Author: Robert Hyatt
Date: 06:25:33 04/13/00
Go up one level in this thread
On April 13, 2000 at 07:05:27, blass uri wrote: >On April 12, 2000 at 18:45:15, Robert Hyatt wrote: > >>On April 12, 2000 at 12:16:59, Tom Kerrigan wrote: >> >>>On April 12, 2000 at 10:28:54, Robert Hyatt wrote: >>> >>>>On April 12, 2000 at 04:54:39, Tom Kerrigan wrote: >>>> >>>>>On April 12, 2000 at 03:51:56, Ritter Rost wrote: >>>>> >>>>>>>And I'm just saying again, "What does it matter to you?" >>>>>> >>>>>>I am interested in an independent and strong SSDF which does not have to fear >>>>>>legal threats if they want to publish test results. This matters a lot to me. >>>>>> >>>>>>>Are you being bullied or prosecuted - No >>>>>>>Are the people being "bullied & prosecuted" publishing the results anyway - Yes. >>>>>> >>>>>>No they are not. See several posts of Bertil Eklunds on the matter. Ossi Weiner >>>>>>forbade them to publish the results of Genius, Zarkov and WChess. >>>>>> >>>>>> >>>>>>>My view from here is you are Anonomously trying to bring a private conflict into >>>>>>>a public forum. >>>>>> >>>>>>For computer chess the authenticity and independence of the SSDF as the only >>>>>>major testing institution is not a private matter. Nor is the behaviour of the >>>>>>major software publishers. >>>>>> >>>>>>> >>>>>>>The major achievement of this is a lot of discussion about the legality of your >>>>>>>original post & the possibility that CCC or the Moderators could be in legal >>>>>>>trouble because of Your post. >>>>>> >>>>>>CCC is a healthy, strong and commercially independent forum which can survive a >>>>>>little controversy as long as it remains free of _personal_ attacks, Tina. Some >>>>>>excitement here and there may even help to raise the number of visits/members. >>>>>> >>>>>>Furthermore there is no need to refer to my unthoughtful original post which has >>>>>>been thankfully deleted. I am simply quoting SSDF member Bertil Eklund in this >>>>>>thread. >>>>>> >>>>>>Bottom line: I would like the SSDF to be allowed to publish all test results. At >>>>>>the moment this is not the case. Bertil Eklund said that Ossi Weiner asked them >>>>>>to suppress the results of Genius, Zarkov and WChess, >>>>>> >>>>>>Ritter >>>>> >>>>>I agree with your entire post. >>>>> >>>>>What I don't understand is why Ossi doesn't want the results of these programs >>>>>published. They're all strong programs and I expect they would all end up within >>>>>a few points of the top program. It seems like that would be terrific >>>>>advertising for Ossi & co. >>>>> >>>>>-Tom >>>> >>>> >>>>If you read the email posted here, the answer is obvious. He believes that >>>>the auto232 software gives programs the opportunity to bias the results by >>>>various means (several have reported moves played under auto232 that are >>>>absolutely impossible to reproduce in testing.) >>> >>>I did read the e-mail, but I dismissed the reason as BS. If Ossi was really so >>>pissed about autoplayer trickery, he wouldn't have eventually allowed Shredder's >>>results in the list. And if I were him and I was convinced cheating was going >>>on, I would raise the issue (LOUDLY) in a public forum. If I had solid evidence, >>>a lot of people would get pissed at my competition... >>> >>>-Tom >> >> >>There _is_ a body of anecdotal evidence that suggests that auto232 both (a) has >>problems and (b) is subject to abuse. Several have looked at SSDF games where a >>program makes an outright blunder than can never be reproduced. Others have >>reported about trickery in using auto232 to bumfuzzle the opponent's learning, >>or break off a lost game, etc... >> >>Whether it happens a lot, a little, or not at all I can't say. But where there >>is smoke there is fire. Or in this case "trash". :) >> >>Auto232 is trash. > >There are mistakes with Auto232 but my impression is that the number of mistakes >is small and having games with auto232 is better than having no games at all. > >I guess that the difference in rating because of autoplayer errors is less than >20 elo. > >Uri OK.. in that case, didn't you mention the problem of Junior 5 not handling under-promotion correctly? I doubt that results in more than a 20 elo slip. Why was it important? Answer: You'd like to see it play _correctly_ all the time, as would I. Same for auto232. _any_ error is one too many, IMHO, if it can be eliminated.
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.