Gibraltar Flash: Hou Yifan resigns in 5 moves!

The probability of Hou playing 7 women during the first 9 rounds with a random pairing is so tiny that I don't believe the pairing was random.

The organisers seem not to recognise they have a problem. It is a probem when many people doubt that the pairing is made neutrally.

The algorithm that was used for pairings should be published and then we could verify whether or not:
1) The algorithm distinguishes between male and female players
2) The algorithm gives the (type of) pairings that were used in the tournament (or, how many times the algorithm should be run to get that type of pairings)

I think Hou Yifan chose a brilliant method to get attention to this problem.

Zookid: and where was this done? Which round, and what should the pairing have been?

Women, on average, are not as good as men in chess. It's not the end of the world or a problem that needs "fixing"... least of all by rigging things behind the scenes.

The pairings were from the machine, Mr. Callaghan said. That sounds reasonable. Just because the pairings are unusual, doesn't mean it can't happen. The childish whining, while perfectly understandable, should stop. Of course, if it turns out the pairings WERE manipulated, this would be a serious scandal.

The female to female pairings is not the point. Manipulating the pairings after the program generates them is the issue. It's too bad Hou Yifan didn't win against Adams in Rd 5, this was the moment to seize if possible.

@TheJackpot --
Good job

@ TheJackpot: Deleted my comment after reading yours. If your analysis is correct I guess there is nothing more to say.

PS: Team Nakamura also made an official complaint about the pairings for the last round btw, which was denied.

@fons

So... what should the pairing have been? Which round and based on what rule?

@Bright Knight

A small fish? She is number 105 in the world. I wouldn't call that a small fish.

Thanks for doing that check of the pairings, TheJackpot. Interesting. Hopefully the tournament directors will release details that confirm what you've found! I'd love to be wrong about this.

@TheJackpot: Great! thank you for your time. That should settle the case... (should... but probably won't, we have entered the post-truth era... But let's hope and see)

I see a lot of remarks saying "well, let them show the machine, so someone can check it", "It must be manipulated, this is so unlikely", etc.

The FIDE Swiss Dutch rules are on the FIDE website, in the handbook. There is pairings.fide.com which has a list of endorsed pairing software, meaning it was tested by FIDE to follow those rules. Why is nobody doing the checks?

Guess what? I did :)

Took the SwissManager tournament file from chess-results.com, created a TRF / FIDE rating report file, imported it, verified the pairings.

Round 1: differences, which is to be expected: people show up late, ratings get corrected, mistakes fixed, etc.
Round 2, 3, 4: equal to the pairing in Gibraltar
Round 5: a few differences in the group of people with 1.5 and 1 out of 4, nowhere near Hou. My educated guess: results of previous rounds were corrected after round 5 was paired
Round 6, 7, 8: equal to pairing in Gibraltar
Round 9: in the lower echelons 2 pairings were adjusted (the black players exchanged), due to (probably) israeli not playing irani
Round 10: equal to pairing in Gibraltar.

Does that count as sticking to the facts?

Sexism in chess, nothing new, she did exactly what's needed to catch the attention. Total support!

I agree totally with David Sprenkle. Someone behind-the-scenes will tell the truth soon.

It is very hard to accept that the pairings were purely the outcome of the "machine's" invisible hand with no human meddling, but an inquiry should be done to verify that. If meddling was involved, then a disservice was indeed done, and frankly done to all the participants for different reasons. Gib's boost to women's chess would be to attract many and have them play among men, as Yifan has chosen to do exclusively, not beside them.

Something is wrong, of course, that she had 7 out of 10 games paired with a female. I applaud her protest. Of course, being a woman, she will take heat from this, while a man would be "standing up for principles". I applaud her protest. The pairings should be absolutely investigated, and this was a good way to spotlight it. Just "speaking up"? Yeah right, the chess world would have said "awwww, she's so nice", patted her on her head, and forgotten about the matter. You go Hou!

As for the probability of Hou being paired against 7 women in 10 rounds in this field by pure chance, a decent approximation can be obtained by back-of-the-envelope calculation. The exact odds cannot quite be known and don't matter. Hou's opponents were among the top 150 players in this field, of whom 27 were women. On this basis, her having to play so many women without deliberate human intervention was less likely than 1 in 100,000.

Tournament organizer Brian Callaghan denied that Hou's pairings were wrong, and condemned her protest. Strictly speaking, he may have stopped short of denying that her pairings had been manipulated; yet clearly he implied it in order to absolve himself and blame her. Are we to believe there was no manipulation of Hou's pairings? I agree with commenter FabulousNight, who independently 'believe[d] him .001 per cent' -- which coincidentally also happens to be one in a hundred thousand.

But this order-of-magnitude figure underestimates the likelihood of manipulation, because so far only Hou's pairings have been considered. And prior to Hou's protest, a similar pattern had already been independently noticed, tending to pair similar players against each other, such as Adams-Short, Topalov-Cheparinov, and Yu Yang Yi-Ju Wenjun. Brit vs. Brit, Bulgarian vs. Bulgarian, Chinese vs Chinese.

