Cheating at chess how to counter it - Page 2

bsquared · 2011-06-10, 21:35

Quote:

Originally Posted by retina

Some people just seem to be concentrating upon getting some arbitrary number as high as they can, in whatever way they can...

Yeah, I thought that's what you were getting at. And when it's something harmless like post counts or "online notoriety" I agree completely with the:

.

Quote:

Originally Posted by retina

My thought is if people really want to cheat like that then just let them at it. There appears to be no harm done to anyone.

There's the rub. As Brian-E said, a high Rating via cheating in some cases can win someone money or fame, invitations to more tournaments, etc. Then there is harm done.

retina · 2011-06-10, 21:48

Quote:

Originally Posted by bsquared

There's the rub. As Brian-E said, a high Rating via cheating in some cases can win someone money or fame, invitations to more tournaments, etc. Then there is harm done.

Oh, so people can use a computer to play chess online and win money. I never realised there was a lucrative usage model there. In that case it is trivial to cheat, and likely almost impossible to detect the smart cheater. Unless I am sitting opposite someone in front of a board I would never put any money at stake.

So the original topic is how to detect cheating. I think that you could only ever detect the stupid and lazy cheats. I expect many of the posters here could, if they were so inclined, avoid detection relatively easily with a little patience and thought. And if there is money at stake then I expect you will have many more cheats than you might have first imagined.

A gestalt of man and computer together would be a pretty hard opponent to overcome by either man or computer alone.

Batalov · 2011-06-10, 22:38

Google for "слюсарчук рыбка Subtitled video".

There was a hilarious piece of pseudo-news, just a month and 1/2 ago.
In short, an amateur played blindly against Rybka under video and won; he was allegedly searched down to underwear before the match; the Rybka was ostensibly set to some silly level and played on some slow 1-cpu computer in the auditorium where this event was set up.

The same charlatan had previously claimed that he remembered 30 million Pi digits or something silly like that. He also is reported to be a 'neurosurgeon'.

Christenson · 2011-06-11, 02:59

And they completely missed the security guard completing the steganographic communications chain from the grandmaster watching the game telegraphing every move by the pattern of footfalls he set nearby on the wooden floor with his boots.

wblipp · 2011-06-11, 03:35

Quote:

Originally Posted by retina

Some people just seem to be concentrating upon getting some arbitrary number as high as they can, in whatever way they can.

In factoring, these are the BOINC participants. Our challenge is to find ways to get them to factor numbers. Yoyo@home, RSALS, and NFS@Home are all pretty effective at this. A few more ideas that don't cannibalize from these would be welcome. I don't see a way to make this work in chess, though - to achieve some goal by encouraging the cheaters to create high meaningless rankings.

ellipse · 2011-06-11, 07:22

I may not know how to prevent cheating at chess, but I do know how to cheat in online chess and get away with it. Just follow these rules:

1.) If possible, use two separate computers: one for making the moves on the chess server, and one for running the chess engine. If you only have one computer, rename the program you're using ("Fritz 12.exe" may be renamed to "scanner2.exe"), and don't run it on all the cores of the computer. But remember that some chess servers don't just check for chess programs and computer resource usage, they also detect excessive switching of windows. You've been warned.

2.) Don't rely on the chess engine to make the first few moves; do them yourself, then copy that position into your chess engine. People will wonder why you take more than 3 seconds to make a simple move, and it'll look even more fishy if it happens a few times in a game. So if there's an obvious move, make the move before you put it into the chess program.

3.) In the middlegame, try to vary the length of time you take to make the moves. No human player takes exactly 30-32 seconds to move for 10+ moves in a row, and people know that.

4.) Avoid playing very short games (called "lightning chess" or "bullet chess"). It's too hard to use engine assistance there, and you'll only attract unwanted attention if you have a large rating difference between games with short time controls and games with longer time controls.

5.) Stay away from tournaments with cash prizes. Not only will cheating detection efforts be ramped up during those tournaments, but other players will also be on the alert for potential cheaters. Besides, if you somehow manage to have a bad day and get caught, you won't want to deal with the legal problems.

6.) Use different engines throughout the game. For example, you could use Stockfish in moves 11-18, Rybka in moves 19-25, and Houdini for the rest of the game. In your next game, switch the engines and the move numbers. Naum could be used for moves 10-22, Shredder could be used for moves 23-30, and you can finish the game with an older version of Stockfish. This prevents you from having a high "hit rate" when others try to compare your moves to that of common chess programs.

7.) Don't get greedy and play a whole bunch of games a day. Chess server admins and experienced players know that masters and GM's won't sit at a computer for several hours in a row just to beat some anonymous, random people on the internet. And don't play one or more games every day either; take a break now and then.

