| 1. | Rollout1 | Bar/21 22/16 | eq: -0.166 |
| Player: Opponent: | 48.72% (G:21.36% B:2.66%) 51.28% (G:14.43% B:0.99%) | Conf.: ± 0.017 (-0.184...-0.149) - [100.0%] Duration: 23 minutes 47 seconds |
|
| 2. | Rollout1 | Bar/15 | eq: -0.211 (-0.045) |
| Player: Opponent: | 47.31% (G:21.27% B:2.44%) 52.69% (G:15.34% B:1.05%) | Conf.: ± 0.014 (-0.225...-0.197) - [0.0%] Duration: 22 minutes 10 seconds |
|
| 3. | XG Roller++ | Bar/21 9/3 | eq: -0.245 (-0.079) |
| Player: Opponent: | 44.89% (G:19.51% B:1.96%) 55.11% (G:14.45% B:0.69%) | |
|
| 4. | XG Roller++ | Bar/21 8/2 | eq: -0.276 (-0.110) |
| Player: Opponent: | 43.97% (G:19.17% B:1.99%) 56.03% (G:14.91% B:0.73%) | |
|
| 5. | XG Roller++ | Bar/21 13/7 | eq: -0.470 (-0.304) |
| Player: Opponent: | 39.23% (G:16.70% B:1.75%) 60.77% (G:19.98% B:1.54%) | |
|
| |
1 1296 Games rolled with Variance Reduction. Moves and cube decisions: 3-ply
|
Seems to me b/15 should be best. This duplicates W's 3s and 1s, making it harder for him to both hit and make his 5pt. This turns out to be wrong, and not by a small amount.