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

BGonline.org Forums

Meta-MET Needed for Best 2 of 3 Matches

Posted By: Joe Russell
Date: Saturday, 30 July 2011, at 1:45 p.m.

In Response To: Meta-MET Needed for Best 2 of 3 Matches (Timothy Chow)

You are right, my thoughts on this subject have been 'weak forms of yours', and 'irrelevant'. I have had trouble 'distinguishing my thoughts from my original suggestions', and I have not tried to add anything to the discussion, but have merely attempted to 'defend' myself instead of 'focusing' more on the points you are trying to make. I guess I am 'not accustomed to reasoning at this level of abstraction.'

As a general principle, you would be better served to make your points without exhibiting such pomposity.

I can't resist one last comment. You stated; 'If XG plays Jellyfish, but doesn't change its behavior from one match to the next, then the unequal strength doesn't have any impact on independence.'

My original question was; 'It seems obvious that each match is independent when the players are equal, but how about when there is a significant difference in ELO?'

I pointed out that decisions made in one match can have an effect on latter matches. This dependence exists because more information about the opponent can be gained by the stronger player or hidden by the weaker player by making a slightly inferior play or cube decision that either lengthens or shortens the current game or match or steers it into, or avoids, a more complicated game. I mentioned that the greater the difference in ELO, the more effective this strategy could be. Neil pointed out he often used this strategy within a single match in the past, when playing an unknown opponent. Bob stated this strategy could be employed with equal players as well.

Obviously a bot can not make use of this information, it does not learn to play against an opponent's weaknesses, as it will only play as if it were playing itself. In the example you state, each match would be totally independent. Yes, that does seem obvious, but is it relevant to the discussion? Do you not agree that decisions made in a game can uncover or hide weaknesses that can be , or fail to be, exploited by humans in latter games or matches?

I think we can agree that each of the matches in a set are independent, unless an an inferior play or cube decision is made that sacrifices some current equity to gain, or prevent the gain, of knowledge that may be used effectively in a latter match.

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.