8.) If you want to cheat, do it from the start. Don't start cheating on the username that you have previously played honest games on, because a sudden rating increase will attract suspicion.

9.) Avoid taunting opponents, and keep chess-related conversation to a minimum. This should be common sense, but you'd be surprised at the number of cheaters who've said things like "You're gonna lose. Mate in 17" or have usernames like "Unbeatable Chess God", only to be reported and caught later on.

10.) Allow no more than one rematch from the same opponent. The more times you let your opponent play you, the more chances he has of finding out that you're cheating. Just say something like "I'd love to play you again, but I have to do (insert activity here)."

11.) If you do get caught, don't just make up another username and try again. Wait for at least a month, because the chess server's engine detection team will now be on high alert for other cheaters. In the meantime, try to figure out what you did that got you caught. Did you offer a draw in a clearly winning position because your engine crashed? Did you successfully analyze complex positions in a matter of seconds in a lot of your games? Or did you use the same engine too many times?

Christenson · 2011-06-11, 14:28

I think questions of motivation are far more interesting than how to do a good job of cheating -- as I said, if you really want to cheat, hire a poor Grandmaster to be your double..

And by the way, I am among those seeking to maximise my GIMPS rating....but within the limits imposed by the seven PCs I have available. The GT440 has been most effective, when applied to TF...

I think it is natural to want to increase one's reputation as much as possible. From that, flows other benefits...especially if the reputation is measured in money...like the best of everything, relatively longer life, etc.

Cheating, as long as it goes uncaught, also increases one's reputation. My younger brother once spent some time in jail and noticed that he was the smartest person there.

I think that the trolls/mischeif makers are all trying to increase their reputations, because of the evolutionary forces involved. I think that places like sci.math don't do anything effective about ensuring that reputation is based on any effective behavior, and the cheaters don't have the internal smarts to recognize that it's going nowhere.

There's also the identity disconnect....you all think my name is "Eric Christenson", but there's only one random person on this board (Bdodson) who has actually met me, and that was years ago. Mr Blazys(sp?) can just walk away from his unbounded crank score, and it won't flow back into his other life, whatever that may be. Anyone else remember "Ludwig Plutonium" and "Time has inertia"?

In addition, I have a little bit of personal experience with a crank inventor. I can tell you that he got to do some fun stuff that he never would have been able to do if he had not "invented" his perpetual motion machine, which consisted of an alternator attached to the wheels of an electric car to recharge alternating halves of the main battery. He also met people such as myself who he would never have met except for this invention; the patent office even granted him a patent, since he carefully avoided using the words "perpetual motion".

What he didn't realize is that we all (or possibly just me, but I told everyone else) knew that the BEST way to extend the main battery life is to use the driving motor as the generator, just like on-the-market hybrids do, and then to start looking for the remaining losses. But that's hard work; the guy got a long way before he ran into someone(me) that said "wait a minute...this is going nowhere!". The car is actually built (it's a rusty pickup truck with a bunch of batteries across the front of the bed, gets 40 miles per charge, and has an operational controller worth $5K from my company that the inventor didn't pay for).

As for getting participants in yoyo@home, etc...I think the right question is how to get participants in "useless" math programs. Somehow, I don't think the back story of GH Hardy fleeing applied math, only to find out that number theory mattered 50 years later, is going to help, even if the theme was repeated when radioactivity was studied.

There's also the difficulty of the math itself. So we are left with various scoring and prize mechanisms. Beyond that, making the software as trouble-free as possible, and able to run on GPUs are the only other ideas I have.

Uncwilly · 2011-06-11, 14:47

Quote:

Originally Posted by Christenson

There's also the identity disconnect....you all think my name is "Eric Christenson", but there's only one random person on this board (Bdodson) who has actually met me, and that was years ago.

Only one that you are aware of.

davieddy · 2011-06-11, 15:19

re Bridge:
Anyone heard of Terrence Reese/Boris Shapiro and the Buenos Aries
(someone please correct my spelling*) Affair?

David

*Falklands (sheepshaggers) War was a long time ago.

PS Ever since the computer beat Gary Kasparov, and solved
"double dummy" problems, my passion for both games has dwindled.

Christenson · 2011-06-11, 15:46

Quote:

Originally Posted by Uncwilly

Only one that you are aware of.

Indeed...and not that I'm really averse to meeting others from here....I just can't see flying to Amsterdam for the purpose right now. [I'll feel differently if I actuall prove a new theorem, but that's far away].

