Re: [computer-go] Computer Go tournament at EGF

[Erik van der Werf <E.vanderWerf@xxxxxxxxxxxxx>]

David G Doshay wrote:

> SlugGo plays to a specified search depth. Because this is a serial
> search I can think of no way to speed this up beyond what we have
> already done with hash tables, other than to make search depth
> a time dependent variable.

You could try to implement something like iterative deepening.

> It seems from the set of replies that computer Go wishes to retain
> what I see an an artificial but generally accepted limitation of one
> hour per program, with possible exceptions in some tournaments
> which allow for byo-yomi. I see the point of the convenience factor
> and have no choice but to either comply or not attend, but I do not
> think it is a good thing for the evolution of computer Go. David
> Fotland has said in an earlier thread that he is not really sure
> what MFoG might do with the extra time, but SlugGo clearly uses
> it to advantage and I have the data that proves it.

With infinite time my program would play perfectly so I don't think 
there's anything artificial about setting a time limit. For the 
evolution of computer Go I think fast games are generally preferable. It 
is much easier to experiment with fast games. With several 
hours/days/weeks per game it just takes too long before you can get any 
meaningful statistics to validate possible improvements.

> It looks to me like people are asking "what is the best blitz Go
> program" rather than "what is the best Go program." That is
> OK too, just a little different.

IMO 10 minutes for a 19x19 game could be considered blitz, one hour is 
already quite slow.

E.

_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/