Author: Mogens Larsen
Date: 11:28:56 04/20/01
Go up one level in this thread
On April 20, 2001 at 12:26:31, Chessfun wrote: >No it isn't. Your argument about knowing DB wasn't available and therefore >wanting publicity can also be argued from the other side if no mention to IBM >was made. Then under your second argument simply checking wihtout making it >public, it would come out during questions at places like this. When they still >therefore get the advertizment you claimed in your original argument. >So it's damned if you do damned if you don't. That only holds water if most people within the business think that Deep Blue exist to this day, several years after the match. Not to mention the commotion raised by Kasparov, where it became clear that the machine couldn't be examined for devious devices because it did no longer exist. There's _no_ bad publicity in receiving the questions at a press conference and answering: "We contacted Hsu as early as possible, but Deep Blue has regretably sieged to exist after the Kasparov match. Therefore it is not invited". Getting publicity for doing something sensible isn't bad. >What? that means Junior can't participate in ANY other events except those >sanctioned by the ICCA? Of course they can. But it isn't ethical to help establish another championship title because you're unable to win the one available. It should be obvious why numerous championships hurt credibility for computer chess in general. Just look at the current human chess situation and professional boxing. >It is the point. The point is that a tournament consisting of x (notice removal >of 10) rounds could remove the best program. The chances of this happening over >an event with fewer programs is less likely. Not if it isn't organized as an ordinary round robin tournament. With knockout or smaller pools, it's possible to increase games between programs without increasing the number of days. That is because they don't all meet each other. Similar to say Tennis. >All the programs you named are non >commercial although maybe not a requirement, I am sure BGN would have thought >it was. I imagine that Bertil would have said something if that was the case. As far as we know: 1) Nothing prevented him from suggesting a non-commercial program. 2) Nothing prevented him from making inqueries to other commercial authors and hear if they had something up and running. 3) Nothing prevented him from posting a notice here or somewhere else. He did neither. That's why he messed up. My personal opinion is that he should have said no to the assignment altogether, because the idea of the strongest program cannot be solved under the conditions imposed. That would have earned him credibility here IMO. >Agh this I agree with a little more than the tournament. You originally wrote >about the thing taking a week or so. To play as you just wrote say 30 games >with 5 engines is 300 games at about 8 day. So even then we are left with 6 >weeks. I can't see BGN wanting a 6 week tournament to decide which program >should play. I see that I need to write my original idea about a knockout out in detail. Darn. Let's try an example (knockout): We have eight participants that wanted to play for the match against Kramnik after reading the public announcement. Let's say that we seed these programs by ICCA tournament results, a RR tournament with reduced timecontrol (at ICC?) or maybe just straws :-). One plays against eight, two against seven and so on. We have eight dual comps (lucky us), so it is possible to start all four matches at once. They play 40 games against each other. After 7-8 days it's over and there are four left. The winner of 1-8 meets the winner of 3-4 and the winner of 2-7 meets the winner of 3-6. After another 7-8 days there are only two left. After another 40 matches we have a winner. That means 8+8+8=24 days at most (except the possibility for tiebreaks) and the winner has played 120 games. Taking the timeframe into account, these can be expanded or reduced. That is fewer games than the current arrangement obviously, but there is fairness and diversity. There are also other possibilities, eg. pools+knockout (NHL), but I'm not going to write them down. >That was all I was able to find. Now I see at http://www.kasparov.com also >a notice but again no mention of World Championship. They use both #1 and "One of the best", so it's unclear. >How do you know BGN didn't take care of the commercial only part? How are they supposed to take care of the commercial part if their expert doesn't recommend any non-commercial programs? Mogens.
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.