Author: Ed Schröder
Date: 02:25:46 03/20/98
Go up one level in this thread
>Posted by Keith Ian Price on March 20, 1998 at 01:37:33: >I also agree with your position. I think programmers spend too much time >improving their programs for the SSDF, instead of for humans. This is >one reason I think Bob and Bruce have it right, and the true testing >agency should be the chess servers. Who cares if MFritzo Pro '98 can >beat GenReb 14 on SSDF if against the humans their rating is 2300 while >playing 40/2 on the chess servers. Of course, the problem is that most >games on the chess servers aren't 40/2, but if one were to arrange a >computer chess qualification tournament to be run every three months on >ICC, I bet that enough IMs and GMs would be willing to participate for >purposes of establishing true ratings. Then again, maybe not. But >without Aegon, something has to be done. Aegon is indeed a great loss for computer chess. Yesterday on the news I heard Aegon made 1.1 billion bucks profit last year. The organizer of Aegon got a new boss. The boss didn't like chess. Aegon gone.... Behind the scenes there are talks for a successor of Aegon. Let's keep our fingers crossed. >But I understand what you are saying. If you are going to resign from >this discussion, don't do it by removing the autoplayer capability from >your next version, as you have stated previously, which will only punish >those of us who like to play them against each other. That's true and I would dislike it very much too. But keep in mind the following, from all commercial chess programmers only a few speak. The other ones are silent. They all read this forum but prefer to keep their silence. Be aware that after this new fashion (secret autoplayers) this will be not without consequences. I prefer to say it in public. But what will the others decide? Or have they already made up their minds? Do you really think nothing will happen? First signs are there, me, the open letter to SSDF signed by 3 main programmers. What will they do in their next versions? An open letter to the SSDF placed into public and signed by three leading chess programmers isn't something you can't ignore. If they don't get a satisfying answer or in worst case no answer at all then what can you expect? One thing what really bothers me is the silence of SSDF until now. I am sure they read every posting here. By keeping their silence they unintentionally may give the impression they don't take the complaints seriously. I would wish Thoralf to say "something". It doesn't matter what. Something like, "Hey, we heard you. We think about it. Be patience". - Ed - >Instead, do it by >implementing the best cheating learner for grabbing the number one spot >on SSDF, and when you do, and they all complain how it's so unfair, >repost this thread for them to read. >I don't pay much attention to the SSDF results, since all the top >programs are so close, it really makes no difference. Did I wait until >it was number one before I bought Rebel? No, I waited until it was on >special for DM 99. Did I wait until Fritz 5 was number one? No, I bought >it right after it lost 6 games to Crafty. I bought it because of its >features, not its strength. Only to find that with all the wonderful >features, it doesn't have an autoplayer, because they are afraid of >losing points on SSDF. Well, no Fritz 6 for me. I'll probably wait for >Fritz 16, when the true ELO improvement is sufficient to warrant an >upgrade, if they add no features, and leave out the autoplayer. Note >this is true for any program whose next version has no more features >than the last, and relies only on its 20-point rise in ELO on SSDF to >sell it, and is not directed toward Fritz itself. So please don't take >features out of Rebel 10, improve them instead. >Thanks, >kp
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.