Well something has to be done because Cheating on chess.com at the higher levels is a major issue stated above by pod1000
I have the Solution to stop cheating in Live Chess!
Sort:
I think having a computer study games by people with a high number of points and a high percentage of wins is a lot simpler, doesn't involve putting more engines to play and could thus prove more efficient, with fewer side-effects.
I actually quite like the idea put forward by TheGrobe: to use a consistent java notification of a loss of window.focus (signifying that the Live Chess window has been minimised), in conjunction with a rating in excess of, say, 1800, to act as a red flag for staff to employ their usual cheat-checking methods.
It should be more obvious if someone's cheating, given the java tool described above, and the shorter time controls (such that if someone's playing at a very strong level, consistent with engine use, it would be impossible to put it down to having spent half an hour analysing each move).
Something is being done -- there is a current program in place to detect and eject cheaters.
The truth is that I don't actually think cheating is actually all that rampant, certainly not at the level at which most of us are playing: Cheaters will quickly propell themselves into the highest rating bracket and once they're there will certainly be subject to more scrutiny by the detection program.
I'm confident that chess.com has an effective program in place to address cheating at this level, and irrespective of whether they do or do not, because it's ocurring at that level it's not affecting me or 99% of the players here.
I'm also confident that chess.com is continually (and quietly) improving on this detection program. I suspect that they silently monitor threads such as this one and pull ideas into their aresnal (such as the focus change detection proposal made here, which I think is a powerful yet easy indicator to gather).
Finally, I'm also confident that they discard many, many more ideas that simply aren't workable. I'm afraid that yours happens to be one of those.
gabrielconroy wrote:
I actually quite like the idea put forward by TheGrobe: to use a consistent java notification of a loss of window.focus (signifying that the Live Chess window has been minimised), in conjunction with a rating in excess of, say, 1800, to act as a red flag for staff to employ their usual cheat-checking methods.
It should be more obvious if someone's cheating, given the java tool described above, and the shorter time controls (such that if someone's playing at a very strong level, consistent with engine use, it would be impossible to put it down to having spent half an hour analysing each move).
Credit where credit is due: Someone else proposed it -- I just clarified the technical feasibility.
Mar 24, 2009
0
#167
pod1000 wrote:
... Recommendations
* The general reporting of engine users or suspects needs to be made much easier.
* Admins should be able to run software to get an idea of matching very quickly of a persons game to an engine.
* The time taken between moves in the opening and endgame should be checked for a strange consistency. Opening moves, and endgames generally are played much quicker by experienced players in blitz. Usually it is only critical situations which require much of a pause in 1-5 minute chess.
* As a final crticial test, the original proposal in this thread may be useful for confirmation of existing evidence of an engine suspect.
I'm no expert on LiveChess; I've played about 3 games. I have no idea (a) how many people get banned from chess.com, nor (b) how many of those who are banned - have been banned for chess engine use, nor (c) any idea of the number or proportion of chess.com members who cheat by using engines, nor (d) the number of my opponents who have cheated - if any. Some chess.com staff will have access to the figues (a) and (b), but (c) and (d) are unknown and unknowable.
I do know something about technology, on the other hand, so I commented on why the proposed technological solutions would not be effective.
(1) pod1000 asserts there are a large proportion of engine users in LiveChess - but he offers no evidence.
(2) He suggests that resporting suspects be made easier; an extra button alongside RESIGN, perhaps, labelled SUSPECT? may be what he has in mind. The current mechanism is to use the Report Abuse link (found at the bottom of every page); you have to supply the name of the member you are reporting, a link to a game and some justification. Not much overhead - but just enough, perhaps, to discourage hotheads from clicking on a SUSPECT? button every time they lose a game.
(3) pod1000 suggests chess.com staff should use the appropriate technology to detect engine use - by comparing game moves with engine recommendations. From Erik's assurances, they are already doing what they can, but naturally, no-one outside chess.com's staff knows the specifics.
(4) pod1000 asserts that the move timing pattern constitutes an engine-use fingerprint. I frequently do other things at the same time as playing Turn-based Chess, so I imagine the same phenomenon might be true for LiveChess. If so, there could easily be a lag of 30 seconds per move as they switch between tasks.
In conclusion, I fail to see how anyone on the outside (of chess.com staff) has enough information to either scope the problem objectively, or know which methods are currently in use.
Mar 24, 2009
0
#168
pod1000 wrote:
(1) pod1000 asserts there are a large proportion of engine users in LiveChess - but he offers no evidence.
On the contrary I have made effort to report people, and at least one person was banned in the last two months as a direct result of one of my reports. He was over 2200 on 3 minute chess and he used things like the exchange french in the opening. If you can see my past game history, then check out where I lost about 4 3 minute games in a row, without much of a chance. Action was taken, but the process was painful. Ideally, players shouldn't have to be proactively collecting evidence by being completely crushed, in order to then find some way of reporting the details of the game. Ideally he would be removed as soon as his games could be analysed proactively.
"(2) He suggests that resporting suspects be made easier; an extra button alongside RESIGN, perhaps, labelled SUSPECT? "
Now you are choosing an implementation of an *idea* which carries a lot of overhead for everyone involved. What about trying to reduce the *need* for reporting in the first place, by automatically running existing resources like the Computer post-mortem tool. on players say >1800 as they complete games of 3 minute chess. If their play is too perfect all the time, then that should raise alarm bells. Maybe such a post-mortem of 3 minute games after they have played would reduce overheads.
