I dont know much about setting filters. But 4-ply w/roller cube says small double. (4-ply, 4-ply about the same)
                       |
| ||||
| XGID=-BBaBBC-B-----A----A--fdb-:1:1:1:00:0:0:3:0:10 | |||||
| Blue on roll, cube action? | |||||
| Analyzed in Rollout | No redouble | Redouble/Take |
| Player Winning Chances: | 69.26% (G:0.00% B:0.00%) | 70.20% (G:0.00% B:0.00%) |
| Opponent Winning Chances: | 30.74% (G:4.00% B:0.03%) | 29.80% (G:4.70% B:0.05%) |
| Cubeless Equities | +0.345 | +0.713 |
| Cubeful Equities | ||
| No redouble: | +0.602 (-0.013) | ±0.033 (+0.569..+0.635) |
| Redouble/Take: | +0.615 | ±0.072 (+0.543..+0.687) |
| Redouble/Pass: | +1.000 (+0.385) | |
| Best Cube action: Redouble / Take | ||
| Rollout details | ||
| 5184 Games rolled. Dice Seed: 3 Moves: 4-ply, cube decisions: XG Roller | ||
| Double Decision confidence: | 62.4% | |
| Take Decision confidence: | 100.0% | |
| Duration: 4 minutes 30 seconds | ||
eXtreme Gammon Version: 2.19.211.pre-release
Variant one pip back. I was curious as Whites 44 and 66 are stronger. But our 66 is better and more hits in general if White leaps.
Redouble/Take
|
| ||||
| XGID=-BBaBBC-B------A---A--fdb-:1:1:1:00:0:0:3:0:10 | |||||
| Blue on roll, cube action? | |||||
| Analyzed in Rollout | No redouble | Redouble/Take |
| Player Winning Chances: | 70.31% (G:0.00% B:0.00%) | 73.01% (G:0.00% B:0.00%) |
| Opponent Winning Chances: | 29.69% (G:5.11% B:0.09%) | 26.99% (G:5.63% B:0.09%) |
| Cubeless Equities | +0.354 | +0.806 |
| Cubeful Equities | ||
| No redouble: | +0.604 (-0.212) | ±0.066 (+0.538..+0.670) |
| Redouble/Take: | +0.816 | ±0.137 (+0.680..+0.953) |
| Redouble/Pass: | +1.000 (+0.184) | |
| Best Cube action: Redouble / Take | ||
| Rollout details | ||
| 1296 Games rolled. Dice Seed: 3 Moves: 4-ply, cube decisions: XG Roller | ||
| Double Decision confidence: | 99.7% | |
| Take Decision confidence: | 99.6% | |
| Duration: 1 minute 06 seconds | ||
eXtreme Gammon Version: 2.19.211.pre-release
I thought I took these deeper than most people that I play based on race lead and what containment probability looked like. Surprised I didn't see the take is trivial in most variants. 2 checkers off matters far less than I expected. One or zero checkers off were mostly takes. So Ill have to review this as some sort of reference.