Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Future man-machine events; inherent problems?

Author: pete

Date: 00:24:52 02/18/00

Go up one level in this thread


On February 17, 2000 at 23:41:55, Mig wrote:

>Thanks to everyone for their feedback, both here and by e-mail. I'm glad to see
>I wasn't wrong in imagine this to be a rational and well-moderated group and I
>received many good ideas. My only wish is that a select few would present their
>solutions to the problem instead of calling me names! At the death it was either
>forfeit Adams, who wasn't going to play that late and couldn't play the next
>day, or forfeit Junior.
>
>This has provoked a lot of discussion about computer play in future events, and
>whether or not the special characteristics of computer players, not just their
>playing strength, should be considered in organizing events. Under perfect
>conditions (Frankfurt last year) objections are minimal, but here we saw a how
>flaws in a system -- our system in this case -- can be amplified.
>
>Should computer players be limited to specific man-machine events, like the
>upcoming match between Xie Jun and Deep Junior? Is play against computers
>different enough to warrant "segregating" them and not mixing them into GM
>tournaments with thousands of dollars on the line? I'm also asking the same
>questions of top GMs, because at the end of the day if they won't play, the
>question is moot. Are specific man-machine events in some way inferior or less
>attractive for programmers or the public?
>
>Thanks, Mig


The only thing I can see is that if you start a new kind of tournament system it
takes some time to come up with sensible rules dealing with all possible
problems ; this was the same in human tournaments which can easily be seen if
you read some of the strange rules that were chosen in very early tournaments
ie. in 19th century . Even in Karlsbad 1929 they still had rules like : everyone
who wanted to agree to a draw had to explain it to the arbiter ( as far as I
remember it was some honourable person named Tietz with limited chess knowledge
and only if he agreed it was a draw ).

Same thing was to some extent true for the first cct1 tournament on ICC ( with
only chessprograms in ) although there all the participants were used to play
over the net and experienced with the inherent problems.

The rule missing was obviously some deadline for connection problems . This is
for sure more difficult if the organizer himself is responsible for setting
everything up ( which is the impression I got through reading the posts ) .

Every rule dealing with special treatment of a computer player should be made
_before_ the tournament starts ; then the programmer had as a fair chance to
decide if under these circumstances he wants to play . If everything posted is
true I see no sensible alternative for the Junior forfeit but if this is decided
on the fly without any rules based on and even worse if it decided against the
published decision of the chief arbiter you sure ask for problems ...

When the tournament has started I think the computer player has to be treated
with just the same respect as any human grandmaster .

I think at the moment we are in the thrilling short period of time where this
kind of tournament ( man-machine ) makes sense as top humans and computers play
in the same league ( at least in g/60 ) ; in 10 years nobody will interested in
this anymore I think as the computers will have taken over , so it would be a
shame to lose the chance to see this kind of competition .

Also from a commercial point of view btw : I am quite sure that Adams-DJ was the
game that attracted most spectators during the round and Kasparov - DJ would
have been a fair test for your server setup :=) .



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.