The hit is a must. The safe alternative 12/9 11/8(2) 6/3 leaves too many good numbers (6 & 3's make 6-prime and 5's escape, as Timothy pointed out).
I like John's approach to start from DMP evaluation. After any reasonable hitting play (11/2* 6/3, or 6/3 5/2*(2) with either 11/8 or 12/9) we have a 4-pt. board. Any of the entering number is a bad swing. It doesn't matter much if we are hit in the process or not.
From this point on it's easy to see that double-blotting play has many advantages over any of the switches. We have no bad rolls next turn, we keep our best-4-pt. board for a few rolls, we have more containment chances, etc.
This big advantage in straight win % is hard to be outweigh by extra gammon losses, and this position is not an exception, as John pointed out.
1. | Rollout1 | 11/2* 6/3 | eq: -0,651 |
| Player: Opponent: | 27,51% (G:9,04% B:0,34%) 72,49% (G:33,54% B:4,90%) | Conf.: ± 0,007 (-0,658...-0,644) - [100,0%] Duration: 15 minutes 12 seconds |
|
2. | Rollout1 | 12/9 6/3 5/2*(2) | eq: -0,679 (-0,028) |
| Player: Opponent: | 23,80% (G:9,16% B:0,32%) 76,20% (G:28,54% B:3,46%) | Conf.: ± 0,008 (-0,687...-0,671) - [0,0%] Duration: 12 minutes 40 seconds |
|
3. | Rollout1 | 11/8 6/3 5/2*(2) | eq: -0,688 (-0,037) |
| Player: Opponent: | 24,54% (G:9,47% B:0,30%) 75,46% (G:31,03% B:3,89%) | Conf.: ± 0,006 (-0,694...-0,681) - [0,0%] Duration: 12 minutes 46 seconds |
|
4. | Rollout1 | 6/3 5/2*(2) 4/1 | eq: -0,747 (-0,096) |
| Player: Opponent: | 21,01% (G:8,74% B:0,20%) 78,99% (G:28,67% B:2,90%) | Conf.: ± 0,005 (-0,753...-0,742) - [0,0%] Duration: 11 minutes 03 seconds |
|
5. | Rollout1 | 12/9 11/8(2) 6/3 | eq: -0,781 (-0,130) |
| Player: Opponent: | 17,28% (G:2,10% B:0,07%) 82,72% (G:19,31% B:1,31%) | Conf.: ± 0,005 (-0,786...-0,776) - [0,0%] Duration: 9 minutes 44 seconds |
|
|
1 2592 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
|
1. | Rollout1 | 12/9 6/3 5/2*(2) | eq: -0,675 |
| Player: Opponent: | 23,53% (G:9,09% B:0,29%) 76,47% (G:27,72% B:3,42%) | Conf.: ± 0,007 (-0,682...-0,669) - [100,0%] Duration: 11 minutes 14 seconds |
|
2. | Rollout1 | 10/7 6/3 5/2*(2) | eq: -0,694 (-0,019) |
| Player: Opponent: | 23,43% (G:9,27% B:0,29%) 76,57% (G:29,00% B:3,91%) | Conf.: ± 0,007 (-0,701...-0,688) - [0,0%] Duration: 10 minutes 34 seconds |
|
3. | Rollout1 | 12/9 10/7(2) 6/3 | eq: -0,701 (-0,026) |
| Player: Opponent: | 20,02% (G:3,48% B:0,15%) 79,98% (G:19,42% B:1,41%) | Conf.: ± 0,005 (-0,707...-0,696) - [0,0%] Duration: 8 minutes 53 seconds |
|
4. | Rollout1 | 6/3 5/2*(2) 4/1 | eq: -0,712 (-0,037) |
| Player: Opponent: | 21,62% (G:9,29% B:0,21%) 78,38% (G:27,16% B:2,65%) | Conf.: ± 0,006 (-0,718...-0,707) - [0,0%] Duration: 9 minutes 11 seconds |
|
|
1 2592 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
|