| |
BGonline.org Forums
Scale of Open/Intermediaate Players
Posted By: Douglas Zare In Response To: Scale of Open/Intermediaate Players (Jason Lee)
Date: Monday, 4 May 2009, at 8:53 a.m.
I've never heard of any strategy discussions that have instructed me to try to get up on the roof against a closed board under the premise that I can't err while sitting on my duff.
I've made that argument. To clarify, the goal is not to get closed out, but a strategy which sometimes leads to getting closed out will be more effective on average than the GNU-measured error rate indicates. If you have to choose between two strategies with the same GNU error rate, you will be giving up less equity if you follow the one which leads to fewer unforced moves per game.
I guess getting closed out is such a horror that people don't want to think about it. Fine. Instead of getting closed out, let's imagine your move is forced because we are talking about a cube decision, and you have already doubled. Is it better to have a cube error rate of 10 millipoints per decision if you tend to double slightly early, so that you have an average of 10 decisions per game, or is it better to have a cube error rate of 8 millipoints per decision by doubling slightly late, so that you have an average of 15 decisions per game? You'll give up less equity with the error rate of 10, and fewer unforced decisions. Doubling early sets up easy (forced) decisions later.
At DMP, if you outplay someone consistently according to a perfect referee's Snowie-style error rate, and you average the same number of moves per game (perhaps due to resignations) then you will win on average. Consistently outplaying someone according to Gnu's measure does not mean you have an advantage.
Douglas Zare
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.