Borislav Ivanov ain't playing so well this time around

General discussion about computer chess...
pgn4web
Posts: 97
Joined: Mon Jun 14, 2010 10:26 pm

Re: Borislav Ivanov ain't playing so well this time around

Post by pgn4web » Thu Jul 11, 2013 1:36 pm

Jeremy Bernstein wrote:
pgn4web wrote:
Jeremy Bernstein wrote:Somewhere in the video he makes what I think is the essence of his argument: something like "if you can't prove it 100%, don't say it, because there's a chance you'll look stupid later". A defensible, logical stance, but one which hasn't served humanity well in the past.
Yeah, setting this logical stance aside, for instance, eradicated the problem of witches: back in the middle ages they burned alive so many witches than none survived to our days. Well done.
That's a great point, especially if you equate banning suspected cheaters from tournament play with murder.
I guessed, if you are a professional player or someone dedicated enough to the game to get to a 2300+ rating, being banned from playing is actually a very big deal.
That is a major concern in the whole discussion about chess and cheating: too many people throwing "easy" solutions to the problem (like banning players on hearsay and circumstantial evidence) without considering the consequences...

User923005
Posts: 616
Joined: Thu May 19, 2011 1:35 am

Re: Borislav Ivanov ain't playing so well this time around

Post by User923005 » Thu Jul 11, 2013 6:07 pm

If I were accused of stealing a balloon, costing one penny, I would want the same diligence and burden of proof to be applied as if I were accused of murder.
What I mean by that is that any attack on my integrity must be completed with utter diligence and care.
Of course, I imagine that I am an outlier here and others would feel differently.

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Borislav Ivanov ain't playing so well this time around

Post by Jeremy Bernstein » Thu Jul 11, 2013 9:25 pm

pgn4web wrote:
Jeremy Bernstein wrote:
pgn4web wrote:
Jeremy Bernstein wrote:Somewhere in the video he makes what I think is the essence of his argument: something like "if you can't prove it 100%, don't say it, because there's a chance you'll look stupid later". A defensible, logical stance, but one which hasn't served humanity well in the past.
Yeah, setting this logical stance aside, for instance, eradicated the problem of witches: back in the middle ages they burned alive so many witches than none survived to our days. Well done.
That's a great point, especially if you equate banning suspected cheaters from tournament play with murder.
I guessed, if you are a professional player or someone dedicated enough to the game to get to a 2300+ rating, being banned from playing is actually a very big deal.
That is a major concern in the whole discussion about chess and cheating: too many people throwing "easy" solutions to the problem (like banning players on hearsay and circumstantial evidence) without considering the consequences...
Take the total number of tournament/semi-professional/2300+ players in the world. Now take the subset of suspected cheaters. Now take the subset of suspected cheaters where the suspicion is solely based on circumstantial evidence. Now take the subset of those who have actually been (temporarily) banned from play on the basis of said evidence. I think we're at 1 (actually, Ivanov was banned for refusing to submit to an anti-cheating inquiry to which he'd previously agreed, not for the alleged cheating directly). The evidence is so strong in this case, and the demonstrated lack of regard for the sports federation (and its elite players and tournament organizers) of which he is nominally a part, that I think it's a thoroughly reasonable remedy, particularly since it's temporary. If tournament organizers want to continue to give him a shot at prize funds after the term of his punishment, that's their problem.

pgn4web
Posts: 97
Joined: Mon Jun 14, 2010 10:26 pm

Re: Borislav Ivanov ain't playing so well this time around

Post by pgn4web » Thu Jul 11, 2013 11:09 pm

Jeremy Bernstein wrote:Take the total number of tournament/semi-professional/2300+ players in the world. Now take the subset of suspected cheaters. Now take the subset of suspected cheaters where the suspicion is solely based on circumstantial evidence. Now take the subset of those who have actually been (temporarily) banned from play on the basis of said evidence. I think we're at 1 (actually, Ivanov was banned for refusing to submit to an anti-cheating inquiry to which he'd previously agreed, not for the alleged cheating directly). The evidence is so strong in this case, and the demonstrated lack of regard for the sports federation (and its elite players and tournament organizers) of which he is nominally a part, that I think it's a thoroughly reasonable remedy, particularly since it's temporary. If tournament organizers want to continue to give him a shot at prize funds after the term of his punishment, that's their problem.
Comparing moves with a random engine (running on a random computer with a random time per move) is nothing more than circumstantial evidence. Especially when plenty of people had a chance to scrutinize him during play, even had him strip-searched and found no evidence whatsoever of how he might cheating.
Convicting someone for failing to prove his innocence after accusations based solely on circumstantial evidence is a legal aberration, again remembers me of the good old witch hunts.
What makes things worse, the Bulgarian federation just took the easy way out in order to avoid bad publicity (as they admit in their statement) rather than going to the bottom of the issue, i.e. letting him play until you can get him in the act. Then you can disqualify him for life with no problems.
If your way of thinking was applied to cycling there would no need of antidoping test, all cyclists would be systematically disqualified...

