Cheating suspicion at the Zadar Open in Croatia

[Don]

Only after a criminal is caught based on some other evidence. It is not used by searching DNA on a database. The difference is huge.

I'm afraid there are some countries that do match DNA traces with a database in order to find a suspect. You are correct about the prosecutor's fallacy, but prosecutors (and judges) still fall for it...

I don't see a problem with searching a database for DNA matches. There are issues of privacy of course that might come up, but aside from that I think it's perfectly valid.

Sometimes the police match partial license plate numbers, so if DNA is invalid then so is this.

The idea of conditional probabilities is to get independent observations, so finding a DNA match in a searched database and then ignoring it is as stupid as you can get. I don't think DNA evidence is ever used to convict a person without other evidence in combination. If your DNA is all over the crime scene but you were in a hospital bed in a coma on the other side of the world, then you are not guilty. Unless there is other evidence tying you to the crime then you probably cannot be convicted. (I am not a legal expert but that is what I have heard.)

In the 10% blood type example, if you KNOW the perpetrator has a certain blood type and you can eliminate 90% of the population for sure, how is that wrong? What is wrong is misinterpreting the data. In this example the blood type is not 90% proof that you are guilty but it can be 100% proof that you are not.

I don't want to get into privacy issues, that is a whole different discussion and not relevant to this discussion.

[Laskos]

The other problem is that "condemning" (i.e. banning etc.) based on statistics set a very dangerous precedent. Not that I have a solution, but I guess my point is that this is complex.

Miguel

DNA profiling is a statistical tool too. Each STR is polymorphic, but the number of alleles is very small. Each STR allele is shared by around 5 - 20% of individuals. The power of STR analysis comes from looking at multiple STR loci simultaneously. The pattern of alleles can identify an individual very accurately. Similar to what these Buffalo guys are doing about Ivanov.

DNA profiling is used in criminal investigations, often as proofs.

Only after a criminal is caught based on some other evidence. It is not used by searching DNA on a database. The difference is huge.

I think it is used by searching DNA database, for example in typical "Familial DNA searching", used in several developed countries. Investigators use software to compare the forensic profile to all profiles taken from DNA databases in order to generate a list of most likely to be a close relative of the forensic profile. Once a suspect has been identified, his DNA profile is then compared to the sample found at the crime scene, to definitively identify the suspect as the source of the crime scene DNA. This, corroborated with other evidence, can be used as a proof.

Usually the judge must also ensure that the jury does not confuse the 'match probability' (reliability of DNA identification) with the 'likelihood ratio' (the probability that a person with matching DNA committed the crime), but ensuring that requires that the jury has some notions of Bayesian statistics.

Why do we have a suspicion about this guy? It is only based on stats or the surprising results he got. In other words, any person in the planet in the last x years who would have had this type of rare results would have set the alarm (it is almost the same as searching from a database). Then the probabilities need to be calculated in a different way, and certain rare episodes do not seem "that" rare anymore. This is what the "prosecutor's fallacy" is a about, and by the way, some people in the past were condemned based on apparently solid statistics, who were later found innocent.

That's indeed the case here, where given a large enough database, damning matches can be found. So yes, another kind of evidence is still needed. But Ivanov by now looks suspicious to me, while before this study was found by Don, I was amiss what this fuss is all about.

Yes, sounds convincing to me, but that is not the point. There should be different levels to convince me and to execute a ban.

Miguel

[Hood]

Hi,

is it not like a paranoia?

Is he playing like Houdini or Houdini is playing like him ?
Is he cheating or Houdini is cheating?

Rgds Hood

[Don]

Hi,

is it not like a paranoia?

A better word for this is McCarthyism. My personal belief is that this is, and will continue to escalate to the point in the general chess community that we will see a form of reverse McCarthyism, where some will viciously go after anyone who believes they have a legitimate accusation. It is hard to understand how you can appear to have so much compassion for one person you will smear as many people as possible in their defense without any regard whatsoever for their feelings or viewpoint.

I call it reverse McCarthyism but it is an irrational fear of McCarthyism to the extent that it cripples people and makes them stupid, just like McCarthysim itself does.

