From | Message | ||
---|---|---|---|
|
![]() So, what does GK do about cheating? Particularly with regard to correspondence games (the only games I play here). I have read that if you suspect cheating, you can provide evidence and appeal to GK. What constitutes evidence? I was recently challenged by a player whose GK rating is over 2300. This player has played over 2500 games here, drawn four of them and lost exactly zero. Average opponent rating 1016. I chose not to accept the challenge, but I went through the player's game history and looked for games that had been analysed by the computer. Result: This player NEVER makes inaccuracies, mistakes or blunders. Simply NEVER. I think that's fishy (and pretty weird too. The whole m.o. - playing low-rated players to inflate your own rating is meaningless. But this player has beaten a few highly-rated players as well.) As an aside - I don't think cheating is a big problem with the correspondence games at GK. The people I play all lose plenty often. I figure cheaters will get tired of winning all the time and quit the game. There was someone named Tamkidd who went undefeated over about 50 games. He cleaned my clock too, never EVER made mistakes. Then quit playing altogether some time later and has since been deleted. |
||
|
![]() gameknot.com GK quite rightly has a very high standard of proof to be met before they will take forward complaints of cheating (I hope I’ve got that right, it’s based on the link above that they’ve provided.) Occasionally, you’ll strongly suspect a person is cheating, but all you can do really is put them on your ignore list. I did this once to a guy who was muddling along at around 1450 elo for years but then suddenly enjoyed a meteoric “improvement.” It’s not nice to be outfoxed by a computer pretending to be an expert. If you have strong evidence, report it here gameknot.com |
||
evader23 13-Dec-19, 15:19 |
![]() |
||
|
![]() Regarding mistakes/etc. as reported by post-game computer analysis -- it's a sliding scale. There is no point to report "inaccuracies" if the game is pretty much won for that player. For example, if a player is already a Rook up in a game, no reason to report a Pawn being lost in the shuffle due to a non-optimal move. So if a player is overwhelmingly stronger than most of their opponents, it wouldn't be unusual that the post-game analysis will report virtually no inaccuracies. And strong players don't make blunders, regardless of their opponents. As for why the player is only playing lower-rated players -- you would need to ask them directly. There could be any number of legitimate reasons for that. Perhaps they are trying to be helpful to the lower-rated players to practice playing vs stronger opponents. Generally speaking, higher-rated players rarely allow someone rated much lower to challenge them to a game, as they risk losing a lot of rating points if they make a mistake. |
||
|
![]() Intl Chess Club apparently has a program that identifies cheating. Does GK? My unbeaten player above may be legitimate, but I consider 2500+ wins again 0 losses to be fishy. (Even Magnus makes mistakes and loses games.) But I certainly don't have concrete evidence. |
||
|
![]() Lots of wins and no lost games is not unusual for players who only play against lower-rated opponents. Even if they make a (very occasional) mistake, they can likely recover and still draw or even win the game. Examples of concrete evidence of cheating: sudden (in a matter of weeks) and significant (400+ points) upward swing in the player's rating, after it has been established for a while; player routinely losing vs lower rated players in the past, and then becoming unbeatable vs higher rated players; significant discrepancy in player stats (e.g. tournament games vs regular games); player admitting to using a chess engine to analyse a game in progress -- "just wanted to make sure I still have a chance to win", "suspected that the opponent was cheating, so I wanted to confirm with a chess engine", etc., or "a friend" doing the same for the player. It takes a lot of time and server resources to investigate a player for cheating, so we cannot do it at a slightest suspicion. We do of course investigate all players that warrant a review. Also, in our experience, it is also highly unusual for someone to avoid detection for any significant period of time. You have to be very careful and analyse every single move, because even a single mistake can result in a lost game at any point. People also don't enjoy the hollow victory of using a chess engine to get a high rating for any significant period of time, so cheaters tend to quit very quickly. Or they get caught very quickly. The end result is that the fear of playing vs someone using a chess engine is way overblown. It is a slightly different story at the very top of the rating range, but for 99.9% of players, if your opponent has been playing for more than several months and their rating is under 2200-2400, it would be *very* unlikely that they are using a chess engine. It might feel sometimes that your opponent is cheating because they are winning, but you should be focusing on learning from the experience instead of taking the easy way out and blaming the loss on your opponent cheating. Not directed at anyone in particular, just a general observation, of course. |
||
|
![]() |
||
|
![]() |
||
yon_cassius 12-Jun-20, 02:57 |
![]() Or a player might have just knuckled down; removed a lot of their bad habits; and studied the game - something I ought to try at some point! |
||
|
![]() |
||
|
![]() That's not entirely correct - some players are very inconsistent. I have weeks when I play where I see everything on the board and weeks when I hang a piece in every game I play because I miss the most basic combinations. My rating fluctuates by a couple hundred points each cycle, and it's so bad several years ago GK warned me against "rating manipulation". I also timeout when I feel burnt out from chess that I have take a long break, and this drops my rating as well (they are the huge dips followed by a long flatline of inactivity in my graph). For instance, I've just beaten a player in his 1900s in less than 20 moves (granted, he made a couple of terrible mistakes which I was fortunate enough to see), however, a few days earlier I blundered this: gameknot.com If these 2 things happen in separate cycles, I could easily be dropping 200 points, gaining them back, etc. |
||
|
![]() |
||
|
![]() As a Team Captain we have signs we look for , all Captains have their own system .. but suspicion isn't proof , you need to take so many more factors into consideration , If somebody's play ever makes you feel uncomfortable I can only suggest you ignore them .. |
||
|
![]() It seems if your rated over elo2000 they check your moves for suspicious activity .... as in the perfect game , which always seems to happen against me These players are then removed from the site ... (unfortunately they can rejoin under a different name lol ) |
||
yon_cassius 25-Aug-20, 08:11 |
![]() <chuckle> As though I'M keeping my rating under 2000, it's my opponents that are keeping my rating down! Seriously though, if someone was playing at that site (and not cheating) they would want to think about whether it was worth hitting 2000 or not. |
||
|
![]() |
||
|
![]() The posts by the "caught cheaters" are rather funny ... " I used swordfish because I thought everyone else was... I'm so sorry " blah blah .... probably because they were caught lol Dave... your logic wasn't flawed it's just hard to prove ... |
||
|
![]() Joking aside, after the A level results fiasco here in the UK, I’d be wary of sites that rely heavily on algorithms and suchlike. |
||
|
![]() |
||
|
![]() I received a notification from a site I... er ... no longer use It's on going probably due to questions on their forum Top players play perfectly against me because I make perfect mistakes ... |
||
wanstronian2 07-Sep-20, 06:38 |
![]() The good news is that he/she has gone so stratospheric (2036 after 15 games), even if I lose my rating will increase |