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

BGonline.org Forums

New GV article

Posted By: Matt Cohn-Geier
Date: Tuesday, 15 December 2009, at 9:44 p.m.

In Response To: New GV article (Timothy Chow)

Common sense would suggest that 108 trials is not a lot of data.

True, but common sense would suggest a few other points: 1) that the bot's eval in this specific position is very good [with or without a variance reduced rollout]; 2) I don't know if you want to call it a Bayesian perspective but I'm more or less 110% certain of the answer here since it can more or less be calculated without a bot; 3) When you have over 44 JSDs and a standard deviation of .0008 after 108 games it probably isn't going to change.

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.