The cube location makes a big difference here. With cube access,
X is looking to cash next turn. 5/4 is a higher-variance play;
it makes the good sequences better and the bad sequences worse,
and in particular if O rolls a 2 then she will be too good. X just
needs his good sequences to be good enough to cash, and 2/1
accomplishes that, while avoiding giving O rolls that leave her
too good.
With the cube turned (or at DMP), the more aggressive 5/4 is XG's
preference.
Note also that if the Jacoby rule were in force, then X wouldn't
have to worry about O being too good, and XG favors the somewhat
unusual-looking 18/12*/11. I haven't tried to figure out why XG
prefers this over 18/12* 5/4.
1. Rollout¹ 18/12* 2/1 eq:+0.260
Player: 60.28% (G:12.29% B:0.13%)
Opponent: 39.72% (G:19.99% B:0.89%)
Confidence: ±0.003 (+0.257..+0.263) - [100.0%]
2. Rollout¹ 18/12* 5/4 eq:+0.178 (-0.082)
Player: 62.47% (G:16.78% B:0.41%)
Opponent: 37.53% (G:25.37% B:1.29%)
Confidence: ±0.005 (+0.172..+0.183) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
--------------------
With the cube turned
--------------------
XGID=--ABBDBB--b-a-----Bbbbbbb-:1:-1:1:61:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | O O O O O O | +---+
| X | | O O O O O O | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | X |
| | | X |
| O X | | X X X X |
| O O X | | X X X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 98 O: 85 X-O: 0-0
Cube: 2, O own cube
X to play 61
1. Rollout¹ 18/12* 5/4 eq:+0.130
Player: 63.28% (G:16.16% B:0.37%)
Opponent: 36.72% (G:25.20% B:1.30%)
Confidence: ±0.006 (+0.125..+0.136) - [100.0%]
2. Rollout¹ 18/17 18/12* eq:+0.032 (-0.098)
Player: 60.22% (G:14.74% B:0.37%)
Opponent: 39.78% (G:26.65% B:1.35%)
Confidence: ±0.006 (+0.027..+0.038) - [0.0%]
3. Rollout¹ 18/12* 12/11 eq:+0.012 (-0.119)
Player: 59.39% (G:14.31% B:0.33%)
Opponent: 40.61% (G:26.58% B:1.29%)
Confidence: ±0.006 (+0.006..+0.018) - [0.0%]
4. Rollout¹ 18/12* 2/1 eq:+0.009 (-0.121)
Player: 60.71% (G:11.50% B:0.12%)
Opponent: 39.29% (G:20.13% B:0.88%)
Confidence: ±0.005 (+0.004..+0.014) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
------
Jacoby
------
XGID=--ABBDBB--b-a-----Bbbbbbb-:0:0:1:16:0:0:1:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | O O O O O O |
| X | | O O O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | X |
| | | X |
| O X | | X X X X |
| O O X | | X X X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 98 O: 85 X-O: 0-0
Cube: 1
X to play 16
1. Rollout¹ 18/12* 12/11 eq:+0.373
Player: 58.28% (G:14.99% B:0.34%)
Opponent: 41.72% (G:27.02% B:1.19%)
Confidence: ±0.005 (+0.367..+0.378) - [100.0%]
2. Rollout¹ 18/12* 5/4 eq:+0.333 (-0.039)
Player: 62.79% (G:15.02% B:0.32%)
Opponent: 37.21% (G:24.15% B:1.09%)
Confidence: ±0.000 (+0.333..+0.333) - [0.0%]
3. Rollout¹ 18/17 18/12* eq:+0.333 (-0.039)
Player: 59.16% (G:14.07% B:0.36%)
Opponent: 40.84% (G:26.12% B:1.09%)
Confidence: ±0.000 (+0.333..+0.333) - [0.0%]
4. Rollout¹ 18/12* 2/1 eq:+0.287 (-0.085)
Player: 59.85% (G:11.77% B:0.14%)
Opponent: 40.15% (G:19.68% B:0.77%)
Confidence: ±0.002 (+0.285..+0.289) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
---
Tim Chow