User923005
Posts: 616
Joined: Thu May 19, 2011 1:35 am

Re: Borislav Ivanov ain't playing so well this time around

Post by User923005 » Fri Jul 12, 2013 12:05 am

On the other hand, I think that both sides have a point in this debate.
1. Circumstantial evidence is not enough to convict
2. If you have an agreement to allow further scrutiny based on the circumstantial evidence and you do not follow through with that agreement, then you have violated your own word.

So both sides have a valid point, as I see it.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: Borislav Ivanov ain't playing so well this time around

Post by BB+ » Fri Jul 12, 2013 1:32 am

1. Circumstantial evidence is not enough to convict
Convicting someone for failing to prove his innocence after accusations based solely on circumstantial evidence is a legal aberration [...]
Most legal scholars, both ancient and modern, would agree that circumstantial evidence [loosely, anything requiring inference] is typically more powerful than direct, particularly when there are independent corroborations involving the matter. http://en.wikipedia.org/wiki/Circumstantial_evidence

For instance, forensic evidence is "circumstantial" (even the proverbial smoking gun requires an inference), while "direct" evidence (eyewitness testimony that actually saw the [alleged] crime committed) is just as prone to error/prejudice, etc. The power of circumstantial evidence is particularly when there is (or appears to be) only one logical conclusion upon its admission (in totality). It can depend on the situation, but it is more typically the responsibility of the defendant (see Evidential burden) to propose plausible alternatives than the accuser to dream them up.
As human behaviour may manifest itself in infinite varieties of circumstances it is perilous to generalize,
but it is not every facile mouthing of some easy phrase of excuse that can amount to an explanation.
It is for a judge to decide whether there is evidence fit to be left to the jury which could be the basis of some suggested verdict...
To take this further: Finding a computing device on Ivanov's person running Houdini on the current board position would be... circumstantial. Finding a mechanism (for instance, teeth/gum electrical contacts a la Keith Taft) for passing information to said computing device would be... circumstantial. If you think this is pedantic, some "collusion" cases in bridge have concluded that while there was direct evidence that information was passed from one player to his partner (eyewitness testimony about foot tapping), there was no direct evidence that such information was actually used during play (which is a higher standard). [Yes, bridge probably does have too many lawyers who are avid players]. Maybe the issue really is that the legal term "circumstantial" is being used in a slightly inaccurate way here, and what should really be said is that "Regan's move matching statistics are not enough to convict", or "Lilov's opinions are not enough to convict", or "inexplicable board behaviour (is that an issue here?) is not enough to convict" -- I think such elucidation would help to clarify the situation.

So while I can't say that I've followed the ins-and-outs of Ivanov to say much about the case, I would say that terming the evidence [whatever it be] to be dismissible merely because it is "circumstantial" w/o proposing a plausible alternative is somewhat a futile scareword. For instance, User923005 came up with the idea that the Houdini/Ivanov move-matching in Zadat might be exaggerated due to opening preparation, which could be plausible (I haven't analysed it, only mildly quoted Regan disputing it). As above, I think that, particularly in a quasi-formal extra-judicial process when the accusers/judges agree that the circumstantial evidence is sufficienly weighty (prima facie as it were), it would be primarily be Ivanov's responsibility (as the accused) to propose such alternatives, which would then be considered/discussed/rebutted further (and ultimately concluded one way or the other by the trier of fact).
again remembers me of the good old witch hunts [...] back in the middle ages they burned alive so many witches than none survived to our days.
This is historically incorrect, the Middle Ages ended in the 15th century (Gutenberg or Columbus), and the witch hunts largely came afterward. Furthermore, it is not too difficult to find self-professed witches today, even in Europe. See http://en.wikipedia.org/wiki/Witch-hunt or http://en.wikipedia.org/wiki/Witch_tria ... ern_period
I might point out that witch trials could have sometimes have direct evidence (eyewitness accounts of practising witchcraft) against the accused.
If I were accused of stealing a balloon, costing one penny, I would want the same diligence and burden of proof to be applied as if I were accused of murder.
I would then expect you to defend yourself with the same diligence. As I say, I haven't followed the Ivanov proceedings closely enough to see what alternative explanations he has given (or if the Bulgarian Chess Federation has even allowed him to defend himself, for that matter).
What I mean by that is that any attack on my integrity must be completed with utter diligence and care. Of course, I imagine that I am an outlier here and others would feel differently.
My own sentiments are closer to yours. But then most modern men don't give a fig about "sacred honour" or whatever those blimey prigs centuries ago blathered on about.

