| |
BGonline.org Forums
The True Test of Skill is Duplicate BG
Posted By: Maik Stiebler In Response To: The True Test of Skill is Duplicate BG (Stick)
Date: Sunday, 26 September 2010, at 8:33 a.m.
nobody takes the time to roll out the problems well past stat. sig. To hold this kind of quiz as I've said before, you need the rollouts to take the variance basically as low as possible, +/- .001 so in the event of two players having a close score it isn't decided by variance which could still be way off on the true value of your errors.
- Variance is a problem. Stat. sig., in itself, isn't (I think that's what you meant, basically, but my point is that if you have a decision with two very close top plays, you don't even have to roll them out to stat.sig., as long as you reduce variance to your threshold.)
- But why should backgammon players shy away from participating in a contest which is sometimes decided by variance?
- The influence of variance can be reduced either by increasing the number of rollout trials for each position, or by increasing the length of the quiz.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.