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

[William Harold Newman <william.newman@xxxxxxxxxxx>]

On Fri, Feb 11, 2005 at 01:10:09AM +0100, Erik van der Werf wrote:
> >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.

This discussion reminds me of the idea that superduperblitz time --
0.5 seconds per move, e.g. -- would eliminate the possibility of many
forms of cheating. (Good luck trying to help your program play at that
speed!:-) Thus, with such a time control, you could run an all-comers
tournament without having to fuss with the usual bozos who think
cheating is satisfying.

It would be unsatisfyingly artificial at some level, and trying to do
it over the Internet would require ping times that we usually
associate with Counterstrike (or at least internet telephony) rather
than Go. But it doesn't seem impossibly hard even today, and it should
become easy not too far in the future.

So far my programs have tended to be slow, so for my programs in
particular these rules would tend to be painful. But my programs have
tended not to be very good anyway.:-| And from an abstract engineering
point of view, this idea appeals to me in the same way as the rules of
the ICFP Functional Programming Contest. Instead of trying to enforce
rules about team sizes and who is qualified to enter and so forth,
they just accept all comers and run the tournament fast. Simple,
self-enforcing, elegant. (And it would have the advantage that you
could run a computer-vs.-computer 7-game match in about the time it
takes to get coffee.:-)

Before anyone gets too indignant about this: Of course I'm not arguing
that traditional tournaments should be changed to this format. That
is, this is not offered in the spirit of "a computer tournament should
have an unambiguous ko rule!" or "0.5 is too small a komi!" Instead,
it is in the spirit of "wouldn't it be interesting to have a 9x9 (or
37x37, for that matter) tournament?"

-- 
William Harold Newman <william.newman@xxxxxxxxxxxxxxxxx>
"A microsecond is not a lot of time to route a packet." -- overheard at IETF
_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/