[computer-go] UCT/MoGo confusion

[Yizao Wang]

Hi,

>Perhaps I am oversimplifying, but if I understand correctly, MoGo 
>primarily gets its strength in 9x9 go by improving upon the random 
>simulations by preferring "good" moves over purely random moves during 
>the random game. Yet I have two results that seem to indicate that it's 
>not really that simple
>
>The first is that we have a purely random UCT version running on CGOS 
>(GoJin) and its rating seems to sit around 1640. But in the paper we are 
>told that the very first Mogo achieved almost the identical rating, yet 
Well, it's possible that the two programs have not the same speed.... this might make a big difference.
>it already had some improvements, such as preferring large captures. Can 
>I conclude that those improvements to the random simulations actually 
>have no effect on the performance of the program?
Surely there would be some effect, but might not be that much as expected.
>
>But even more confusing to me is that we've tried some simple 
>improvements to the random program that have had no effect. The ones 
>that I was certain would improve performance were versions that changed 
>random simulations so that moves near existing stones would be preferred 
>over stones placed too far away from the action. Many versions have been 
>tried, e.g., moves that must be adjacent to some other stone, moves that 
>must be no more than 1 space away from existing stones, etc. Surely on 
Well, I think we have tried some similar things. Personally I don't think this will lead to a significant improvement, and it is what the result shows in our experiment.

>average these are going to be better moves than purely random moves -- 
>or is this, indeed, the flaw in my logic? Shouldn't these versions 
>outperform the purely random versions? In almost every case the modified 
No not necessary.
>version performed identical to the purely random version -- no worse and 
>no better -- at least according to the self tests. Does this really 
I dont know what do you mean by 'identical' and 'self tests' here. Intuitively it is not a good idea to have your program test against itself. I believe every change will bring something different, but sometimes not very significant.

Yizao
>sound right?

>
>-Richard
>_______________________________________________
>computer-go mailing list
>computer-go at computer-go.org
>http://www.computer-go.org/mailman/listinfo/computer-go/
>