Brian-E · 2013-01-05, 18:42

Here's an article about an open tournament last month in Croatia with many grandmasters and international masters participating, at which suspicions have been raised about possible cheating by a player who performed far better than you would have expected from his rating. Apparently he was searched for incriminating evidence, but nothing was found, and nor was his behaviour at the board unusual in any way. The suspicion of cheating, which is continuing despite no chips or audio devices being found on him, seems to be based entirely on the unusually high standard of his play.

This seems quite depressing to me. Players do, of course, sometimes play much better than they normally do, especially young players like this one. If top international chess is unable to allay suspicions of cheating in its tournaments and matches, I think it could be doomed. Or am I being too pessimistic?

The article also has many links to other articles about cheating or possible cheating in the recent past.

http://www.chessbase.com/newsdetail.asp?newsid=8751

2011-06-11, 15:19	#20
davieddy "Lucan" Dec 2006 England 194A₁₆ Posts	Hope this isn't too "off topic" re Bridge: Anyone heard of Terrence Reese/Boris Shapiro and the Buenos Aries (someone please correct my spelling) Affair? David Falklands (sheepshaggers) War was a long time ago. PS Ever since the computer beat Gary Kasparov, and solved "double dummy" problems, my passion for both games has dwindled. Last fiddled with by davieddy on 2011-06-11 at 15:32

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Vote chess game 4: To be decided? Some chess variant will be interesting to consider with!	Raman	Chess	6	2016-12-06 06:50
Reset the milestone counter: New SoB & PSP was found !!	Aillas	Prime Sierpinski Project	4	2016-11-08 04:24
Visitor counter	Rodrigo	Forum Feedback	9	2010-09-14 14:14
Counter-examples, please	Numbers	Miscellaneous Math	4	2005-12-29 11:03
A Counter example, anyone?	devarajkandadai	Math	27	2005-05-24 04:37

2011-06-10, 22:38	#14
Batalov "Serge" Mar 2008 San Diego, Calif. 3²×7×163 Posts	Google for "слюсарчук рыбка Subtitled video". There was a hilarious piece of pseudo-news, just a month and 1/2 ago. In short, an amateur played blindly against Rybka under video and won; he was allegedly searched down to underwear before the match; the Rybka was ostensibly set to some silly level and played on some slow 1-cpu computer in the auditorium where this event was set up. The same charlatan had previously claimed that he remembered 30 million Pi digits or something silly like that. He also is reported to be a 'neurosurgeon'.

2011-06-11, 02:59	#15
Christenson Dec 2010 Monticello 1795₁₀ Posts	And they completely missed the security guard completing the steganographic communications chain from the grandmaster watching the game telegraphing every move by the pattern of footfalls he set nearby on the wooden floor with his boots.

2011-06-11, 07:22	#17
ellipse Nov 2010 2×3×5 Posts	I may not know how to prevent cheating at chess, but I do know how to cheat in online chess and get away with it. Just follow these rules: 1.) If possible, use two separate computers: one for making the moves on the chess server, and one for running the chess engine. If you only have one computer, rename the program you're using ("Fritz 12.exe" may be renamed to "scanner2.exe"), and don't run it on all the cores of the computer. But remember that some chess servers don't just check for chess programs and computer resource usage, they also detect excessive switching of windows. You've been warned. 2.) Don't rely on the chess engine to make the first few moves; do them yourself, then copy that position into your chess engine. People will wonder why you take more than 3 seconds to make a simple move, and it'll look even more fishy if it happens a few times in a game. So if there's an obvious move, make the move before you put it into the chess program. 3.) In the middlegame, try to vary the length of time you take to make the moves. No human player takes exactly 30-32 seconds to move for 10+ moves in a row, and people know that. 4.) Avoid playing very short games (called "lightning chess" or "bullet chess"). It's too hard to use engine assistance there, and you'll only attract unwanted attention if you have a large rating difference between games with short time controls and games with longer time controls. 5.) Stay away from tournaments with cash prizes. Not only will cheating detection efforts be ramped up during those tournaments, but other players will also be on the alert for potential cheaters. Besides, if you somehow manage to have a bad day and get caught, you won't want to deal with the legal problems. 6.) Use different engines throughout the game. For example, you could use Stockfish in moves 11-18, Rybka in moves 19-25, and Houdini for the rest of the game. In your next game, switch the engines and the move numbers. Naum could be used for moves 10-22, Shredder could be used for moves 23-30, and you can finish the game with an older version of Stockfish. This prevents you from having a high "hit rate" when others try to compare your moves to that of common chess programs. 7.) Don't get greedy and play a whole bunch of games a day. Chess server admins and experienced players know that masters and GM's won't sit at a computer for several hours in a row just to beat some anonymous, random people on the internet. And don't play one or more games every day either; take a break now and then. 8.) If you want to cheat, do it from the start. Don't start cheating on the username that you have previously played honest games on, because a sudden rating increase will attract suspicion. 9.) Avoid taunting opponents, and keep chess-related conversation to a minimum. This should be common sense, but you'd be surprised at the number of cheaters who've said things like "You're gonna lose. Mate in 17" or have usernames like "Unbeatable Chess God", only to be reported and caught later on. 10.) Allow no more than one rematch from the same opponent. The more times you let your opponent play you, the more chances he has of finding out that you're cheating. Just say something like "I'd love to play you again, but I have to do (insert activity here)." 11.) If you do get caught, don't just make up another username and try again. Wait for at least a month, because the chess server's engine detection team will now be on high alert for other cheaters. In the meantime, try to figure out what you did that got you caught. Did you offer a draw in a clearly winning position because your engine crashed? Did you successfully analyze complex positions in a matter of seconds in a lot of your games? Or did you use the same engine too many times?

