| |
BGonline.org Forums
Cube action - RO
Posted By: Rick Janowski In Response To: Cube action - RO (Mr Majestyk)
Date: Sunday, 30 December 2012, at 10:08 a.m.
I investigated the effect of the blots on the on the 14 and 16 point by looking at all combinations on the 14-17 points, i.e. outside of direct shot range of the opponent’s blot on the bar-point. XGR++ analysis indicates that only the 14 & 16 point gives a marginal double/pass (I had expected the 14 & 15 point combination to be slightly worse in reality, so this was surprising to me).
Although passing appears to be technically correct in the position, by a small margin, I tend to think that for practical purposes taking is correct for three reasons. Firstly, one cannot hope to pick out all the nuances in every specific position, and in this small family of similar positions (as investigated above), the general tendency is for taking to be correct, and if this was your judgement you will tend to make less errors overall in similar positions. Secondly, not only does this position appear to be the exception rather than the rule, but the scale of error is also negligible. Thirdly, snapping up close passes can often be beneficial in influencing subsequent late doubles from opponents, and where such games are turned around, may provoke some element of steaming.
From the rollouts I calculated that the taker’s cube efficiency is about 65% in this position. In 1-ply equity evaluations 70.5% was assumed by XG. At the deepest delving ply levels (7-ply, XGR+ and XGR++) the evaluations are such that equivalent to cube efficiency levels of about 67%. XGR+ and XGR++ are almost invariably very effective in this regard. Excellent work here by the developers.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.