| Thanks to Dmitriy, David, Nack and Tim for helping me out on this one. I had missed the point entirely.
My original post mentions my observation that the back checkers have already been split. I was thinking that moving out to the bar point with 24/18 was like resplitting (another mistake I've been known to make). Instead of resplitting the checkers that have already been split, I thought, I must use this roll productively in some other fashion. Now I see that the already developed midpoint plays an even more important role in the decision. The risk of a six-split with 62S is more worthwhile when you can develop an overstacked midpoint at the same time. But here, the 51S opener has already developed the midpoint somewhat. The other big takeaway is David's warning to avoid putting the rear checkers on the 23pt and 18pt in the early going. You give your opponent a six, while leaving none for yourself. Here's my rollout. As a bonus I rolled out Tim's variation, 52S-42P-62. (I had it in the oven before Dmitriy posted his.) This one includes a few extra low-equity alternatives. Mike |
51S-42P-62 [R $33 S38 C110] "<=5
| It looks like plays 1 and 3 were rolled out using the seed 0, while plays 2 and 4 use a random seed. But I'm not sure. The two rollouts were made separately, but I have the "random seed" choice selected. I don't understand how the seed could have been set to 0. Furthermore, I thought I could add new moves to a rollout without causing them to use a new seed. Perhaps, then, they both use the same seed. |
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
eXtreme Gammon Version: 2.01
52S-42P-62 [Z S18 R30 C37 r46 $101 c105 s110] "<=5
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
eXtreme Gammon Version: 2.01