1 edit
Although I agree that GauravV's games are highly suspicious. I also agree with Tengu that making the methods of discovering cheaters secret will make people feel uncomfortable, especially genuinely strong players who are worried that they will be suspected.
I am aware of the argument that revealing the methods will help cheaters to make it "appear" that they are not, but how do we know they aren't able to do that anyway? Some people might have guessed how moderators detect them and taken action against it.
If you get falsely accused of murder, you would expect to see evidence. Not the judge say, "Sorry, we can't reveal our methods, otherwise it would help other people to commit murder and get away with it".
Originally posted by lauseyLook come on, if kasparov told the world how he studied then all his opponents would be able to prepare for him! The police don't disclose how they catch pedophiles, if they did then loads of them would get away with it!! Take a look at the games Sicillian smaug posted, it doesn't take a genious to work out that he was BLATANTLY cheating! Look at all the tournie wins. take a peak at the "tournament players hall of fame" thread in the tournaments forum. He wasn't even on it 8 months ago, now he's top! He won like 20 tournaments in 6 months...
Although I agree that GauravV's games are highly suspicious. I also agree with Tengu that making the methods of discovering cheaters secret will make people feel uncomfortable, especially genuinely strong players who are worried that they will be suspected.
I am aware of the argument that revealing the methods will help cheaters to make it "appear" that th ...[text shortened]... our methods, otherwise it would help other people to commit murder and get away with it".
if you match up 85%> of the time, your in trouble. Not even Kramnik, Topalov, Fischer Botvinnik can match up with computers that much.
Kramnik is the world champion and Danailov accused him of 87% matchup rate, which would have made a compelling arguement to have Kramnik booted if it wernt for Danailov manipulating the #s
1 edit
Originally posted by MarinkatombLike I said, GauravV was particularly obvious to everyone observing. Just that with some players banned it might not be so obvious. I am just wondering if the moderators use the same methods of detecting cheaters in the same way that we suspect (e.g. starting off very bad and suddenly improving, as well as matching up a lot of the time with engines). If that is the case, then I cannot see what would be the harm in revealing those methods, as it would be what we know already.
Look come on, if kasparov told the world how he studied then all his opponents would be able to prepare for him! The police don't disclose how they catch pedophiles, if they did then loads of them would get away with it!! Take a look at the games Sicillian smaug posted, it doesn't take a genious to work out that he was BLATANTLY cheating! Look at all the He wasn't even on it 8 months ago, now he's top! He won like 20 tournaments in 6 months...
As for pedophiles getting caught. The evidence will be revealed AFTER they have been caught.
Originally posted by Sicilian SmaugDANG-IT! I just resigned two games to him! ðŸ˜
Have a look at a few of his early games on the site:
Game 470644
Game 464816
Game 491449
His graph only shows last 300 games, if it showed the whole history the 'improvement' would be shown as the dodge fest it really is.
Of course this is not proof in itself but I'm sure the game moderators will have used the usual method o ...[text shortened]... the time this has taken. People have been raising there suspicions about him since 2004.
1 edit
Originally posted by HomerJSimpsonPerhaps in the 13% that don't match up Kramnik found an even better move!
if you match up 85%> of the time, your in trouble. Not even Kramnik, Topalov, Fischer Botvinnik can match up with computers that much.
Kramnik is the world champion and Danailov accused him of 87% matchup rate, which would have made a compelling arguement to have Kramnik booted if it wernt for Danailov manipulating the #s
Edit: How long would you need to leave Fritz on to calculate moves at Kramniks level? 10 minutes a move / 15 minutes. Clearly this would be impossible with 100 games. What level will Fritz play if given no more than 1 minute per move (surely it won't found the best move unless obvious then and if obvious won't most reasonably strong human players find it). I imagine there can be long sequences in a game that will match up because the moves are obviously best - in openings there will be a lot of matches because we can use opening books. The percentage matching up will change based on thinking time so it should not be raw percentages that count (except perhaps over a very long sequence of play) but rather more subtle methods such as that resignation is a drawn position (that is beyond the computers horizon), playing on in an obviously lost book ending (again beyond the computers horizon), an exchange of material that wins material now but which creates an obvious long term strategic weakness that a strong human player would avoid (I think I see this in Scotts game that I mentioned earlier that Scott won), obviously complex and non intuiative tactical combinations that win but which even the strongest players would miss.
Individually these are hard to assess but over a reasonable number of games the number of occurrances of such instances would point to engine abuse.
Perhaps the moderators place a cookie on your computer watching what programmes you are running. This might be illegal and perhaps the Firewall would pick it up but it would be meaningless anyway as you could be using a Opening Database (legal) or analysing past games (legal) or you could simply use another computer.
1 edit
Originally posted by HomerJSimpsonI agree with you. Although No1 accused me once of using assistance in the only game I beat him at. I put it in Fritz afterwards to see if he was correct and it did have a high percentage of matched moves. However, both of us followed the database for about 20 moves.
if you match up 85%> of the time, your in trouble. Not even Kramnik, Topalov, Fischer Botvinnik can match up with computers that much.
Kramnik is the world champion and Danailov accused him of 87% matchup rate, which would have made a compelling arguement to have Kramnik booted if it wernt for Danailov manipulating the #s
Originally posted by arrakisIf you play enough games there will be the odd game with a high percentage of matches surely? Just not every game or even the majority of games.
I agree with you. Although No1 accused me once of using assistance in the only game I beat him at. I put it in Fritz afterwards to see if he was correct and it did have a high percentage of matched moves. However, both of us followed the database for about 20 moves.
Originally posted by Dragon FireA minor point as you revealed it wouldn't be a viable option anyway, but cookies cannot detect what applications you are running.
Perhaps the moderators place a cookie on your computer watching what programmes you are running. This might be illegal and perhaps the Firewall would pick it up but it would be meaningless anyway as you could be using a Opening Database (legal) or analysing past games (legal) or you could simply use another computer.
A cookie file is just pure data which websites can read/create/modify. Unless software applications are specifically programmed to create or modify certain cookies (which I very much doubt Fritz or any other chess engine will do), they will be quite useless.
Also, creating cookies isn't illegal. What would be illegal is to hack into someone's PC to install executable code (which can possibly detect what applications are running).
2 edits
Originally posted by arrakisMatch up rates in themselves don't really tell you all that much, I'm pretty convinced most people on this site will agree with Fritz or Crafty or whatever on what move to play next as white in this position:
I agree with you. Although No1 accused me once of using assistance in the only game I beat him at. I put it in Fritz afterwards to see if he was correct and it did have a high percentage of matched moves. However, both of us followed the database for about 20 moves.
I assume that the games mods have some way of assessing how easy a move is to find - where the example above counts as very easy...
In some ways the clearest fingerprint of engine use is in the mistakes they make, for example the null move heuristic fails in zugzwang positions.
Edit: I played GuaravV a number of times and lost each game - I remember putting some effort into the games, and its annoying that my work was wasted on an imbecile with an engine.