| |
BGonline.org Forums
How to improve your game
Posted By: Rich Munitz In Response To: How to improve your game (Bob Koca)
Date: Tuesday, 28 July 2009, at 8:48 p.m.
I've already commented on several occasions that I don't generally care for long rollouts for the purposes of trying to improve one's play.
Long rollouts to the point of statsig are useful when developing benchmark positions where you want to find the exact take/drop cusp in a position type. Long rollouts are good for openings where the exact same position appears over and over again and where it may be useful in obtaining some conceptual insights.
However, the rest of the time, I think they usually are not the most effective use of time. Either two plays will be sufficiently close so that it doesn't make a big difference in the grand scheme of things, or one play will be proven to be clearly correct by a good margin with a short rollout.
My goal is to identify my LARGE errors where BOTH the bot analysis AND a short rollout agree that my play was wrong by a lot - AND where my second look at the position without looking at the analysis result also produces the incorrect play again. Sometimes I get it right the second time and if so, I obviously must have some understanding of what's going on. But if all three tests fail, that position is one that I will spend a good amount of time on trying to understand the key concept or technical aspect of the position that I missed and try to learn from it.
There is no shortage of blunders to be found in my play from which to learn from, and spending time to try to fix those misunderstandings should reap the largest rewards. Why waste time on the 0.02 errors? I'll note them, mostly to ensure that I keep my mind open to consider similar plays in the future (and know that the result is probably not statsig anyway), but I won't dwell on them.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.