This came up in a match vs. XG the other day. I figured that hitting was certainly the DMP play but was afraid it lost too many extra gammons to make it right in unlimited games. A 10K RO seems to agree but it's not stat sig. Hitting wins about 2.7% more games but loses ~ 10% more net gammons.
           
 
         
 | | White is Player 2
score: 0
pip: 140 | Unlimited Game
Jacoby Beaver | pip: 162
score: 0
Blue is Player 1 |
|
| XGID=-a---AD-C---cEaa-c-d-bB---:0:0:1:52:0:0:3:0:10 |
| Blue to play 52 |
| 1. | Rollout1 | 8/3 5/3 | eq: -0.470 |
| Player:
Opponent: | 37.44% (G:7.75% B:0.27%)
62.56% (G:12.46% B:0.48%) | Conf.: ± 0.005 (-0.475...-0.465) - [91.4%]
Duration: 2 hours 05 minutes |
|
| 2. | Rollout1 | 22/15* | eq: -0.475 (-0.005) |
| Player:
Opponent: | 40.16% (G:8.70% B:0.41%)
59.84% (G:23.41% B:1.17%) | Conf.: ± 0.005 (-0.480...-0.470) - [8.6%]
Duration: 2 hours 12 minutes |
|
| |
1 10368 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller
|
eXtreme Gammon Version: 2.10
If you unclick Jacoby, ++ says there's a larger difference between the plays. Makes sense since there are some sequences after 22/15* where White will clearly be TG (i.e. White responds with 33 or 55).
| | White is Player 2
score: 0
pip: 140 | Unlimited Game
Beaver | pip: 162
score: 0
Blue is Player 1 |
|
| XGID=-a---AD-C---cEaa-c-d-bB---:0:0:1:52:0:0:2:0:10 |
| Blue to play 52 |
| 1. | XG Roller++ | 8/3 5/3 | eq: -0.469 |
| Player:
Opponent: | 37.21% (G:7.26% B:0.21%)
62.79% (G:11.75% B:0.39%) | |
|
| 2. | XG Roller++ | 22/15* | eq: -0.499 (-0.031) |
| Player:
Opponent: | 40.00% (G:8.36% B:0.34%)
60.00% (G:23.01% B:1.08%) | |
|
eXtreme Gammon Version: 2.10