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

BGonline.org Forums

Some questions/comments re: opening roll RO's

Posted By: Maik Stiebler
Date: Friday, 14 December 2012, at 10:17 a.m.

In Response To: Some questions/comments re: opening roll RO's (Chuck Bower)

... the humans who get compared are likely the most vociferous in claiming that they are stronger than bots of similar error ratings. The flip-side of that is that the bots are the judges of their own strength, as well as human strength, when you simply use the calculated error rates.

That's where unbiased measures of skill come in handy.

The general rule is "take the best information available" but I'm questioning whether or not mega-bot vs. mega-bot is really that.

Regarding choosing one's move, I agree that the idea that it's always best to do what the highest level rollout calls correct is questionable. If the choices are close, you need only the slightest suspicion that you will handle the followup better or that your opponent is more likely to butcher it to justify deviating from the rollout advice. On the other hand, you probably will have to do this kind of adjustments for lower level rollouts as well, and then you also have to figure in (more) bot weaknesses, or work on the assumption that the bot weaknesses are similar to human weaknesses, which is dubious.

I would find it interesting to study the influence that rollout settings have on the result. Unfortunately, except for the early opening and the Depreli positions (for a comparison of XG and gnubg rollouts), I think data is scarce.

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.