Originally posted by greenpawn34 Interesting - and you don't mind?
I believe you, but some of these things are very strong these days.
(I'm an old hacker who can remember when these things were naff)
What happened with rating points - did you get them back?
(I'm sure others would like to know this as well - I'm just curious).
Well I don't mind as long as the rating represent the strength of the computer. I would mind if I play against a 1200 player who suddenly use Fritz or else.
As far as I know, nobody got the rating points back when someone is banned for engine use.
There will never be a cheat detection system with 100% accuracy. We may get close, but never 100% as such. So supposing we have to compromise, what balance would people prefer?
A) a system that detects 99% of cheats, but 3% of the time wrongly identifies an innocent player as cheating
B) detects 80% of cheats, and 1% of the time is wrong in the innocent case
C) 50% of cheats, 0.5% in the innocent case
i.e. what is your preferred balance between putting up with cheats versus risking innocent players getting banned?
Originally posted by Varenka There will never be a cheat detection system with 100% accuracy. We may get close, but never 100% as such.
This is the real issue, isn't it? My understanding is that over a certain number of games it's impossible for any human player to matchup with a strong engine to a certain percentage level when you're out of database theory. If this is so, then it is in fact 100% accuracy.
But how many games and what percentage level? I'm not sure there's any statistics available regarding for example human world champions or modern days GM's. I've seen some experimenting with the games of a pre-computer days CC master in another thread, but that's about it.
Originally posted by Zipdrive Nightmare This is the real issue, isn't it? My understanding is that over a certain number of games it's impossible for any human player to matchup with a strong engine to a certain percentage level when you're out of database theory. If this is so, then it is in fact 100% accuracy.
But how many games and what percentage level? I'm not sure there's any ...[text shortened]... with the games of a pre-computer days CC master in another thread, but that's about it.
>> If this is so, then it is in fact 100% accuracy
I agree, but I was unclear in my initial post. When I say “cheat”, I mean all cheats, not just blatant cheats but also occassional cheats, etc. When considering the overall problem, it’s not easy to get to 100% in all cases.
I posted this game a long time ago Game 3689355 and said that I was suspicious of 25. ... Bxh2 because any experienced player would recognise that the bishop is trapped after 26. g3. (Yes, I know Fischer walked into a similar trap against Spassky, but he thought he could get his bishop out!).
At the time I thought I had worked out a way to win the bishop but I didn't need to play it as my opponent was banned. I can't find the winning line now.
Well I didn't see it coming... that means I'm not even an ordinary master...
This may be a stupid question but why didn't you just take his rook in move 38? I've looked for ages and I can't see the trap. Probably why my rating is in the 1400s
Originally posted by gearoidmm This may be a stupid question but why didn't you just take his rook in move 38? I've looked for ages and I can't see the trap. Probably why my rating is in the 1400s
In retrospect 39.Qxc2 would have been better. But then Black plays Bxa3 and is likely winning with two bishops, safer king, and a better pawn structure. Taking on f8 first seems powerful. If the bishop is recaptured then the Rc2 drops and white is a piece up. A better try for black is retreating the R, and then white has two pieces for the rook. In both cases White is a lot better. Alas, there is a third possibility. The kind of move that a computer would not miss but most players would, perhaps even some masters.
The point I was making is that all this needed to be seen when playing ...b6 several moves earlier. It's suggestive of computer use but by no means definitive.
24 Jul '08 22:55
Spotting a cheat move/game?
Back to Top