David Bridges
In his extensive and animated interview with Alejandro Ramirez the soaring US grandmaster assured his innocence offered to prove it. How?
“This absolutely fuels me and this makes me want to win the tournament even more. I can come to the game and completely strip. You can do any fair play check to me you want, I don’t care, because I know that I’m clean. If they want me to strip fully naked, I’ll do it. I don’t care because I know that I’m clean… You want me to play in a closed box with zero electric transmission, I don’t care. Name whatever you guys want.”
Play chess games completely naked? Sounds convincing (and interesting). But then the world’s most successful billionaire took time of his Space X program and his Tesla electric car production to weigh in on the situation. He did this by retweeting a chess meme:
Before I comment on the subject I would like to say I believe that Hans Niemann is innocent. It is of course not easy to pass judgement from 7,500 kms away. Right now I can only study pictures of the venue, listen to him in video interviews, and look at the “Weighted Error Value” provided in ChessBase’s Live Game service:
The Weighted Error Value is calculated by our software when it machine-annotates games. It doesn’t consider the opening and ignores “obvious” moves (recaptures, forced replies). It compares the results with the choice of moves proposed by different engines. The lower the value it comes up with, the lower the error quotient in the game.
This is what I found for Niemann:
- Round 1: 0.09 (flawless)
- Round 2: 0.06 (flawless)
- Round 3: 0.09 (flawless)
- Round 4: 0.11 (very precise)
- Round 5: 0.06 (flawless)
- Round 6: 0.61
It certainly looked suspicious, but further checking put the results into perspective: Wesley So, who is currently leading in the event, had the following error values:
- Round 1: 0.03 (flawless)
- Round 2: 0.07 (flawless)
- Round 3: 0.23 (precise)
- Round 4: 0.01 (flawless)
- Round 5: 0.07 (flawless)
- Round 6: 0.08 (flawless)
This makes one realize that the level of play in top GM events (compare it to other lower-level games) is generally close to perfection.
And then I checked Magnus. After 0.04 (flawless) in round one, and 0.01 (!) in round two (flawless), in round three the World Champion had an 0.42 accuracy. The software ranks this as below 2700 level of play.
I also and listened to the very intense interviews that Hans Niemann has given. Based on two decades of experience in the field it has led to my purely subjective initial conclusion: he is innocent.
Regarding undetectable cheating devices: that is a subject I brought up in my proposal to FIDE over a decade ago. But I will leave a discussion of this subject to a future article.
Mason Thompson
Tessa Everly
