The Dark Knight System

1... Nc6 against nearly everything

Cheating scandal in grandmaster online chess

The topic of cheating is as old as online chess itself, in fact as old as the availability of powerful engines on standard smartphones. Because the first attempts at cheating using prohibited computer assistance did not take place online, but in close chess over the board (with the classic – unfortunately no joke – of the hidden mobile phone in the men’s toilet and a conspicuous urge to urinate in critical positions, on the board of course).

The online chess portals each have their own methods to take more or less consistent action against cheaters (I have summarised this in my booklet “Online chess for amateur and hobby players”).

Now a real scandal has hit the scene. An Armenian TOP grandmaster, who with a rating beyond the 2,600 limit really shouldn’t need to cheat, was convicted of cheating in an official (and highly remunerated) online competition. This was the semi-final and final of the Pro Chess League played on, and his conviction led to the loss of his team’s actually won final, the exclusion of his team from the next Pro Chess League and his personal lifetime ban on and for the Pro Chess League. A good summary of the events can be found in the German online chess magazine “Perlen vom Bodensee” (don’t ask me why the magazine is called 😊 ).

The fact that individual players, including title holders, are convicted of cheating and banned on the respective portal is nothing new. But this usually takes place relatively discreetly in the background. The present case takes place in the spotlight of the chess public, fueled by personal and drastic tweets of involved grandmasters. And the question arises whether convicted players should/must not also be banned by the FIDE (World Chess Federation) for official FIDE tournaments (none other than top grandmaster Hikaru Nakamura has asked this question out loud in a video). Online chess is no longer a Wild West where you can do whatever you want.

Cheating is like doping, even worse than that. In most sports doping improves the chances of success considerably, but victories are not guaranteed by doping. It’s different with chess: no human player (Magnus Carlsen may forgive me) has a chance against a powerful engine. Cheating in chess is cheating on a higher level than doping. Therefore there can only be a zero-tolerance limit – if you seriously want to counteract it. Whether a ban has to be for lifetime or whether a limit of x years is sufficient as a deterrent is debatable.

Tigran Petrosian, the Armenian grandmaster concerned (not to be confused with the former world champion Tigran Petrosian!), of course affirms his innocence, albeit in a very emotional, not to say insulting way. There are two arguments of different weight against him:

  1. The analysis of his games shows not only that he played better than his level (that would be a weak argument), but that he played concretely like an engine in too many places. Human moves and engine moves differ, not in every position, but in a convincing way if there is a sufficient amount of data. Those responsible at must and will be absolutely certain, because in the worst case they will have to defend their decisions legally. So the Petrosian case shows (and this is the positive thing about the whole thing) that it is possible to determine whether everything went right even after a game has been played. This relieves the referees, for example, with regard to permanent video surveillance of players during a match (such a system is only possible in very few top online tournaments).
  2. Petrosian remarkably often lowered his gaze during the affected games, as if he had a tablet with the engine running in front of him. He will certainly not have been reading a book or the newspaper on the side in his final games (he didn’t claim to). This behaviour alone would only be suspicious, at best an indication, not proof. But in combination with the game analysis it is an indication supporting the judgement.

Of course, it is always possible that the Armenian is innocent and that he actually played like an engine in the positions analysed and was in the form of his life in all these games. Possible yes, but also very unlikely. Even an athlete convicted of doping may be right in claiming that his blood or urine specimen was mixed up in the laboratory. This may even be the case with A and B specimen, but it is extremely unlikely.

Consistent and deterrent action by the competent arbitral tribunals is imperative. Cheating is a massive threat to the game of chess, whether online or over the board. However, the methods of surveillance and proof must also be further developed and refined and ultimately made transparent, because the broad acceptance of decisions and barring is a prerequisite for the effectiveness of their deterrence.

1… Nc6 against titled players

One of the special attractions of online chess (at least on the major portals such as Lichess) is to play against a FIDE titled player in a tournament. What happens very rarely to the amateur player over the board is much more common in online chess (especially in Lichess Arena tournaments, where titled players also enter later and are thus paired against nominally weaker opponents with the same number of points).

The Dark Knight System must also hold its own against titled players as a serious and solid opening system, because if it were a pure “spoof system” it would be torn apart against these opponents.

In the last three years I have played a total of 32 games 1… Nc6 against titled players on Lichess, from Bullet to Blitz and Rapid to Classical: 8 * NM, 7 * CM, 11 * FM, 5 * IM, 1 * GM (Viacheslav Tilicheev).

The result:

  • 6 victories
  • 4 draws
  • 22 defeats

Makes a point yield of 25%. For an amateur player against titled players (all games logically with Black) a good result (as an amateur I consider every half point against such opponents a success).

More interesting for the practical evaluation of the TDKS is not only the result at the end of the game (especially in Blitz games, a tragic time trouble can overturn the whole play), but also the analysis of the game development, especially in the moment of the transition from the opening to the middlegame. That’s why I have summarised all games in a Lichess study and subjected them to a quick analysis with Lichess’ own server engine (Stockfish 11+).

Below each game you can follow the progress of the score graphically; this is practical and helpful at first sight. Here the graph of my rapid chess game against GM Tilicheev:

As you can see, Tilicheev came out of the opening with an advantage (+0.9), built on it in the middle game, made a blunder (is that even allowed to say at a GM?) in the 19th move and took advantage of a blunder of mine in the 33rd move to win the game.

The graph of my draw against IM Sergey Klimenko is a visualisation of the wishful thinking of a TDKS game:

Admittedly, Klimenko made a typical TDKS careless mistake on the 4th train, but it’s good to know that this can even happen to titled players in the TDKS.

Relaunch Website

The relaunch of the website was successfully completed.

The website now runs under WordPress and is bilingual (German/English). The English translations are handmade and probably as bumpy as most chess videos by non-English people 🙂 – thus in the best of company.