Forgive me the dumb original position. Of course it's nonsense with cube centered. Blue had a big double before the roll, which White would have to drop. With such low gammons, Blue cashes next turn if missed, so Bagai play of 8/2* 6/2 makes no sense. Reward is the same, while risk is higher than with 8/2* 8/4.
        
 
         
 | | White is Player 2
score: 0
pip: 78 | | Unlimited Game | pip: 65
score: 0
Blue is Player 1 |
|
| XGID=-BaCDBB-B----------cccbab-:0:0:1:64:0:0:0:0:10 |
| Blue to play 64 |
| 1. | Rollout1 | 8/4 8/2* | eq: +0,456 |
| Player:
Opponent: | 66,18% (G:1,74% B:0,03%)
33,82% (G:0,91% B:0,02%) | Conf.: ± 0,001 (+0,455...+0,457) - [100,0%]
Duration: 2 minutes 27 seconds |
|
| 2. | Rollout1 | 8/2* 6/2 | eq: +0,401 (-0,055) |
| Player:
Opponent: | 71,32% (G:1,62% B:0,02%)
28,68% (G:4,94% B:0,14%) | Conf.: ± 0,001 (+0,400...+0,402) - [0,0%]
Duration: 5 minutes 47 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Moves and cube decisions: XG Roller+
Search interval: Huge
|
eXtreme Gammon Version: 2.10
W/o acces to cube, the Bagai play wins a lot more (if you're missed, you have pretty secure bearoff, while if you play 2 from the 8, most 6x and 5x blot next turn), so it compensates the gammon losses well.
| | White is Player 2
score: 0
pip: 78 | | Unlimited Game | pip: 65
score: 0
Blue is Player 1 |
|
| XGID=-BaCDBB-B----------cccbab-:1:-1:1:64:0:0:0:0:10 |
| Blue to play 64 |
| 1. | Rollout1 | 8/2* 6/2 | eq: +0,363 |
| Player:
Opponent: | 71,58% (G:1,33% B:0,02%)
28,42% (G:4,93% B:0,14%) | Conf.: ± 0,001 (+0,362...+0,364) - [100,0%]
Duration: 1 minute 11 seconds |
|
| 2. | Rollout1 | 8/4 8/2* | eq: +0,254 (-0,109) |
| Player:
Opponent: | 66,56% (G:1,58% B:0,03%)
33,44% (G:0,99% B:0,02%) | Conf.: ± 0,002 (+0,253...+0,256) - [0,0%]
Duration: 1 minute 00 second |
|
| |
1 1296 Games rolled with Variance Reduction.
Moves and cube decisions: 4-ply
Search interval: Large
|
eXtreme Gammon Version: 2.10
In the next diagram, I moved one blue spare from 4pt to 7pt. The big difference is not the risk of 3 blots that Bagai leaves. After all Bagai vs non-Bagai play here is 3 blots vs 2 blots. In the original position it was 2 vs 1. The big difference is the diversification of numbers. The non-Bagai 8/4 8/2* doesn't produce as many blotting numbers next turn, as it did in the original position. So the Bagai play doesn't gain enough.
 
| | White is Player 2
score: 0
pip: 78 | | Unlimited Game | pip: 68
score: 0
Blue is Player 1 |
|
| XGID=-BaCCBBAB----------cccbab-:1:-1:1:64:0:0:0:0:10 |
| Blue to play 64 |
| 1. | Rollout1 | 8/4 8/2* | eq: +0,281 |
| Player:
Opponent: | 68,62% (G:1,65% B:0,03%)
31,38% (G:5,65% B:0,14%) | Conf.: ± 0,002 (+0,279...+0,284) - [57,9%]
Duration: 1 minute 48 seconds |
|
| 2. | Rollout1 | 8/2* 7/3 | eq: +0,281 (0,000) |
| Player:
Opponent: | 69,09% (G:1,69% B:0,04%)
30,91% (G:7,07% B:0,19%) | Conf.: ± 0,002 (+0,279...+0,283) - [42,1%]
Duration: 2 minutes 29 seconds |
|
| 3. | Rollout1 | 8/2* 6/2 | eq: +0,265 (-0,016) |
| Player:
Opponent: | 68,99% (G:1,42% B:0,03%)
31,01% (G:9,81% B:0,40%) | Conf.: ± 0,002 (+0,263...+0,267) - [0,0%]
Duration: 2 minutes 02 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Moves and cube decisions: 4-ply
Search interval: Large
|
eXtreme Gammon Version: 2.10
Last diagram I find interesting. It's the original position again, but white spares from 4 and 5pt were moved back into the outfield. The cube is in the middle, and the plays of 8/4 8/2* and 8/2* 6/2 are about even in equity.
What happens is that after a non-Bagai play, we are cashing in case of a miss, and loose via White's cash in case of hit-dance. After a Bagai play, we play on for the gammon in case of a miss, but also if she hits and we dance, she can play on. This trade-off favors the non-Bagai play, since it doesn't fall behind despite what cubeless numbers would suggest.
| | White is Player 2
score: 0
pip: 91 | | Unlimited Game | pip: 65
score: 0
Blue is Player 1 |
|
| XGID=-BaCDBB-B----a-a---cbbbab-:0:0:1:64:0:0:0:0:10 |
| Blue to play 64 |
| 1. | Rollout1 | 8/4 8/2* | eq: +0,492 |
| Player:
Opponent: | 68,37% (G:9,03% B:0,12%)
31,63% (G:1,09% B:0,02%) | Conf.: ± 0,002 (+0,490...+0,494) - [84,9%]
Duration: 1 minute 11 seconds |
|
| 2. | Rollout1 | 8/2* 6/2 | eq: +0,490 (-0,002) |
| Player:
Opponent: | 74,22% (G:8,87% B:0,10%)
25,78% (G:6,01% B:0,17%) | Conf.: ± 0,003 (+0,488...+0,493) - [15,1%]
Duration: 3 minutes 09 seconds |
|
| |
1 1296 Games rolled with Variance Reduction.
Moves and cube decisions: 4-ply
Search interval: Large
|
eXtreme Gammon Version: 2.10