Author: Shep
Date: 04:12:35 08/02/99
Go up one level in this thread
On August 02, 1999 at 06:40:03, Micheal Cummings wrote: >I still do not get why wait to release them, you know there are people here who >would be willing and able to test these settings against other Chessmaster >settings and other programs. It will be stupid to play games in tournaments and >say the settings only to have other people test them and find very different >results. > >What is the point of hiding them, if you can tell me that then I would be happy, >there are many settings being tested, and building these up to something I think >a immediate release would be better so the more the merrier in testing. > >I have got the kasparov personity to beat shredder2, so playing these settings >against other programs, say 5 or so games, means nothing. release them so we can >get down to some serious testing, adding to both of yours tests. First, I did not imagine my decision to withhold the settings until the end of testing would cause a longer discussion (and no, I am *not* doing it just to artificially raise interest!). One of the points is what you mention above: on one hand, you say "serious testing", on the other you talk about "testing these settings against other Chessmaster settings". That's what I feared about releasing them too early: People would run 6666 against another CM personality, probably at 10/game on a 133 MHz machine (OK, I'm exaggerating a bit ;-), report a 2:8 score (this has actually happened, BTW, against CM 6000 SS10) and conclude the settings are weak. Other people would disagree and bang! would we have another big discussion, similar to the "Is CM rightfully the SSDF #1/#2?" threads. That's why I wanted to back the settings up with some serious tournament results. Another reason is - if you may forgive this slight vanity on my part - the reputation that is at stake. When I created CM 5555, it was pure luck that it turned out to be so very strong. Now everyone expects "something" when I step forward again with a personality that is intended to surpass the former. That is why I want to make absolutely sure that there is nothing wrong with it and that the good results are not just a coincidence (like the recent Crafty-Hiarcs match (from Tania?) where Crafty started out 3:0, only to lose 6:18 or so). And reiterating my reasoning above, more people testing it does not necessarily mean more relevant data are produced to decide if this personality is "worth" being called CM 6666. --- Shep
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.