2011-06-11, 14:28	#18
Christenson Dec 2010 Monticello 5·359 Posts	I think questions of motivation are far more interesting than how to do a good job of cheating -- as I said, if you really want to cheat, hire a poor Grandmaster to be your double.. And by the way, I am among those seeking to maximise my GIMPS rating....but within the limits imposed by the seven PCs I have available. The GT440 has been most effective, when applied to TF... I think it is natural to want to increase one's reputation as much as possible. From that, flows other benefits...especially if the reputation is measured in money...like the best of everything, relatively longer life, etc. Cheating, as long as it goes uncaught, also increases one's reputation. My younger brother once spent some time in jail and noticed that he was the smartest person there. I think that the trolls/mischeif makers are all trying to increase their reputations, because of the evolutionary forces involved. I think that places like sci.math don't do anything effective about ensuring that reputation is based on any effective behavior, and the cheaters don't have the internal smarts to recognize that it's going nowhere. There's also the identity disconnect....you all think my name is "Eric Christenson", but there's only one random person on this board (Bdodson) who has actually met me, and that was years ago. Mr Blazys(sp?) can just walk away from his unbounded crank score, and it won't flow back into his other life, whatever that may be. Anyone else remember "Ludwig Plutonium" and "Time has inertia"? In addition, I have a little bit of personal experience with a crank inventor. I can tell you that he got to do some fun stuff that he never would have been able to do if he had not "invented" his perpetual motion machine, which consisted of an alternator attached to the wheels of an electric car to recharge alternating halves of the main battery. He also met people such as myself who he would never have met except for this invention; the patent office even granted him a patent, since he carefully avoided using the words "perpetual motion". What he didn't realize is that we all (or possibly just me, but I told everyone else) knew that the BEST way to extend the main battery life is to use the driving motor as the generator, just like on-the-market hybrids do, and then to start looking for the remaining losses. But that's hard work; the guy got a long way before he ran into someone(me) that said "wait a minute...this is going nowhere!". The car is actually built (it's a rusty pickup truck with a bunch of batteries across the front of the bed, gets 40 miles per charge, and has an operational controller worth $5K from my company that the inventor didn't pay for). As for getting participants in yoyo@home, etc...I think the right question is how to get participants in "useless" math programs. Somehow, I don't think the back story of GH Hardy fleeing applied math, only to find out that number theory mattered 50 years later, is going to help, even if the theme was repeated when radioactivity was studied. There's also the difficulty of the math itself. So we are left with various scoring and prize mechanisms. Beyond that, making the software as trouble-free as possible, and able to run on GPUs are the only other ideas I have.

2013-01-05, 18:42	#22
Brian-E "Brian" Jul 2007 The Netherlands 3278₁₀ Posts	Here's an article about an open tournament last month in Croatia with many grandmasters and international masters participating, at which suspicions have been raised about possible cheating by a player who performed far better than you would have expected from his rating. Apparently he was searched for incriminating evidence, but nothing was found, and nor was his behaviour at the board unusual in any way. The suspicion of cheating, which is continuing despite no chips or audio devices being found on him, seems to be based entirely on the unusually high standard of his play. This seems quite depressing to me. Players do, of course, sometimes play much better than they normally do, especially young players like this one. If top international chess is unable to allay suspicions of cheating in its tournaments and matches, I think it could be doomed. Or am I being too pessimistic? The article also has many links to other articles about cheating or possible cheating in the recent past. http://www.chessbase.com/newsdetail.asp?newsid=8751