Nack played 11/8 8/5, and in Modern Backgammon (page 297) Robertie writes:
"A slight technical error. Better is 11/7, leaving a spare on the 6-point. The actual play gains on 44, but loses on 62, 63 and 64."
Lets call Nack's play A and Robertie's play B.
Here's a table with the equity difference between A and B after the next roll according to XG 5-ply eval. Only rolls where there is a difference are in the table.
| Roll | Diff(A-B) |
|---|---|
| 44 | +0.166 |
| 55 | +0.105 |
| 54 | +0.067 |
| 66 | +0.052 |
| 43 | +0.041 |
| 53 | +0.035 |
| 51 | +0.025 |
| 31 | +0.020 |
| 61 | +0.011 |
| 33 | +0.002 |
| 41 | -0.007 |
| 52 | -0.010 |
| 64 | -0.097 |
| 63 | -0.110 |
In his analysis Robertie mentions the two most favorable rolls for play B, the most favorable roll for play A and also a play (62) where there is no equity difference, but he fails to mention several other rolls that are favorable for play A.
XG rollouts, 3-ply and 4-ply.
                        |
| |||||||||||||||
| XGID=cBBBCBC----A-------bbbc-c-:2:-1:1:31:5:13:0:25:10 | ||||||||||||||||
 to play 31 | ||||||||||||||||
is Player 2| 1. | Rollout1 | 11/8 6/5 | eq: +2,084 | |||
| ||||||
| 2. | Rollout1 | 11/7 | eq: +2,076 (-0,008) | |||
| ||||||
| 3. | Rollout1 | 11/8 4/3 | eq: +2,070 (-0,014) | |||
| ||||||
| 1 10368 Games rolled with Variance Reduction. Dice Seed: 36095669 Moves and cube decisions: 3 ply | ||||||
eXtreme Gammon Version: 1.14, MET: Rockwell-Kazaross
| 1. | Rollout1 | 11/8 6/5 | eq: +2,088 | |||
| ||||||
| 2. | Rollout1 | 11/7 | eq: +2,078 (-0,010) | |||
| ||||||
| 3. | Rollout1 | 11/8 4/3 | eq: +2,073 (-0,015) | |||
| ||||||
| 1 5184 Games rolled with Variance Reduction. Dice Seed: 36095669 Moves and cube decisions: 4 ply | ||||||
eXtreme Gammon Version: 1.14, MET: Rockwell-Kazaross