A good starting point in our reasoning could be the following third roll position arising as 42P-61P-21:

              
 
              
 | | White is Player 2
score: 0 pip: 160 | Unlimited Game Jacoby Beaver | pip: 161 score: 0
Blue is Player 1 | |
| XGID=-b--B-D-B---dE---bbe----B-:0:0:1:21:0:0:3:0:10 |
| Blue to play 21 |
| 1. | Rollout1 | 24/23 13/11 | eq: +0.073 |
| Player: Opponent: | 51.38% (G:16.39% B:0.74%) 48.62% (G:13.86% B:0.64%) | Conf.: ± 0.004 (+0.069...+0.078) - [100.0%] Duration: 1 hour 31 minutes |
|
| 2. | Rollout1 | 13/11 6/5 | eq: +0.062 (-0.011) |
| Player: Opponent: | 51.57% (G:16.00% B:0.74%) 48.43% (G:14.81% B:0.98%) | Conf.: ± 0.005 (+0.057...+0.067) - [0.0%] Duration: 1 hour 38 minutes |
|
| 3. | Rollout2 | 24/21 | eq: +0.035 (-0.038) |
| Player: Opponent: | 50.66% (G:14.49% B:0.64%) 49.34% (G:13.61% B:0.53%) | Conf.: ± 0.007 (+0.028...+0.042) - [0.0%] Duration: 31 minutes 21 seconds |
|
| 4. | Rollout2 | 24/23 24/22 | eq: +0.010 (-0.063) |
| Player: Opponent: | 50.32% (G:14.37% B:0.59%) 49.68% (G:13.91% B:0.53%) | Conf.: ± 0.007 (+0.003...+0.018) - [0.0%] Duration: 30 minutes 48 seconds |
|
| |
1 15552 Games rolled with Variance Reduction. Dice Seed: 70051551 Moves: 3-ply, cube decisions: XG Roller
2 5184 Games rolled with Variance Reduction. Dice Seed: 70051551 Moves: 3-ply, cube decisions: XG Roller
|
Rollout by Igor Erovenko 2013-Sep-14 eXtreme Gammon Version: 2.10 42P-61P-21-ie.xgp [S $11 U38 V63] "<=15*5 |
eXtreme Gammon Version: 2.10
In the actual fourth roll position, white has split to the 23pt. First of all, this naturally eliminates slotting from the candidates. I think we aren't sufficiently back in the race to risk being hit here with 23 numbers. Next, splitting also goes down in relative equity. I believe one of the reasons for this is in the fact that splitting is a wrong reply to 51S-21 (Up -- 24/21 is the right move). The blot on the 11pt is subject to 6 shots. Compare this with 52S-21 when splitting is best! (Fewer hitting numbers and duplication of 64.) I use this observation as a general rule of thumb that with the back formation 24-23 having a blot on the 11pt is not desirable, while with the 24-22 formation it's much safer.
Here is the rollout of the OP.

              
 
              
 | | White is Player 2
score: 0 pip: 159 | Unlimited Game Jacoby Beaver | pip: 161 score: 0
Blue is Player 1 | |
| XGID=-aa-B-D-B---dE---bbe----B-:0:0:1:21:0:0:3:0:10 |
| Blue to play 21 |
| 1. | Rollout1 | 24/21 | eq: -0.032 |
| Player: Opponent: | 49.25% (G:13.17% B:0.55%) 50.75% (G:13.94% B:0.58%) | Conf.: ± 0.002 (-0.034...-0.030) - [92.9%] Duration: 20 hours 35 minutes |
|
| 2. | Rollout1 | 24/23 24/22 | eq: -0.034 (-0.002) |
| Player: Opponent: | 49.32% (G:13.28% B:0.53%) 50.68% (G:14.42% B:0.54%) | Conf.: ± 0.002 (-0.036...-0.032) - [7.1%] Duration: 19 hours 52 minutes |
|
| 3. | Rollout2 | 24/23 13/11 | eq: -0.043 (-0.011) |
| Player: Opponent: | 48.94% (G:14.50% B:0.57%) 51.06% (G:15.02% B:0.69%) | Conf.: ± 0.003 (-0.046...-0.040) - [0.0%] Duration: 10 hours 52 minutes |
|
| 4. | Rollout3 | 13/10 | eq: -0.070 (-0.038) |
| Player: Opponent: | 47.83% (G:14.65% B:0.57%) 52.17% (G:14.66% B:0.81%) | Conf.: ± 0.008 (-0.078...-0.062) - [0.0%] Duration: 2 hours 13 minutes |
|
| 5. | Rollout3 | 24/23 6/4 | eq: -0.092 (-0.060) |
| Player: Opponent: | 47.92% (G:13.60% B:0.54%) 52.08% (G:15.31% B:0.58%) | Conf.: ± 0.008 (-0.100...-0.084) - [0.0%] Duration: 2 hours 08 minutes |
|
| |
1 62208 Games rolled with Variance Reduction. Dice Seed: 62811902 Moves: 3-ply, cube decisions: XG Roller
2 31104 Games rolled with Variance Reduction. Dice Seed: 62811902 Moves: 3-ply, cube decisions: XG Roller
3 5184 Games rolled with Variance Reduction. Dice Seed: 61295014 Moves: 3-ply, cube decisions: XG Roller
|
eXtreme Gammon Version: 2.10