• Red to play  

            • Equities at 3-away: White, 3-away: Red

			13/11(2) 6/4(2)	+0.025
			50.1%   15.9%  1.8%    49.9%   15.5%   1.4%   CL: +0.007    CF: +0.025
			24/22(2) 6/4(2)	+0.016 ( -0.009)
			49.7%   13.2%  1.2%    50.3%   11.3%   1.0%   CL: +0.025    CF: +0.016
			24/22 13/11 6/4(2)        (936 games)	- 0.098 ( -0.123)
			48.9%   14.0%  1.5%    51.1%   14.8%   1.6%   CL: - 0.040    CF: - 0.066

            Full cubeful rollout, 1296 games, Play: 0-ply/expert, Cube: 2-ply/33%, using Jacob/Trice Met

            • Equities at 3-away: White, 3-away: Red

			24/22(2) 6/4(2)	+0.018
			49.5%   13.5%  1.2%    50.5%   11.5%   0.9%   CL: +0.023    CF: +0.018
			13/11(2) 6/4(2)	- 0.019 ( -0.037)
			49.8%   16.0%  2.1%    50.2%   15.7%   1.9%   CL: - 0.006    CF: - 0.019
			24/22 13/11 6/4(2)        (936 games)	- 0.083 ( -0.101)
			48.7%   14.7%  0.9%    51.3%   14.7%   1.5%   CL: - 0.036    CF: - 0.083

            Full cubeful rollout, 1296 games, Play: 0-ply/expert, Cube: 2-ply/33%, using Kazaross/Shaw Met (Ver.121503)

            • Equities at 3-away: White, 3-away: Red

			24/22(2) 6/4(2)	+0.017
			49.5%   13.6%  1.3%    50.5%   11.6%   0.9%   CL: +0.025    CF: +0.017
			13/11(2) 6/4(2)	- 0.014 ( -0.031)
			49.9%   16.0%  2.1%    50.1%   15.9%   1.9%   CL: - 0.004    CF: - 0.014
			24/22 13/11 6/4(2)        (936 games)	- 0.090 ( -0.107)
			48.6%   14.4%  1.2%    51.4%   14.8%   1.6%   CL: - 0.039    CF: - 0.090

            Full cubeful rollout, 1296 games, Play: 0-ply/expert, Cube: 2-ply/33%, using Mec26 Met

Usually when a rollout has a reversal of plays between cubeless and cubeful results (as showed on Trice Met here), it's from one of two reasons: 1) The position is a cash or a near cash. 2) The cubeful choice wins more games than the cubeless choice. 1) doesn't apply here, of course. As for 2), 13/11(2) 6/4(2) wins 0.4% more games than 24/22(2) 6/4(2), cubeful. My rule of thumb is: If the reversed play wins more than 0.020 (cubeless equities) than the cubeful play, I consider that it is mostly due to a rollout fluke of cube handlings which have happened during the rollout process. Standard deviation of the Trice Met rollout here is less than 0.020 (cubeful), 0.010 (cubeless) for both plays and the single-win difference is 0.018, so a longer-trial rollout (something like 5184 games) might make them reversed. In any case, I believe 24/22(2) 6/4(2) is the best play on this match score as Kazaross and mec26 Mets rollouts confirm.

GG

GS

DMP

Money