Chess Player Strip Searched

[Chris Rice]

Reviving this old thread as the investigation into cheating allegations against the Bulgarian player Tetimov reveals links and a photo with our dear friend, Borislav Ivanov. An article published on the Russian chess news site also shows that Tetimov, who refused to be searched for electronic devices I understand, is not the only Bulgarian who is under the spotlight either which seems to focus specifically on a tournament in Benidorm last year. http://chess-news.ru/en/node/18966

We know from the other thread that FIDE have sent a letter to the Bulgarian Chess Federation asking them to cooperate with the investigation but Danailov, its President isn't playing ball. http://www.ecforum.org.uk/viewtopic.php ... 4&start=75 and he's repeatedly accused FIDE and the ACC of incompetence. http://chess-news.ru/en/node/18954

The deadline for cooperation has now passed and we wait to see what FIDE will do about it.

[Roger de Coverly]

Reviving this old thread as the investigation into cheating allegations against the Bulgarian player Tetimov reveals links and a photo with our dear friend, Borislav Ivanov.

Thanks to the Indian guy, we now have another possible method detailed. It needs a co-conspirator armed with a chess engine. So possible moves for both players are transmitted by the co-conspirator via a hidden earpiece. A motion sensitive device enables Yes/No signals to be returned. If the cheater is a reasonable player, then the obvious pauses can be avoided by running ahead of the computer. For that matter the advisor can read out the engine's line, rather than single moves. It's even easier when the tournament does the transmission job itself by non-delayed live board coverage.

[MartinCarpenter]

Anyone doing that last thing should probably think about it rather hard.

Co-conspirator easier/cheaper but I can't actually think why you couldn't (in principle anyway) fully automate some version of that Indian method easily enough. Plenty of good speech synthesisers/desktop automation software etc.

Not actually even sure you'd need the - relatively easy to find on search - microphone Just signal which file the moves are on using say tiny electrodes for each of your toes or something and some signal back again. Non trivial engineering effort of course but not insane.

The delay before each move wouldn't be good, but then you'd probably do it with someone who's a half decent player to start with so they can play truly obvious moves at speed.

As for the Bulgarian chess federation, well I won't cry if Danailov gets into trouble

[Chris Rice]

The Bulgarian Chess Federation have complied after all, sort of. Statement from the Anti-Cheating Committee including an email from the Bulgarian Chess Federation attached.

http://www.fide.com/images/stories/NEWS ... om_ACC.pdf

[Chris Rice]

ChessBase report here on the Tetimov scandal. You'll have seen most of this already so just scroll down to the end which contains an analysed game, Benidorm tournament standings and the "Quality of Play" index much criticized on this forum already....

http://en.chessbase.com/post/fide-threa ... e-bulgaria

[Roger de Coverly]

Reviving this old thread, there's a recent paper which demonstrates a more sceptical approach to the issue of finding computer usage by move matching.

https://kar.kent.ac.uk//44719/1/1-s2.0- ... 5-main.pdf

[JustinHorton]

The game between Weiss and Burille given in Appendix B.3
is a good case in point. It was played in the 6th USA Congress
in New York between Miksa (Max) Weiss (the co-winner of the
Congress, with Chigorin) and the position after 13 half-moves
had been encountered before in our historical database. From
that point on, Mr. Weiss won the game by playing 26 consecutive
moves that exactly match our engine's choices at depth
20. This is a much larger series of perfect moves than that we
saw in the Morphy examples. Surely due to computer cheating,
except for the fact that the year was 1889.

[Stewart Reuben]

Nobody in FIDE is suggesting that a high computer match in itself proves cheating. But, if there is other evidence to go with it, then it points towards guilt. DNA sampling, as a way of establishing guilt is similar.

[Roger de Coverly]

Nobody in FIDE is suggesting that a high computer match in itself proves cheating.

FIDE Arena, the online stuff isn't really FIDE, despite what they may infer, but they do claim that they can detect computer use.

Equally I thought the Regan claims from last year were that if you detected moves played above their rating station, that was evidence of cheating. I could believe that more realistic and less patronising views have prevailed.

[Stewart Reuben]

PokerStars sends 'feelers' into your computer to try to ensure you are not using a pokerbot. That is just as much against the rules in poker as in chess.
If it were a high level chess event, it could be required that there be a video link showing the player in action in the chess game. FIDEArena may have other ways they have yet to reveal.

Pokerstars wrote to a player and said that they were going to ban him. He was playing 17 tables at once and this was clearly humanly impossible. So he sent them a video of his playing online. That is a decision approximately every 3 seconds. He was not only able to do so, but also changed the music on his ipod while playing. I can't even play two hands at once satisfactorily.

[Ian Thompson]

PokerStars sends 'feelers' into your computer to try to ensure you are not using a pokerbot.

I hope my computer's anti-virus software would block it if I accessed a site that did that.

Isn't it impossible to prevent cheating at online poker? Couldn't you just use a second computer to keep track of cards that have been played in previous hands, and to calculate the chances of winning with your own hand?

[NickFaulks]

PokerStars sends 'feelers' into your computer to try to ensure you are not using a pokerbot. That is just as much against the rules in poker as in chess.

Is that something they admit, or just the belief ( true, no doubt ) of players?

[Roger de Coverly]

Is that something they admit, or just the belief ( true, no doubt ) of players?

It's been alleged, or probably known, that software that supports on-line chess can also detect task switching and possibly check what else is running.

[Stewart Reuben]

When playing poker online, you could, of course, use a second computer to keep track of everything. That is not what is against the rules. It is against the rules to use a pokerbot to play every aspect of the whole hand.
Pokerstars claim their machines can detect this. They believe it is a selling point.

You can cheat at online poker. e.g. You and a friend play in the same room on two different telephone lines from two different accounts. You go online to play in a particular game and your friend follows you to the same table. Now you can collude, by showing each other your hole cards. You have KHQH. Your friend has KD JC. He should pass.

From memory, I think I listed 9 different ways to cheat when playing live poker in my book 'Need to Know Poker', published by Harper Collins.

[Simon Ansell]

Bots in online poker are starting to become a much bigger problem than Stewart is implying. I helped analyse the data in this recent case, where suspicions were raised by the regular players and several accounts subsequently banned by PokerStars. Note this is PLO, which is far more complicated from a game theory perspective than No Limit Hold'em, or other games, so the bots are still relatively young here.

There are many other cases, just search the twoplustwo forums. And PokerStars are far and away the best site for anti-cheating and botting measures; on smaller sites with fewer resources there is likely all kinds of stuff going on.

Yes, PokerStars run software on your machine to look for certain things (which you will agree to when accepting the T&Cs when you create an account and install the software - if the software is not running you will not be able to play), but it's clear that it's not good enough by itself. It might detect casual cheats, but how can you detect someone scraping the screen with hardware via a video output cable, for example? Or running any "supplementary" software in a VM? In the case linked above, the player's statistical analysis of (strictly speaking, illegally obtained) hand histories showed playing patterns or tendencies of the suspected accounts common to them and unlike any human players - not unlike Ken Regan's methods for detection in chess - but as far as I know this is the first time a method like this has been used in poker, and PokerStars have a lot of work to do here.

Collusion, that Stewart also mentions above, is not nearly as significant and much easier to detect.

edit: the link I gave above is a poorly written piece tbh, will try and find a better article.

edit 2: this is much better piece about the same case