Author: Robert Hyatt
Date: 19:01:26 11/15/00
Go up one level in this thread
On November 15, 2000 at 17:09:30, Bruce Moreland wrote: >On November 15, 2000 at 16:44:27, Robert Hyatt wrote: > >>On November 15, 2000 at 12:57:54, Bruce Moreland wrote: >> >>>If you want to use numbers, be more careful with the numbers. You could have >>>said it the way it is, or you could have understated the distance between the >>>ratings, or you could have overstated the distance. >>> >>>It doesn't help your case that you start out by overstating. Was the computer's >>>rating 3000? Was the human's rating 1800? This can easily be verified by >>>looking at the ICC history. >>> >>>The human's rating was 1884, prior to the game. The computer's rating was 2645. >>> This is a big delta, but not the 1200 points you suggest. It is 761 points. >> >>I generally consider _all_ decent computers to be around 3000 on ICC. Some >>are not due to operator actions such as switching hardware, etc. But if they >>all played for a good long while, I can't imagine any really staying below >>3000 or so... > >The program runs on a 450 mhz machine and the account will play low-rated >people. > >>>If you look today, which you can do, nobody is stopping you, the human is up to >>>2001 points. He beat Hossa in a 4 0 and he beat another human near his rating >>>in several games of 3 0. >>> >>>The game against Hossa is appended at the end of this post. It is a human game. >> >>I don't doubt that. However, as I said before, this has nothing to do with >>the first game discussed. Cheating on one of every N games is not uncommon. >>And when I originally wrote that first post, 'cheating' wasn't really the >>issue. IE I wasn't trying to label the human as a cheater. I was trying to >>point out that drawing a conclusion from an 1800 player beating a computer is >>not always safe, because many times, that 1800 player has some silicon behind >>him... > >http://www.icdchess.com/forums/1/message.shtml?138536 > >Your first post on the issue. > >"It is highly likely that the opponent used a computer here. I base this on a >couple of things. Near the end, there is a deep mate. He played it >_perfectly_. Which I don't think an 1800 player could do. If I were betting, I >would bet that white is a computer." > >and > >"However, on the other hand, Tiger was playing a computer in human clothing. >Nothing good can come of that, and drawing conclusions is harder. I would >_never_ believe than an 1800 player can beat today's programs." > >There is more. The whole post is essentially an accusation that the guy was >using a computer. > No it isn't, but I'm not going to argue about it further. The whole post was about the fact that I suggested that rather than assuming tiger played like a patzer, perhaps the opponent was something other than what was assumed. I would bet, in 99% of such cases, that rather than the computer playing like an idiot, the opponent was cheating, when such games show up on the chess servers. That leaves 1% for legit wins, and that is rare enough to fit the Elo stats reasonably. I really wasn't trying to pin the guy as a cheater, although I do agree it sounded like it. I am neither convinced he did or didn't, in fact. I would need more data. If you think I "was after him" so be it. I didn't intend that. >>The thread diverged at that point, partly my fault no doubt, because I didn't >>word things very carefully. But the above was my original point, _not_ the >>fact that I think the human cheated. I ran a quick tactical check, on a single >>game, didn't find any errors, and suggested that this _could_ be a computer >>rather than a human. One game isn't enough. But for the original point of >>the thread, it was a (to me) reasonable remark to make... > >See above. > >>>>>And the guy's ICC name was featured in the original post, so he is being tarred >>>>>with this. >>>> >>>> >>>>I didn't look at the PGN, which was unfortunate. And you are right in that >>>>respect. However, I am also not against calling a spade a spade, and in this >>>>game, to me, based on my (and computer) analysis, it looked a bit fishy. It >>>>might be perfectly legit. Or it might not. >>> >>>This is insane, Bob. Yesterday you as much as admited that your "mate in 7" >>>argument doesn't fly, because you actually looked at the end of the game and >>>realized that a C player could play those moves. And you looked at the rest of >>>the game, and the other game I showed. What, other than the result, are you >>>having a hard time swallowing? >> >>The fact that for 30+ moves I found _zero_ tactical mistakes of any kind, in >>a fairly fast blitz game. Unusual. Very unusual. > >There were only a few actual moves in the game. He lumped around in the >middlegame and then the computer killed itself. > >>>>>The guy isn't a cheater. He's someone who played a decent attack against a >>>>>computer and it worked. >>>> >>>> >>>>I don't think you can conclude that from one game, any more than I could >>>>conclude that he did cheat. >>> >>>Things happen but you can't start talking about evil and devious intent until >>>you have some reason to do it. >> >>I didn't suggest evil or devious intent at all. I pointed out that when a >>computer loses to a supposedly weak human, quite often things are not what they >>seem. And concluding that Tiger had therefore played weakly might not be a >>good assumption. IE when I wrote the post, the subject was "did tiger play >>bad enough to get beat by a weaker human?" Not "did this human cheat to beat >>tiger?" I merely suggested that this latter conclusion is possible, so that >>the original conclusion about tiger playing weakly could be seen "in context". > >Just look at the game. It is very obvious what happened. > >bruce
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.