                         |
| ||||
| XGID=-a-BB-C-B---cDa-bc-e----B-:0:0:1:42:0:0:3:0:10 | |||||
| Blue to play 42 | |||||
eXtreme Gammon Version: 2.10
I'll guess that most of the people getting this wrong are playing 24/18. On most days, that would probably my over-the-board play.
Interestingly, a blot on the 18pt is just as vulnerable as one on the 20pt. The opponent White has four direct numbers targeting each of them. I count 16 point-on-head rolls that make the 20pt, and the same number that make the 18pt. When you factor in the return shots after a loose hit, it may be that leaving a blot on the 20pt is actually better than putting one on the 18pt. A critical disadvantage of starting the 20pt, of course, is that White will have made his most valuable inside point when he rolls one of the 16 POH numbers.
In the early game, after your opponent has escaped one of his rear checkers, it is frequently correct to unstack the midpoint when you roll a 2. A checker placed on the 11pt is ideal for containing the final runner. You will get hit when your opponent rolls 64, but when he tries running with any smaller number, you are assured of getting at least a double shot. In this position, I like playing down to the 11pt.
I also like spltting the rear checkers. Not splitting seems like the slow way to lose. Allowing the opponent to improve his prime without being challenged could be fatal. After a split, White will usually attack, but the 3-point board owned by Blue may lessen the danger. Forcing White to attack, before his prime is filled in, is probably better than waiting.
Playing two down would be my choice if I decided splitting were too dangerous. I don't really like it, however, because it strips the midpoint, and also allows the opponent to build his prime without any disturbance from me. A checker on the 10pt is an unnecessary target that adds little or nothing to help contain White's last runner.
This analysis leaves me with three candidates:
- Split to the 18pt: 24/18
- Split to the 20pt: 24/20, 13/11
- Play two down: 13/11, 13/9
I will try split-and-down: 24/20, 13/11.
Mike