FM Borislav Ivanov Disqualified

That's a good point. One shouldn't seriously consider the minute probability that he is innocent. The huge probability of him cheating is the most important factor here.

Take this example:

There is a lottery ticket. It costs 1 000 000 $

It has a 99.999999% chance of winning 1 000 000 000 $
It has a 0.000001% chance of winning nothing.

Let's assume that all you have is 1 000 000 $
Will you risk all your wealth to win 1000x more money?

Nobody said not to consider it.  He is suspended so that the matter can be investigated further.  He has not been found guilty of anything, yet.

There is good chance he did cheat. From the way he has been playng etc But it seemed to be a summary 'execution'. Just a local 'court' quickly passing the judgement under pressure from the senior GMs.

I think we as Human Beings in general are trying to overcome this kind of tribal mentality, the "Lord of the Flies" and Franz Kafka kind of stuff. This happens in certain countries. Where there is no due process. Someone with a high social standing will say that so-and-so commited a crime against the state, then the entire onus is on the accused to prove himself innocent, and if he CAN'T then he is definitely guilty, for if he was innocent he SHOULD have been able to prove it.

As Human Beings he have to build a better world, and sometimes we have to sacrifice our personal sentiments or what we are losing (in this case the GMs are losing money and being dishonored by having to take someone seriously who is cheating them which a slap in everyones face true), but in spite of all this we have to uphold justice and truth.

To not to sink to his level. IF he is cheating, we will not cheat in carrying out justice. We must do it the right way, thought the right way may be time consuming, and may cost us money.

Let us not be reduced to some jungle law. Now perhaps they did consult with their lawyers and a judge as to how to proceed. IF they did and then all fair and good. If it was internally decided then how much influence did the protesting GMs have on that decision; then in that case it is a shameful thing.

It's funny to see the "this is a lynch mob" responses. You've got these people who haven't reviewed the massive amounts of evidence, and yet, are acting like the righteous choice is to continue defending someone who is clearly cheating -- until the authorities physically find his method of cheating. These people will be hilarious to listen to when Ivanov is finally caught. Here is what they will be saying, without realizing it: "I'm glad we let the cheater continue to cheat all those people for years, while we knew he was cheating, but couldn't prove it completely." You think that is true justice?

Let me tell you something. You can't always wait for 100% proof to take action. Right now, there is about 99.9% proof of cheating. If we needed 100% proof of crime in order to pursue defensive action, 99% of criminals would not get caught, prosecuted, or punished. You realize if we needed 100% proof in order to ban cheaters, then WE COULD NOT BAN PEOPLE FROM CHESS.COM? There is always a sliver of chance that all those people are just as good as computers! How do we know they aren't????? Injustice!

Do you know what "Beyond a reasonable doubt" MEANS?

http://en.wikipedia.org/wiki/Reasonable_doubt

On some days, Ivanov plays 100% like a computer, LITERALLY BETTER THAN ANY CHESS CHAMPION WHO EVER EXISTED. On other days, he plays like a coffee shop chess enthusiast. When it comes to the matter of whether or not Ivanov is guilty of cheating, there is nothing left to discuss. Ivanov's case is BEYOND REASONABLE DOUBT.

You are the one making a statistical error. You are treating Borislav Ivanov as a single data point and then arguing against the use of statistics.

Instead, each move is a data point. When you compare the moves Ivanov made to the moves made by any human, you do indeed have enough data points to draw conclusions. This isn't a misapplication to an individual, it is an application of statistics to a large population of moves made. That is fully within the realm of statistics.

Yeah, again, he hasn't even been "convicted" of anything yet.  However, there is more than enough evidence to suspend him for investigation, as has been done.

Within a single game the moves are highly correlated. This complicates analysis considerably; it's not a matter of "37-out-of-40, he's cheating."

You have to exclude moves (i) from known opening lines and (ii) that are obvious/forced/quasi-forced in the position. Evaluation of "obvious" requires assessment by a top-level player (ideally more than one). After pruning each game's moves in this manner, I expect there would be very few moves in each game that could be considered independent. Building a sampling model for this is non-trivial.

Like Al Capone. They had to put him in jail for income tax violation, although he was a known crime mob boss. I wonder why they did that way? Interesting.

woah this topic switched so hard!

See post #122. The moves within a game are not independent samples. The effective sample size, taking correlation into account, is much smaller than the number of moves in the game.

How that ?

Otherwise, each game is only about 5 exploitable moves (taking out opening and forced moves), but on the course of all the games, it remains a good sample.

Its not really about what I think the right number is, I think the method that presents the lowest chance for misjudgement should be used.

I gotta admit that I have no clue how the Bulgarian law handles stuff like this, I always assumed the TO could search players that are suspect to cheating, at least if the announce before the tournament that random searches may occur. 

If there is no way to do that then I gotta concede that banning him based on game analysis is probably the next best thing to do.

You apparently don't understand that chess.com bullet ratings are deflated and chess.com correspondence ratings inflated.  Whereas I understand the difference between USCF correspondence and OTB ratings -- for me the former is about 250 points higher than the latter.  But 600 points?  Really, you think a 1700-ish USCF OTB player can possibly achieve a 2300+ ICCF rating WITHOUT an engine?

Yes. It's not because you cannot do it that none can.

My standard FIDE rating is roughly 300 points lower than my rapid rating, which is huge - most of the people I know have a 100-point difference at most. Does it mean I am using computers in rapid games ?

Look, some more horrible analogies and straw man arguments.

WMDs and different chess rating pools have nothing to do with the fact that there is a guy who, one day plays with perfect computer accuracy and smashes grandmasters, and another day, would struggle to win a local tournament.

It's called: cheating.

Wake up.

To think otherwise is not justice -- it is simply ignorant, naive, and irresponsible.

Well said.

For example, if you entered a forcing line that, once started, both players had no choice but to complete (or be down a piece, or something similarly decisive), all of the moves of the forcing line after the initial move are correlated nearly perfectly to the first move. They offer no information; only the first move does. If the game is analyzed 'statistically' as if all of those moves were uncorrelated, the probability assessment would be wrong (not just a little wrong -- a lot wrong). That's where a human assessor is needed, to intervene and say "Just take the first move of this sequence". I think (but I'm not sure) that chess.com has someone of IM caliber do this job before someone is tossed for engine use.

If he wasn't cheating, it seems would volunteer to be searched on his own.