I agree with neilkaz that GNU 4 ply should be the standard, even with its faults. BGB needs major improvement, especially in regard to cube decisons. XG isn't currently up to the job, in my opinion, mainly because its fixed narrow move filters sometimes allow even major blunders through without being picked up as wrong at all. As in, below, 3ply analysis not letting 13/9 12/9 through the 1 ply filter.
| | is eXtremeGammon
score: 2 pip: 100 |
| | | | | | | | | | | | | | | |
| | |  | | | 5 point match |
| | | | | | | | | | | | | | | | | pip: 173 score: 2
is Tod |
| | XGID=-----bE-B-BaAB-----bcbcCb-:0:0:1:43:2:2:0:5:10 |
to play 43 |
| 1. | 3 ply | 12/8 6/3 | eq: -0.683 |
| Player: Opponent: | 38.69% (G:6.83% B:0.33%) 61.31% (G:28.23% B:1.17%) | |
|
| 2. | 1 ply | 13/9 12/9 | eq: -0.859 (-0.175) |
| Player: Opponent: | 37.51% (G:5.73% B:0.16%) 62.49% (G:32.46% B:1.42%) | |
|
| 3. | 1 ply | 12/9 6/2 | eq: -1.000 (-0.317) |
| Player: Opponent: | 33.32% (G:5.11% B:0.15%) 66.68% (G:37.63% B:1.77%) | |
|
| 4. | 1 ply | 13/10 12/8 | eq: -1.000 (-0.317) |
| Player: Opponent: | 32.41% (G:4.26% B:0.12%) 67.59% (G:35.59% B:1.39%) | |
|
| 5. | 1 ply | 6/3 6/2 | eq: -1.028 (-0.345) |
| Player: Opponent: | 32.14% (G:4.70% B:0.14%) 67.86% (G:40.24% B:2.03%) | |
|