pruning for 1-ply moves. | |
| | Cube: 2-ply cubeful prune [world class] | |
Now you can see why this position fails my criteria for a good quiz problem. But it's still an interesting position!
Variant (gammon save):
| | White is Player 1
score: 0
pip: 89 | 2 point match
Crawford | pip: 120
score: 1
Blue is Player 2 |
|
| XGID=--BBB-B-CaBa-------dcbdB--:0:0:1:51:1:0:1:2:10 |
| Blue to play 51 |
| 1. | Rollout1 | 10/9* 8/3 | eq: -0.349 |
| Player:
Opponent: | 46.12% (G:8.26% B:0.35%)
53.88% (G:13.57% B:1.46%) | Conf: ± 0.008 (-0.357...-0.341)
Duration: 41 minutes 03 seconds |
|
| 2. | Rollout1 | 10/9* 9/4 | eq: -0.385 (-0.036) |
| Player:
Opponent: | 43.28% (G:7.33% B:0.27%)
56.72% (G:12.54% B:1.04%) | Conf: ± 0.009 (-0.394...-0.376)
Duration: 36 minutes 22 seconds |
|
| 3. | Rollout1 | 10/5 10/9* | eq: -0.393 (-0.044) |
| Player:
Opponent: | 48.65% (G:9.73% B:0.42%)
51.35% (G:18.29% B:3.42%) | Conf: ± 0.009 (-0.402...-0.384)
Duration: 43 minutes 05 seconds |
|
| 4. | Rollout1 | 23/18 10/9* | eq: -0.422 (-0.073) |
| Player:
Opponent: | 51.40% (G:12.05% B:0.70%)
48.60% (G:22.49% B:2.70%) | Conf: ± 0.011 (-0.433...-0.411)
Duration: 39 minutes 48 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
|
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross