[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

Random thoughts on adjustment of play, PR level and expected win rate

Posted By: Timothy Chow
Date: Friday, 25 November 2011, at 9:36 p.m.

In Response To: Random thoughts on adjustment of play, PR level and expected win rate (Stick)

Stick wrote:

What do you have of Mikael's, a couple hundred matches tops? Near meaningless esp. against equal like opp's.

This overstates the case. Suppose A and B are equally strong. The probability that A wins 140 or more matches against B out of 200 matches is less than 1 in a hundred million.

Even the probability that A wins 65 or more matches against B out of 100 matches is less than 1 in 500.

It's possible that you're confusing the number of trials needed for a statistically significant rollout result with the number of matches needed to determine if A is stronger than B. If you want to leverage your intuition from doing rollouts, consider comparing two DMP plays where Play A gives you (say) 15% higher match-winning chances than Play B. It should be clear that you don't need 46K trials to observe a statistically significant difference between the two plays.

Messages In This Thread

 

Post Response
Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:
 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.