         
 
         
 | | White is Player 2
score: 0
pip: 158 | Unlimited Game
Jacoby Beaver | pip: 160
score: 0
Blue is Player 1 |
| | XGID=-a---aE-C--AeD---cAd----aA:0:0:1:51:0:0:3:0:10 | | Blue to play 51 |
I took a look at e = each [bar/20, 6/5*], but rejected it as too loose. All numbers except 2 would give White a return hit somewhere.
Neither was I excited by R = Run [bar/24*, 18/13], thinking it too passive, but I was going to play it nevertheless. Then I noticed the double hit, K = Kill [bar/24*, 6/1*]. I thought K better than R because it unstacks rather than stacking, and it takes away White’s roll of a 6 from the bar. Playing R gives White a hit when he rolls any non-doublet 6 from the bar. Combining these with the tactical opportunities offerred by the double hit, I figured K had to be at least as good as R.
Imagine my surprise when the XGR++ eval put K behind R by 0.014. A truncated GnuBg rollout that I have setup to be very similar to XGR++ put K even further behind, at 0.024.
So I was a bit confused. My analysis seemed OK to me. It’s embarrassing to think you know what’s best, and then be wrong. Knowing you don’t know is far better. When I have announced a low confidence in my judgement, I don’t mind being wrong.
Well, the rollout came to the rescue. K is not so bad after all. In the 62k rollout below, R and K are too close to call.
62S-52X-31 is an interesting variant. K beats R by a mile. That’s because hitting on the 3pt is better than hitting on the 1pt, and also because running to the 15pt is not as good as running to the midpoint. I am posting the rollout below.
Mike
62S-54X-51 — Unlimited Games — 3-ply XG2 Rollout
| | White is Player 2
score: 0
pip: 158 | Unlimited Game
Jacoby Beaver | pip: 160
score: 0
Blue is Player 1 |
| | XGID=-a---aE-C--AeD---cAd----aA:0:0:1:51:0:0:3:0:10 | | Blue to play 51 |
| 1. | Rollout1 | Bar/24* 18/13 | eq: +0.1663 | | Player:
Opponent: | 53.85% (G:14.41% B:0.71%)
46.15% (G:11.41% B:0.42%) | Conf.: ±0.0024 (+0.1639...+0.1687) - [60.0%]
Duration: 10 hours 02 minutes |
| | 2. | Rollout1 | Bar/24* 6/1* | eq: +0.1659 (-0.0004) | | Player:
Opponent: | 53.65% (G:15.50% B:0.55%)
46.35% (G:11.82% B:0.50%) | Conf.: ±0.0025 (+0.1634...+0.1684) - [40.0%]
Duration: 10 hours 48 minutes |
| | 3. | Rollout2 | Bar/20 6/5* | eq: +0.1193 (-0.0470) | | Player:
Opponent: | 53.45% (G:14.33% B:0.72%)
46.55% (G:14.10% B:0.60%) | Conf.: ±0.0097 (+0.1096...+0.1291) - [0.0%]
Duration: 41 minutes 42 seconds |
| | 4. | Rollout2 | Bar/24* 13/8 | eq: +0.1190 (-0.0473) | | Player:
Opponent: | 52.78% (G:14.33% B:0.66%)
47.22% (G:11.85% B:0.50%) | Conf.: ±0.0090 (+0.1101...+0.1280) - [0.0%]
Duration: 41 minutes 47 seconds |
| | 5. | Rollout2 | Bar/24* 11/6 | eq: +0.0680 (-0.0983) | | Player:
Opponent: | 51.35% (G:14.36% B:0.66%)
48.65% (G:12.35% B:0.47%) | Conf.: ±0.0088 (+0.0592...+0.0768) - [0.0%]
Duration: 41 minutes 10 seconds |
| | | 1 62208 Games rolled with Variance Reduction.
Dice Seed: 65942463
Moves: 3-ply, cube decisions: XG Roller
2 5184 Games rolled with Variance Reduction.
Dice Seed: 65942463
Moves: 3-ply, cube decisions: XG Roller
|
Rollout by Taper_Mike
2012-Dec-14
eXtreme Gammon Version: 2.03
62S-54X-51-tm.xgp
62S-54X-51 [R=K e47 Z47 C98] "<=62*5
|
eXtreme Gammon Version: 2.03
62S-54X-51 — Unlimited Games — XGR++ Eval
| | White is Player 2
score: 0
pip: 158 | Unlimited Game
Jacoby Beaver | pip: 160
score: 0
Blue is Player 1 |
| | XGID=-a---aE-C--AeD---cAd----aA:0:0:1:51:0:0:3:0:10 | | Blue to play 51 |
| 1. | XG Roller++ | Bar/24* 18/13 | eq: +0.1700 | | Player:
Opponent: | 54.10% (G:14.14% B:0.64%)
45.90% (G:10.89% B:0.37%) | |
| | 2. | XG Roller++ | Bar/24* 6/1* | eq: +0.1565 (-0.0135) | | Player:
Opponent: | 53.68% (G:14.87% B:0.49%)
46.32% (G:11.24% B:0.46%) | |
| | 3. | XG Roller++ | Bar/24* 13/8 | eq: +0.1227 (-0.0473) | | Player:
Opponent: | 52.92% (G:14.13% B:0.57%)
47.08% (G:11.27% B:0.43%) | |
| | 4. | XG Roller++ | Bar/20 6/5* | eq: +0.1205 (-0.0495) | | Player:
Opponent: | 53.36% (G:13.90% B:0.66%)
46.64% (G:13.38% B:0.54%) | |
| | 5. | XG Roller++ | Bar/24* 11/6 | eq: +0.0783 (-0.0917) | | Player:
Opponent: | 51.71% (G:14.07% B:0.58%)
48.29% (G:11.94% B:0.44%) | |
|
eXtreme Gammon Version: 2.03
62S-52X-31 — Unlimited Games — 3-ply XG2 Rollout
| | White is Player 2
score: 0
pip: 160 | Unlimited Game
Jacoby Beaver | pip: 160
score: 0
Blue is Player 1 |
| | XGID=-a-a--E-C--AeD---cAd----aA:0:0:1:31:0:0:3:0:10 | | Blue to play 31 |
| 1. | Rollout1 | Bar/24* 6/3* | eq: +0.2972 | | Player:
Opponent: | 56.58% (G:17.52% B:0.99%)
43.42% (G:10.74% B:0.49%) | Conf.: ±0.0085 (+0.2887...+0.3057) - [100.0%]
Duration: 1 hour 30 minutes |
| | 2. | Rollout1 | Bar/24* 18/15 | eq: +0.2248 (-0.0724) | | Player:
Opponent: | 55.09% (G:15.79% B:0.81%)
44.91% (G:11.29% B:0.49%) | Conf.: ±0.0089 (+0.2159...+0.2337) - [0.0%]
Duration: 1 hour 32 minutes |
| | | 1 5184 Games rolled with Variance Reduction.
Dice Seed: 60710437
Moves: 3-ply, cube decisions: XG Roller
|
Rollout by Taper_Mike
2012-Sep-16
eXtreme Gammon Version: 2.03
62S-52X-31-tm.xgp
62S-52X-31 [K R72] "<=5
Ballard-Weaver Diagram A59 |
eXtreme Gammon Version: 2.03
|