Incidentally, I think that a higher diligence and burden of proof would be applied for the [petty] theft in any criminal proceedings (compared to civil ones), but that it would still be a somewhat lesser standard, as issues of property are not as considerable as issues of liberty (which are not as considerable as issues of life). Similarly, grand theft is a higher standard than petty theft, first degree murder (involving premediation, perhaps even malice prepense) requires more evidence than second degree, etc. From the court's standpoint, ordering (criminal) restitution is not as weighty as a temporary jail sentence, which is not as weighty as capital punishment and/or life w/o parole. But I point out that the measures here have varied over times and places (for instance, "banishment" seems no longer to be a punitive measure anywhere, or a millennium ago depriving someone of property could reduce them to such a state of destitution that e.g. public flogging/ridicule was preferred).
2. If you have an agreement to allow further scrutiny based on the circumstantial evidence and you do not follow through with that agreement, then you have violated your own word.
Again I'm not following this story too much, or relying on press accounts, but it is not clear to me that Ivanov was given "proper notification" of what exactly the tests might be (in said agreement), and perhaps he decided not to go through with the testing because of that. On the positive side for him, I don't think anyone could say that Ivanov is personally mounting a "defense via the press" in any sense, but I do think he should strive to resolve the issues either with the BCF or FIDE (both paragons of integrity -- NOT!) if he thinks he is being unfairly treated.

User923005
Posts: 616
Joined: Thu May 19, 2011 1:35 am

Re: Borislav Ivanov ain't playing so well this time around

Post by User923005 » Fri Jul 12, 2013 2:49 am

In this case, the evidence is pure inference.
IOW, the accused played a sequence of moves very similar to a computer (though not identical).
Therefore, the accused has cheated.

The only evidence is the inference itself.

The inference is based on probability and relies on several assumptions that may or may not be true.

A fingerprint, at least, is physical evidence.
One might claim that the game score itself is analogous to the fingerprint and that the similar nodes of play are analogous to the 16 identifying markings on a physical fingerprint identification.
I see it differently.
I might toss 20 heads in a row in a game of "heads or tails."
This is not evidence that I cheated, though it could easily arouse suspicion.
Arguments from probability assume that improbable things to not happen. Actually, they happen quite often. It's an artifact of there being so many collections of events in the world. Given enough sets, some of them will be outliers.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: Borislav Ivanov ain't playing so well this time around

Post by BB+ » Fri Jul 12, 2013 3:03 am

The only evidence is the inference itself. The inference is based on probability and relies on several assumptions that may or may not be true. A fingerprint, at least, is physical evidence.
I think Regan made a similar comment recently, that a 1-in-1000 statistical match with corroborating "physical" evidence (such as abnormal board behaviour, he suggested) might well be considered more substantial than 1-in-a-million statistics with nothing else apparent. Again, I'm not up to speed on the issues enough to say whether the verbal testimonies of other players regarding Ivanov's antics are of much value.

bluefever
Posts: 22
Joined: Wed Jul 10, 2013 8:23 am

Re: Borislav Ivanov ain't playing so well this time around

Post by bluefever » Fri Jul 12, 2013 7:35 pm

marcelk wrote:
Jeremy Bernstein wrote:Here's a rebuttal from Andrew Martin: http://www.youtube.com/watch?feature=pl ... zbPytRKY#!

Very tricky situation. I disagree with Martin's stance, but I understand his reluctance to damn Ivanov on the basis of Lilov's videos. The statistical evidence is pretty tough to refute, though. See also: Jens Kotainy, about whom a very lengthy (and very ironic) article in last month's "Schach" was written (with rebuttal by Jens, as well).

jb
Is there a real statistical analyses anywhere? I'm clicking around and I see 1. examples, 2. observations and statements, but no 3. stats other than stuff like "99% the same moves as Houdini". The last is above the level of many engines' self similarity, so something else must be meant (or it is made-up). But where is the analyses, taking all of his games, calculate similarity game by game, and do the same with a control group?
Agree. I watched one Lilov video. It started with a statement saying "100% Houdini moves are made", and as he analysed, each time a move was different, he said "well, the computer speeds are different" or "he must have been low on time here". It was terrible.

He should present the case for both sides before reaching a conclusion. As it is, it looks like completely biased opinion.

User avatar
simonhue
Posts: 14
Joined: Sun Sep 02, 2012 12:02 pm
Real Name: Simon

Re: Borislav Ivanov ain't playing so well this time around

Post by simonhue » Fri Apr 04, 2014 6:31 pm

Also pretty much obvious from his games' QoP:

http://chess-db.com/public/analyzedgames.jsp?id=2903741

Post Reply