I'm gonna play to 11 and 3. I want to maximize outfield coverage for his jumps.

11 and 10. I don't want to break up the wall. 11 and 10 lets me get the pieces closer to safety. If 10 gets captured there is less ways that the other vulnerable piece can get captured than if you played 12 and 9 where 9 gets captured.

All the future possible rolls next turn seem better for you than 12 and 9. This assumes you don't get captured next turn.

I think slotting 12/7 is worth a look. A lot of White's 6s are going to be pretty good in any case, and you may get a shot at the checker on the 23 if you get hit, so you aren't paying full price to getting hit.

Don't get me wrong, you are paying pretty big to the 6s on average compared to say, 12/10 14/11, but you also gain pretty big on the non 6s. I'm thinking you have a good cube on a lot of White's numbers after playing 12/7 where you don't after 12/10 14/11. This could be something like if White owned the cube, you would prefer a safer play, but because you get a good cube in on White's bad numbers, you can make the more volatile play.

1. XG Roller+ 12/7 eq:+0.213
Player: 57.45% (G:7.21% B:0.16%)
Opponent: 42.55% (G:13.06% B:0.32%)

2. XG Roller+ 14/9 eq:+0.173 (-0.039)
Player: 56.21% (G:6.74% B:0.15%)
Opponent: 43.79% (G:9.56% B:0.26%)

3. XG Roller+ 14/11 5/3 eq:+0.155 (-0.058)
Player: 55.57% (G:6.73% B:0.16%)
Opponent: 44.43% (G:10.10% B:0.28%)

4. XG Roller+ 14/12 5/2 eq:+0.112 (-0.100)
Player: 54.52% (G:6.31% B:0.17%)
Opponent: 45.48% (G:8.89% B:0.22%)

5. XG Roller+ 14/11 12/10 eq:+0.105 (-0.107)
Player: 54.09% (G:6.38% B:0.17%)
Opponent: 45.91% (G:9.25% B:0.26%)