18 Aug 07
If you have an engine like Fritz or Rybka, how do you use it to screen for engine users? When people say that a certain persons moves matched up with an engines by 98% or so exactly how is that calculated? What time settings are used? Is there an easier way to do it or does someone actually put a game in move by move and see how the engine replies?
18 Aug 07
Originally posted by amolv06Careful. If you admit to using an engine to check on your opponent during a game you are as likely to be banned as the alleged cheat. DO NOT under any circumstances use that engine on a game in progress.
If you have an engine like Fritz or Rybka, how do you use it to screen for engine users? When people say that a certain persons moves matched up with an engines by 98% or so exactly how is that calculated? What time settings are used? Is there an easier way to do it or does someone actually put a game in move by move and see how the engine replies?
18 Aug 07
Originally posted by KeplerI'll keep that in mind. However I'd like to be able to go over some people's games after I play. When someone who has consistently been rated 500-600 points below me completely crushes me and I don't see any obvious mistakes on my part it makes me somewhat suspicious.
Careful. If you admit to using an engine to check on your opponent during a game you are as likely to be banned as the alleged cheat. DO NOT under any circumstances use that engine on a game in progress.
18 Aug 07
1 edit
Originally posted by amolv06no it isn't. you're a 1500, which means you make obvious mistakes in every game. if they're not obvious to you, look harder. which would be extremely beneficial, as the mistakes you have problems seeing are likely to be the ones that keep you in the 1500's.
When someone who has consistently been rated 500-600 points below me completely crushes me and I don't see any obvious mistakes on my part it makes me somewhat suspicious.
basically, the way you check for engine use is first you count out database moves, maybe also forced moves, then analyze the rest with an engine and see what you've got. bear in mind that a single game doesn't necessarily mean anything, no matter what the matchup rate. but if you think you have some kind of evidence, send it to the game mods and they'll do the proper investigation.
also, you have to pick up a rigid time limit for analyzing a single move, and stick to it no matter what. especially if you use 'infinite analysis' or comparable. the reason is that if you analyze 60s per move, and at 63s the evaluation suddenly jumps to engine choice #1, if you now count that as a match, you'll be statistically biasing the results. cherry picking. -so you need to discard the match at 63s to get meaningful evidence.
I wouldn't even bother to check anyone under 2000 though, it's just paranoia when you suspect weaker players.
19 Aug 07
Originally posted by amolv06sometimes those people have sketchy internet access, and time out alot, meaning there real rating is higher.
I'll keep that in mind. However I'd like to be able to go over some people's games after I play. When someone who has consistently been rated 500-600 points below me completely crushes me and I don't see any obvious mistakes on my part it makes me somewhat suspicious.
Also, I've been beat by lower ranked players when I thought I made no "obvious" mistakes, but after careful review I find that I had.
19 Aug 07
Originally posted by wormwoodDid these "over 2000 players" join RHP at that rating?
I wouldn't even bother to check anyone under 2000 though, it's just paranoia when you suspect weaker players.
Under 2000s have been banned for engine use so I disagree with the paranoia theory. It's a fact.
Also, contrary to some peoples belief, one doesn't necessarily have to be of a high rating yourself to spot the engine users..
19 Aug 07
1 edit
Originally posted by Dr Strangelove"Originally posted by wormwood :
Did these "over 2000 players" join RHP at that rating?
Under 2000s have been banned for engine use so I disagree with the paranoia theory. It's a fact.
Also, contrary to some peoples belief, one doesn't necessarily have to be of a high rating yourself to spot the engine users..
I wouldn't even bother ..."
it was a personal opinion based on experience.
of course everybody starts at 1200, that's trivial and not even worth a mention. excluding the few self-confessed low rated 3bs, the lowest one I've seen in two years was tolikcheturi at 1800+, but the vast majority of them have been 2000+. if you fancy chasing ghosts, analyze the games of 1200's all you want. I wouldn't bother because it would be wasted effort 99.9% of times. the ones I've checked have turned out to be cheaters about 20-30% of times. there's quite a huge difference there.
19 Aug 07
i agree with strangelove.....i like it when they cheat.....when my woman sets the board up......all cheaters and talkers get crushed by me.....anyone better than me....1500 otb...never talks there game up......so when i hear a guy talkin trash....i say...."let's play"....
I win......strong players otb....don't talk trash....they prefer to let the board and there moves do the talking...
19 Aug 07
Originally posted by amolv06People rated 500-600 points below you can beat you from time to time but if they are rated that low isn't it highly unlikely that they are using an engine.
I'll keep that in mind. However I'd like to be able to go over some people's games after I play. When someone who has consistently been rated 500-600 points below me completely crushes me and I don't see any obvious mistakes on my part it makes me somewhat suspicious.
If they are that far below you and unbeaten the rating is of no consequence as it is not yet their true rating but if it has stabalised at that level my guess would be they had a good game and you made a mistake.
The only indication of engine use would be if they were rated 1300 for ages and suddenly show a sustained improvement (coupled with) a totally different style of play. Real people tend show gradual improvement over a long period with the occasional set back.