But also maybe in your implementation proposal dreamt up just for the sake of flogging the concept of "easier reporting", could be qualified by both players being over 1800 only. That is if it is not possible to do without the reporting model.
"(4) pod1000 asserts that the move timing pattern constitutes an engine-use fingerprint. I frequently do other things at the same time as playing Turn-based Chess, so I imagine the same phenomenon might be true for LiveChess. If so, there could easily be a lag of 30 seconds per move as they switch between tasks."
I was talking about 1 to 5 minute games where time is absolutely critical!
"In conclusion, I fail to see how anyone on the outside (of chess.com staff) has enough information to either scope the problem objectively, or know which methods are currently in use."
Serious blitz chess players have played on other sites like ICC, and playchess for several years. On those sites, such as playchess, there is very visible alerting for a computer abuser, and a public announcement of their rating being reset. You just need to play there for an evening to witness that. On ICC there are several people made into (C) accounts. Again very visible.
Coming back to the main concept which I think immortal gamer is alluding to: It would be nice for strong players e.g. lets take IM Andrew Martin, to play on the 3 minute chess facility, and not be completely crushed. It would be nice if he wasn't a "guinea pig" and have to report being crushed after the fact. As well as damaging to his ego, he probably wouldn't play here again. Do you think titled players like to be used as guinea pigs for Yahoo style kids with powerful computers?
You totally understand where I'm coming from and perhaps it is because I was a member on ICC for so long that I want a better live chess area when it comes to cheating.
Like I already said, no one is denying that Live Chess has people cheating on it. No one is denying that it needs to be sorted out.
The idea of having sanctioned people roaming around deliberately cheating as a way of testing other players (like it's Mad Max, or something) is ridiculous.
The probem with chess online (& the same problem happens with Tiger Woods online golf which I also play) is that there's a rating system, which is good, but it encourages people with low self esteem to cheat, because they have to have good ratings to feel good. I've come to relalize that there is no way around this. I am a novice player, so I just play with people around my ratings, because if those people were cheating, their rating would be high. It's unfortunate for the good players whose ratings are high, that you will always have the stigma from others who wonder if you are really good online or are using a chess program to hep you get high ratings. I never use my chessmaster program while I am playing any game, as I do not want a computer to be player the game for me, to me that would be pointless.
Mar 24, 2009
0
#171
gabrielconroy wrote:
The idea of having sanctioned people roaming around deliberately cheating as a way of testing other players (like it's Mad Max, or something) is ridiculous.
Then why would police officers pose as prostitues or drug dealers to catch the actual criminals?
I'm not saying the chess is as important as those things...just making a point on what works.
The police officers who pose as prostitues and drug dealers don't actually sell sex or drugs to the general public.
Mar 24, 2009
0
#173
Yes they do. They just make arrests (e.g cancel accounts) before the perp gets away.
Painterroy> I am a novice player, so I just play with people around my ratings, because if those people were cheating, their rating would be high.
Correct. I've won 19 out of my last 20 games, and that loss was to a human. My experiences here lead me to believe the population of engine-users below 1500 is tiny.
Mar 24, 2009
0
#175
immortalgamer wrote:
Yes they do. They just make arrests (e.g cancel accounts) before the perp gets away.
The cop doesn't have sexual intercourse. They arrest them before that. This idea engages in approved cheating before it can make a determination of whether there is wrong doing or not.
Innocent people will get utterly crushed and then later told "Sorry, we had to see if you were cheating."
I don't see how anyone cannot see how retarded this is.
Mar 24, 2009
0
#176
I don't know how many freaking times I've said this: If the person is not cheating the game will be resigned automatically...
i personally can't believe that this is a problem! what kind of a loser would you have to be to come on a site that is purely for the love of chess and cheat??!!
they might win a few games cheating but then get them on an otb and they would crumble. whats the point?????
Mar 24, 2009
0
#178
You'd need to play quite a few moves to even get out of the opening. You gonna call someone a cheater just for knowing the first 15 book moves? Where do you draw the line? If you have a mating attack you resign?
You are definitely more worried about cheaters than you should be. Your rating on here is completely meaningless. Play over the board if you want a legitimate game.
You've missed my point:
The analogy is flawed because everyone who buys drugs or sex is committing a crime, so the undercover police officers only deal with people who are offenders. If you're transacting with one, you're guilty by definition.
Here, there are legitimate transactions in the mix (games between two non-cheaters). To extend your analogy to better fit your proposal, it's as though undercover cops went around offering people cigarrettes that were actually drugs, but only arrested those that agreed to buy more.
People who just wanted a cigarette (i.e. a game with a real human) will end up unwillingly doing drugs (i.e. having a game with an engine). Not a very good way to fight the war on drugs....
Mar 24, 2009
0
#180
mwill wrote:
You'd need to play quite a few moves to even get out of the opening. You gonna call someone a cheater just for knowing the first 15 book moves? Where do you draw the line? If you have a mating attack you resign?
You are definitely more worried about cheaters than you should be. Your rating on here is completely meaningless. Play over the board if you want a legitimate game.
oh brother...use your common sense. It's about creating a good chess.com environment.
No one disagrees it's a problem, and that it needs to be addressed; I just don't agree that using a "kill squad" (what is, the favella?) is a good solution.