I don't think that XGR++ can be trusted in this position but it does illustrate that XG does see merit in kamikaze plays sometimes (at least at DMP). Note that the plays below are not the top 6 XGR++ plays. The first couple didn't even make it through the move filter and had to be manually added. The play 24/23 7/6 5/4 2/1* is what XG 3-ply plays (note: what XG 3-ply plays is not necessarily the same as the move that has the highest XG 3-ply equity).
         
 
       
 | | White is Player 2
score: 0
pip: 134 | | 1 point match | pip: 178
score: 0
Blue is Player 1 |
|
| XGID=-aBa-BBB-aB--a-a--bbbbbBC-:0:0:1:11:0:0:0:1:10 |
| Blue to play 11 |
| 1. | XG Roller++ | 24/23(2) 10/9*(2) | eq: -0.605 |
| Player:
Opponent: | 19.77% (G:5.17% B:0.39%)
80.23% (G:48.04% B:6.28%) | |
|
| 2. | XG Roller++ | 10/9* 5/3* 2/1* | eq: -0.608 (-0.003) |
| Player:
Opponent: | 19.62% (G:0.54% B:0.03%)
80.38% (G:57.79% B:22.56%) | |
|
| 3. | XG Roller++ | 24/23 10/9* 5/3* | eq: -0.619 (-0.015) |
| Player:
Opponent: | 19.04% (G:2.39% B:0.24%)
80.96% (G:52.08% B:13.70%) | |
|
| 4. | XG Roller++ | 24/23 7/6 5/4 2/1* | eq: -0.664 (-0.059) |
| Player:
Opponent: | 16.81% (G:0.57% B:0.02%)
83.19% (G:51.98% B:12.23%) | |
|
| 5. | XG Roller++ | 10/9* 9/8 2/1*(2) | eq: -0.719 (-0.115) |
| Player:
Opponent: | 14.03% (G:3.57% B:0.22%)
85.97% (G:49.94% B:8.43%) | |
|
| 6. | XG Roller++ | 10/9*(2) 2/1*(2) | eq: -0.722 (-0.118) |
| Player:
Opponent: | 13.88% (G:3.66% B:0.22%)
86.12% (G:49.06% B:7.46%) | |
|
eXtreme Gammon Version: 2.19.208.pre-release, MET: Kazaross XG2