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

In the Lance Armstrong situation there was a backlash against Lances accusers, according to them they were "going after" Lance out of petty jealousy and they were the scum of society to victimize poor Lance. To them this was also clear proof that the organizations involved were highly corrupt. Many still defend Lance even after he admitted publicly what he had done with an apology.

Is he playing like Houdini or Houdini is playing like him ?
Is he cheating or Houdini is cheating?

That is a strange viewpoint. Are you saying that maybe Houdini is taking moves somehow from Ivanov? Now THAT is what I call paranoia.

Rgds Hood

[Adam Hair]

Hi Adam. I went through thr game with Kurajica some time ago.
Here is a summary of my check. Is it similar to yours?

Hardware, OS and settings: i7 / Win7
GUI Fritz 11/Hash 256 Mb/ Houdini 2.0.c/1 thread/ depth 18
Legenda:
first row:: moves from.. to..
second row: 1 means a matching between move played and Houd's first choice
third row: 1 means a matching between move played and Houd's second choice
fourth row: 1 means no matching, or matching between move played and third or worse
anywhere : ----- means never happened Houd's move

Ivanov
10-19
I 111111111
II 1
- ----
20-29
I 111111111
II 1
- ----
30-34
I 11111
II ----
- ----

Kurajica
10-19
I 111
II 111
- 1111
20-29
I 111111
II 111
- 1
30-34
I 111
II -----
- 11

TOTAL
Ivanov I 23 II 2 - 0
Kurajica I 12 II 6 - 7

Hi Enrico,

I also allowed Houdini to search to depth 18, but I started at move 15 and only used single pv. For me, every move made by Ivanov agreed with Houdini 2.0c. 21 matches for 21 moves. 13 out of 20 of Kurajica's moves agreed with Houdini. However, 3 of the non-matching moves occurred near the end of the game after Houdini had found checkmate. Before that point, Kurajica and Houdini agreed on 11 of 15 moves.

The agreement with Houdini in this game does look quite damning for Ivanov. However, when I also look at Kurajica's moves, the evidence does not look quite as damning.

[noctiferus]

Thx, Adam.
First, I'll try to clear off my summary according to your choices.

Second, I hope to find some time to run the same analysis on a few games of the last tournament, whose games can be found here:

http://www.messaggeroscacchi.it/?p=4116#respond

already mentioned in
http://talkchess.com/forum/viewtopic.php?t=47933

Don't hold your breath, in the meantime...

[Rebel]

I am undecided on the isuue.

I will not easily forget the case of Lucia de Berk a woman life sentenced for murder in 2003 based on statistics and exonerated in 2010.

From the wiki:

law psychologist Henk Elffers, who was used by the courts as expert witness on statistics both in the original case and on appeal, was also interviewed on the programme and stated that the chance of a nurse working at the three hospitals being present at the scene of so many unexplained deaths and resuscitations is one in 342 million.

After this news 99.99% of the Dutch believed the woman was guilty as charged including the judges. And in the end they were wrong due to a few persistent stubborn people.

Exceptions do happen, in fact Murphy's law guarantees it. If things are probable they will happen sooner or later. Whole evolution theory is based on that.

Living in a world of multiple proven cheating cases every exceptional tournament performance immediately is highlighted as suspect.

[noctiferus]

Hi, Adam.
Here is my summary for Ivanov Kurajica, adjusted for your analysis:
Ivanov I 19 II 1 ---- 0
Kurajica I 12 II 4 ---- 4

[Don]

I am undecided on the isuue.

I will not easily forget the case of Lucia de Berk a woman life sentenced for murder in 2003 based on statistics and exonerated in 2010.

From the wiki:

law psychologist Henk Elffers, who was used by the courts as expert witness on statistics both in the original case and on appeal, was also interviewed on the programme and stated that the chance of a nurse working at the three hospitals being present at the scene of so many unexplained deaths and resuscitations is one in 342 million.

After this news 99.99% of the Dutch believed the woman was guilty as charged including the judges. And in the end they were wrong due to a few persistent stubborn people.

