If you consider that all top GMs nowadays play variations checked with top computers, you could also consider a game played by 2700+ players as an output of engine analysis.

So it's quite difficult to draw the line...

costelus, please call it something other than "a form of offline cheating" -- because it's only cheating if it's against the rules.

Maybe instead I would propose "a form of offline computer engine assistance".

PerfectGent: Yes absolutely, a little odd. I guess there are just so many positions though, and you're somewhat unlikely to encounter that same Bishop or Knight question in a subsequent game. And if you did, it would only be for ONE game.

I wonder if this problem would be a LOT more pronounced in a thematic tournament...

Yes, it is odd. Basically, in critical positions, the possibility of using cyborg databases is very important. That's why I say that allowing huge databases from ICCF games is a form of offline cheating. It will allow you at least an edge over your opponent. In a sharp opening this is all you need...

i have pondered this myself. eg if i play a game today and i have a critical move where i can capture with B or N. I choose N and lose the game. after the game i analyse with rybka and find that i have an easy win if i capture with B instead. Now this game with its analysis goes into my games dbase. Later i reach the same position in another game and there waiting for me is my previous game analysis. I obviously capture with the B and win.

To me this means that i cannot use an engine on the original game but i can use its output on subsequent games!!!!!! So in essence i am only forbidden to use the engine the first time i meet a position. Seems a little odd to me.

Does chess.com have an official position on the usage of cyborg databases? It would seem to me that they would allow them. But I am wondering.

PerfectGent:

1. 2200 is not something fixed. Maybe Erik chose this number to give the impression that there are not so many cheaters. They also banned cheaters with lower ratings (who did not have the chance to increase their rating too much).

2. I also said this many times that cheating or not cheating can only be saw by analyzing games. A high rating can be only an indication, but ... common, this is not witch-hunting :)

3. This is also an issue. A good database with many cyborg games will offer you an edge over an opponent who does not have such a database. Especially in a very sharp opening like KG for instance. Basically, the usage of such a database is a form of "offline" cheating.

4. yes, you are right. Because here you have the possibility of chosing your opponents, you can increase your rating very much.

In short, I do not blindly believe that ALL the high-rated players cheat. But I also know that I chose Chessnut AT RANDOM from the top 10. I also know that on ICC the players are much weaker than here, even if many of the top players there have an ELO.

It's an interesting point you make.

I don't think correspondence ratings can be accurately compared to OTB ratings anyway, as you suggested already there are vast differences between the two. Some players who are brilliant Chess players OTB might be poor on Chess.com because they don't see the games as very important and as a result do not scrutinise over them to the extent of a lesser OTB player who has a higher rating on Chess.com.

It is true to say though that the people who do cheat are very likely to be over 2200 as if they are not then the chances are that a program isn't involved. The very idea of cheating is to get a higher rating (I think), at least that appears to be the only motive since they can't be enjoying not actually playing Chess so they must be enjoying the ego boost instead. This does not mean that everyone over 2200 is a cheat though, and anyone who actually thinks that is just ignorant.

A

you mean your real name is not ozzie c cobblepot???????

Yes - I think costelus's point is that if a titled player wants to play anonymously they can still do it even if they have a nice shiny NM or FM next to their name.

and i wasnt referring to knowing titled players names. I meant titled players who did not declare such information. so looked like players with no official rating

Costelus wrote "ah, for issue 1 in your post: that's perfectly fine. Tell me, do you know the real name of Ozzie? I don't and I bet that the majority of people here don't know it either. But that is not the situation with the vast majority of top players on chess.com, who are *NOT* titled."

You are correct. There are 5 people on chess.com who know my real name.

Ozzie

i agree that there are cheats at the top. I also appreciate that you are of a mind to realise this and are not implying all high raters with no official rating are cheating. Though I think you would agree that some in the official thread are conducting a witch hunt against all high rated players. My blog was not entirely about me but was for all those who have come under suspicion for simply being above 2200.

Batgirl

Im glad that you agree with me, however the majority (on this and other sites) that i have spoken with are of the opinion that i am unusual. Perhaps they just dont like to have their predjudices challenged by accepting the reality.

In judo i got my 1st dan in 1968 when gradings at that level meant a 600 mile round trip to London and a hope that enough players turned up to have a valid exam.

Costelus

1: You may not consider me a high rater but in erik's original post he made the distinction that those under 2200 were highly unlikely to cheat which by inference suggested that those over 2200 were possible suspects. Unfortunately there are those who just take this figure as gospel and assume that if such a rating is not accompanied by an official rating then that player MUST be cheating.

2: my rating has been in the 2400s at some time - does that mean that for that period you would consider me a cheat??

3: i agree that this is an indicator requiring investigation. However i would put a lot of the better performance down to the use of computers to search much larger databases.

4: dont know about icc but my previous site chessworld.net had players much higher (over 3000 in a few cases) and with better stats (one player was unbeaten and with only 1 draw). It depends upopn your opposition. If you only ever play lower rated opponents then you will climb the ratings slowly but look very good in your games as you are not being challenged by good players.

Perfect Gent, may I say some words on your blog?

1. First of all, I do not think you are high-rated. An amateur player, given the time an amateur can invest in chess, can reach to 2200 ELO. Therefore, an online rating of 2300-2400 is not something unusual. And definitely, not something to be labelled by default as cheating.

2. What is extremely suspect (not in your case) is the professional level (2300+) displayed by amateur players. Common, you invested so much time to get to at least the level of a FIDE master, and you don't even have a Fide rating?

3. issue no. 2 in your post: Please explain me why, even the best correspondence players in the past (some of them were strong OTB players also), did not achieve the level of play of the amateurs on chess.com? They made blunders, they do not see tactical shots, while the top players here NEVER make such mistakes. How come?

4. How come the players here are much much better than the players of correspondence chess (no computers allowed) from ICC?

ah, for issue 1 in your post: that's perfectly fine. Tell me, do you know the real name of Ozzie? I don't and I bet that the majority of people here don't know it either. But that is not the situation with the vast majority of top players on chess.com, who are *NOT* titled.

"When i mention that i dont have an official rating because i never had time to join a club or travel to tournaments i am told i am unusual."

 

Really?  The VAST majority of chess players whom I know have never set foot in a chess club, nor played in a tournament, myself included.

My suspicion is that some higher rated players cheat, some don't. Some lower rated players cheat, some don't. I can't tell who might cheat from their rating, but, then again, a higher rating is at least a goal that might inspire one to cheat - so, I would look at the games of a higher rated player a little more carefully.

What dan are you in judo?

Great post man, I totally concur to your opinion. I don't have a rating either, I just want to play chess free of any pressure and insinuations. And I would definitely choose not to have a rating if such an option exists on chess.com. Cheers!

A good and well-reasoned post! I have played you, and of course I have the utmost respect for you, for your playing style, and for your integrity. As I wrote in the "official" forum topic, I think that the lack of titled players at the very high levels is suspicious. But the very high levels are 2600+, 2700+, and 2800+, not the 2200+ group. I find all of your explanations plausible. Of course the implied #0 is computer assistance. One could, in theory, go through all top 200 players and find out an explanation for them, and then make a histogram. I guess what I'm saying is that I would be surprised if all of the top players fell into #1-#3.