This lucky roll turns the game around.
Bar/23 is forced and 11/9* is obvious. After one examines the options
a bit, 23/21 also becomes pretty obvious. The last deuce is the
tricky one.
Along with several others here, I played 9/7. This seems natural,
making a four-prime and reducing shots. However, we need to assess
the position as a whole. Though X now has an advantage, having just
put O on the bar against his four-point board, he still has two
checkers back, and still has quite a bit of work to do to lock up the
game. Since X has made his 1pt and 2pt, his position is more suited
to blitzing than to priming. Playing 11/9 for the fourth deuce gives
X builders from three different points to bear on his 3pt and 4pt, at
the cost of giving O 43 to hit from the bar. 11/9 gains more than it
risks. According to the rollout, both 9/7 and 7/5 are blunders.
=======
Rollout
=======
1. Rollout: Bar/21 11/9*(2) eq:+0.508
Player : 58.91% (G:36.73% B:0.60%)
Opponent: 41.09% (G:12.33% B:0.91%)
Confidence: ±0.012 (+0.496<E<+0.520)
2. Rollout: Bar/21 11/9* 7/5 eq:+0.429 (-0.079)
Player : 56.22% (G:34.13% B:0.61%)
Opponent: 43.78% (G:13.20% B:1.07%)
Confidence: ±0.015 (+0.414<E<+0.444)
3. Rollout: Bar/21 11/9* 9/7 eq:+0.422 (-0.086)
Player : 56.41% (G:33.11% B:0.59%)
Opponent: 43.59% (G:12.45% B:0.88%)
Confidence: ±0.017 (+0.405<E<+0.439)
1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply
eXtreme Gammon Version: 1.21
---
Tim Chow