| |
BGonline.org Forums
Containment: Might 7-ply outperform XGR+?
Posted By: Casper van der Tak In Response To: Containment: Might 7-ply outperform XGR+? (Timothy Chow)
Date: Friday, 21 January 2011, at 10:32 a.m.
I am not sure what outperform would mean here. Predict equities more accurately? For me, there is little reason to assume that 7-ply analysis would do better here than a rollout. The position normally is not going to change dramatically over 7 plies.
I looked at GNU, and I believe GNU plays this OK, also at the lower plies. Make the ace, block high doubles, hit and create builders for the high points in the board, concentrate on not letting the last checker escape, and within those constraints slot and cover.
What do we expect as winning chances? I'd expect to complete a close-out from here about 55%-60% of the time. With 80% GWC from there, I'd expect winning chances in the 44%-48% range, more probably in the lower end. That is not so far off from evaluation and low level rollout results (for GNU).
GNU is weak when rolling a snake home, but this is never going to be a shake.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.