|
BGonline.org Forums
How to design a statisctally significant bot head-to-head?
Posted By: Bob Koca In Response To: How to design a statisctally significant bot head-to-head? (Timothy Chow)
Date: Sunday, 10 January 2010, at 6:45 p.m.
You would want to switch the dice by using the same rolls but switching the player that wins the opening roll. Then consider those pairs of games as matched pairs and look at the equity difference player A - Player B. Then use a t-test on the single list of numbers that results. This can be used in addition to the usual variance reduction techniques.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.