Thursday, July 5, 2012

History, analysis and "psychoanalysis" of cheater "Mr. F.B."

I feel sorry for my readers, I am not going to reveal true identity of this guy this time. He wasn't banned for cheating by the site officially. He is an elderly man from Germany, he lived in the USA for a few years, he graduated from mathematics at Liepzig University, he works as hosting associated professor at university in Beirut. He used two accounts on he admitted cheating with computer by mistake in a few games in his first account, he apologized then, closed this account and came back with new account having promised he would play honestly. He made it to top rated players with his new account and he became top#1 for a brief period after Ouachita and several other obvious cheaters were banned in Autumn 2009. He was accused of cheating by Costelus and after his much younger mistress (or wife) was banned for cheating, he closed his second account in early 2010 and left by himself.

He started as honest player in 2008 playing mostly free games for fun. His rating fluctuated around 2100. Then he started playing more prestigious tournaments.

I have analyzed his tournament games from this era by my usual setup and obtained these results:
(a few games didn't make the cut)
top1 ... 154/339 ... 45.43 %
top2 ... 223/339 ... 65.78 %
top3 ... 267/339 ... 78.76 %
As we can see, his top3 matchup rate is bellow average of elite unassisted CC players, thus we can tell his play was clear. His blunder graph indicates he was quite average player and he made several terrible mistakes coming perhaps from lack of practise in his recent years and fast play with only shallow analysis.

He attracted attention in the forums when he responded to a question about cheating. He wrote:
"sorry, what do you mean by cheating ? I do not suppose that the use of chess programs is forbidden!!!
We play correspondence chess, and everyone knows that almost everybody in that 'sport', 'asks' programs, at least from time to time.
So what else could cheating mean ?"
Yes, his amazement was real I suppose. He forgot to read site rules and he supposed engines are allowed as they are legal in ICCF. He quickly realized he was wrong, but we can see first signs of doubt as part of future rationalization:
"I think there is massive use of chess engines at all stages of the game on this site.
But I will follow the rule from now on
 I have to say that it is a very bad rule: you cannot prove BEYOND ANY DOUBT that someone used a program during the game.
That's why many players will never further follow the rule, and they will profit."
He was right that was a mess back then. The rest could be a matter of discussion and I would disagree with him wholeheartedly. Later he wrote:
"I admit to have used Fritz 8, since September 2008 (not before, I didn't have one !),  never without checking the variants 'he' suggested. I must also say that I have learned a lot from the 'conversation' with the program. In doing this, my rating increased from around 2100 to around 2350, in short time.
I decided to use the program because I didn't know it was forbidden. Since obviously to me, many of my opponents use programs, I didnt suspect doing something immoral, also because it is explicitely allowed in correspondence chess, and the overwhelming majority of correspondence players uses a program.
Now, unexpectedly for me, some people on this site feel hurt, which I regret, because they did have other expectations. I will therefore leave the site immediately, and not come back. I however hope you guys will discuss this and come to another conclusion sometime."
No matter he confused some dubious ICCF internal bad habits with rules of online and correspondence chess overall, I suspect here is the crucial point where the Light and Darkness started the battle of his soul. He regarded the no-computer rule as stupid, because he thought it cannot be detected and this helps cheaters. It would be relevant opinion, if cheating couldn't be detected. And he was wrong here, because he failed to realize cheating can be detected. Despite he left the site, he was invited back by Erik:
"i did not ask *** to leave the site. he is leaving on his own accord. i let him know that he could stay if he agreed to follow the rules as he clearly is a strong player."
So he came back with new account...
"I am back, with cleared rating, I will not use a computer anymore, may be I will not play at all, because so many others here use it!!!!"
And here it came. Not only he came back and resumed playing, he also made it to the top year later. Perhaps he couldn't stand derisive posts like this:
"You are just cranky because you miss your program."
"I leave now because i am a target of self-announced moral apostols."
Although he lost credibility in many eyes, I admit I felt sorry for him and I also believed he would take his promises seriously.

During the following months he didn't participate in discussion about cheating and concentrated on play instead. The problem was he failed to realize that cheating is huge offence, he underwent internal rationalization and came to conclusion cheating is something common, no-computer rule is for fools and beginners. As I will show you later, he joined The Dark Side and became dedicated cheater!

About a year later, Costelus wrote in the Cheating Forum group:
"Most likely player ABC realized quickly that at the top, all the players here use Rybka. Since nobody enjoys losing, he started checking his moves for blunders too. Another thing I noticed: ABC was very active in the forums, posting interesting chess-related things; XYZ never posted anything. Maybe he's trying to keep a low profile?"
ABC and XYZ are old and new accounts of our dude. Costelus continued:
"XYZ has nothing special. We could also call him Fruit 2.3, but at very high depths (at least 16-17). This appears to be his success story: running the engine for much longer periods of time than the "normal" players here."
Costelus checked his games in a primitive way: he let them run through an UCI with engine on the background. Our dude appeared soon - noticed by a fellow cheater Schachgeek - and responded:
"I realize that many of the people in this group are bad persons, and dont WANT to be convinced. Therfore i am leaving."
Such a complain! But I still felt it would have been huge loss if an excellent honest player had been driven out from the site by the mob so I wrote:
"Well, we should make a comparison. If XYZ has coincidence with computer higher than ABC (the same player occassionaly cheating), then it is clear he has cheated, falsely promised and lied.
But until anybody proves it, we should be polite.
Honestly i don't have much taste to do it."
I wrote the same to him in a private message. I recommended him to ask Steve Collyer to analyze his games by sound top3 method and clear the suspicion caused by Costelus.

He ignored it and left Before this however, he stated his wife was unjustly banned for cheating and he didn't trust staff and members anymore.

I owed him this analysis, so I have completed it recently. The results of his new "honest" account:
(only from the last page of tournament games, 34 games with 30 or more moves, about 4-5 didn't make the cut)
top1 ... 584/843 ... 69.28 % (first account ... 45.43 %)
top2 ... 724/843 ... 85.88 % (first account ... 65.78 %)
top3 ... 782/843 ... 92.76 % (first account ... 78.76 %) ... 7.78 sd above average% of top human CC play

This is the proof of blatant engine contamination, the probability of such result obtained by honest excellent human play is 1 of 139,000,000,000,000. Costelus was right. Our Mr. F.B. falsely promised and cheated.

What is the moral of this story? There is none. Only perhaps lack of cane during his childhood.


  1. Costelus has left the site now because chess.coms methods are rubbish. There are so many cheats still in operation. When you go to the home page and see the top 5 players, it makes me sick.
    Bill Evans? the guy has been cheating on different servers e.g. FICS for years.

  2. Yes, the situation is not good, even Julio Becerra cheated at some point. methods may be rubbish, but above all they are lazy to scan top rated CC players for cheating regularly and focus on Live chess instead. This goes hand in hand with easy free registration when banned cheaters just reappear several times the same day keeping their detection system clogged up. They refuse even to block creating new accounts from cheater's IP adresses' pool!

    Regarding the validity of their detection methods, I suspect they use double measures. I mean using two or more control datasets for catching "titled" cheaters and mere amateurs. I find such differentiation wrong in CC in principle. Even the blatant idiot cheater Falko Bindrich wasn't banned after being thoroughly reported two times. And kohai said "inconclusive". Haha, yes, z-score more than 5 sd (!!) and "inconclusive".