Exceptions do happen, in fact Murphy's law guarantees it. If things are probable they will happen sooner or later. Whole evolution theory is based on that.

Living in a world of multiple proven cheating cases every exceptional tournament performance immediately is highlighted as suspect.

The problem with expert testimony is that the expert is payed for his expertise. Even if he were not payed there is still bias because he would not even have been called if it were not known in advance what he was going to say.

There are some travesties of justice that probably will never be avoided unless you simply don't arrest or jail anyone because even it the odds are properly calculated and it's 1 in 200,000,000 - that's better than winning the lottery. Rare events do happen. But if multiple people are dying (and will continue to die) and you fail to act, then you have a much more serious travesty of justice and it's universally understood that you have to have balance if you want to have true justice. Most countries consider it far more grievous to jail an innocent person than to let a criminal go and that is how it should be since the whole point is to protect the innocent. It's not possible to never make a mistake unfortunately, unless you throw justice out the window.

This is why earlier I said that if I knew a suspected pedophile was living in my neighborhood, I would not let her babysit my children. I would keep an open mind about her possible innocence, but I would not take a chance with the life of my children. That is why I think innocent until proven guilty is sound bite that doesn't reflect either reality or common sense. If we really believed in that we would never put anyone in jail or even arrest people. As you know people are arrested before they are proved guilty or innocent. How stupid is that to arrest innocent people?

The principle should be (and actually is) careful escalation. If there is a suspicion there also needs to be level of evidence for an arrest, then there has to be a hearing to determine whether to proceed farther and then the possibility in most countries of a re-trial by a higher court. You want to abort the process as early as possible when innocence is established but you also have an obligation to protect society.

[noctiferus]

I was doing my home work, as promised to Adam, when I stepped into a unexplicable move (at least, for a patzer like me).
It was a game in Tarnovo "Old Capital" tournament, 2013, played between Ivanov and IM A.Rombaldoni.

[d]r6k/3PN1pp/8/p7/P1P2PbP/1P4P1/8/4R1K1 w - - 0 42"


While before this move, Ivanov's performance was, as almost always, at the best of Houdini 2.0c (details will follow), here he plays Nc6.
While, before this move, he was playing first or second Houdini's moves, this one is below the 25th at depth 18, up to at least depth 20 (stopped here).

Actually, Houdini's suggestions were (depth 18)
Houdini 2.0c x64: 1)} 42. f5 Rf8 43. Ng6+ hxg6 44. Re8 Kg8 45. d8=Q Rxe8 46.
Qxe8+ Kh7 47. Qxg6+ Kh8 48. Qxg4 Kg8 49. f6 Kf7 50. fxg7 Ke7 51. g8=Q Kd6 52.
Q8c8 Ke5 53. Qce6# {[%eval 32755,19]})
({Houdini 2.0c x64: 2)} 42. Nf5 Bh5 43.g4 Bg6 44. Ne7 Be8 45. dxe8=R+ Rxe8 46. Ng6+ Kg8 47. Rxe8+ Kf7 48. Rf8+ Ke6 49.
Ne5 h6 50. Kf2 Kd6 51. Rg8 g5 52. hxg5 hxg5 53. Rxg5 Kc5 54. Nf3+ Kb6 55. Rb5+
Kc7 56. Rxa5 Kd8 57. Rd5+ Kc7 {[%eval 1666,18]})
({Houdini 2.0c x64: 3)} 42.Ng6+ hxg6 43. Re8+ Kh7 44. Rxa8 Bxd7 45. Rxa5 Bg4 46. Kf2 g5 47. hxg5 Kg6 48.
c5 Kf5 49. b4 g6 50. Ke3 Bh5 51. c6+ Ke6 52. b5 Kd6 53. Ra7 Bg4 54. c7 Kd7 55.
b6 Kc6 56. a5 Bc8 57. Kd4 Bf5 {[%eval 1123,18]})

Under hypothesis 1 (Ivanov is cheating) what happened? transmisstion problems? Wrong understanding of audio transmission?
Under hypothesis 2 (Ivanov not cheating) how it is possible? Not even a patzer like me would give such a pawn for nothing...

I would really appreciate your opinion. Thx.