Of course some such matchups will occur by chance, but in conjunction with Hou's pairings it is more likely that the organizers deliberately manipulated her pairings, and tried to hide it, than that it was all mere happenstance. That Hou's pairings were manipulated to pit her against other women seems beyond a reasonable doubt -- even though I am a bleeding-heart criminal-defense attorney.

Concluding that manipulation is virtually certain to have occurred, brings us to the question of whether it was wrong. Mr. Callaghan said the pairings were not wrong. Maybe for him they weren't but he can't speak for Hou Yifan, who like Judit Polgar regards being a Chess player as a greater thing than being a chess player of a certain sex. Both Judit and Hou avoided competitions not open to all, and at some cost. Mr. Callaghan ought to respect their policies, and if he presents Gibraltar as an OPEN tournament, and Hou enters it expecting to take on all comers, and he puts her against other women more than two-thirds of the time . . . that sounds like fraud to me. It also seems dishonest, demeaning, and doubly-degrading because it is unnecessary.

Such manipulation of the pairings is wrong for other reasons as well. Pairing countrymen against each other obviously invites collusion, and threatens the integrity of the tournament. The sort of thing Bobby Fischer correctly pointed out. Depending on how the games happen to go, the fact that the individual players are above any such stooping to cheat does not always spare them from the appearance of possible impropriety, and may subject them to malign speculation ever after.

Hou Yifan was wronged, and seriously. Brian Callaghan finds fault with the way she protested. In so doing, he added insult to injury. Hou's protest was not too much, it was too mild. Henrik Ibsen would have supported it, saying that 'One should not put on one's best pair of pants when going out to fight for freedom and truth.' [Both are implicated here]. Well done Hou!

Mr. Callaghan wronged Hou, not the other way around. He should of course admit it, apologize to his victims, and ask what he can do to repair and not repeat it.

It's all so crass. Compatriots who see each other all the time, travel to Gibraltar to play others from around the world, only to find that the organizer thinks it's really neat to make them play each other at every opportunity.

Hou's protest is likely to improve the game of chess, and help protect organizers from themselves. She should be thanked.

Why get upset about playing women? As long as they all have the same score from the Swiss system then what does it matter?
The sponsors of this tournament give a huge women's prize that's why this tournament attracts a lot of women . You'd expect her to grateful to Gibraltar for the huge extra prize women get which she has won in the past.

It's pretty obvious that pairings were manipulated for a variety of reasons in this tournament. The problem is that, by not pairing according to a standard algorithm, you're handing some players easier pairings than others. Hou Yifan herself, for example, probably actually received easier pairings than she should have because she was playing a woman pretty much every round one was available for her to be paired with. Kudos to her for protesting anyway, even though it was against her own interest in that way--and kudos to the other competitors for not making a stink about it, even though they legitimately could have. There's no reason for the Gibraltar organizers to be manipulating so much--it just leads to hard feelings. And there's certainly no reason for Mr. Callaghan to be so disingenuous in a way designed to protect himself and the directors at Hou Yifan's expense.

better way to protest ..is to crush the man-GM and continue protesting by claiming that she should play men because she is better than women in chess

@ all the conspiracists: why should the organizers rig the draw? I don't see why it should be in their interest if Hou constantly plays females...
It's just a coincidence and it's strange for me that she overreacted that way.

abdekker

You really need to take statistics more seriously !...
I support how hou protested...It is the only way to make headlines, other wise the organizers can easily get away..

@ ConwyCastle:

I can give you a statistical analysis. Among the 83 players who had a score similar to Hou Yifan (ie. pool of potential opponents), 17 (20.4%) were female. However, this doesn't mean she has a 20.4% of being paired against a woman in any given round. That's only true initially. Once she's paired against a few women, these women are removed from the pool of potential opponents, and the proportion of women among potential opponents becomes even smaller. For example, after she was paired against women 6 times, the probability for it to happen a seventh time was only 14.3%.

With all that in mind - if the pairings are fair - the probability to be paired against a woman seven or more times out of ten is: <drumroll>

(17/83 x 16/82 x 15/81 x 14/80 x 13/79 x 12/78 x 11/77)*10!/(7!*3!)

Which is 0.056% (1-in-1,779)

Now, we go to Bayesian probabilities and ask ourselves, what is more likely: that this sequence of pairings was created as a part of honest pairings (which, as we know is 1-in-1779) or that the organizers deliberately skewed the pairings? Each can answer to himself.

1. If the pairings are done using a computer program, then easy to resolve, publicize the program, version and in a short period of time, the pairing pattern could be replicated to see if that is what the program really will yield. 2. There must be some more to the whole situation. She had played a great game two rounds earlier with a deep queen sac, and lost a tough game last round. ? what else is going on ?

I'm curious about this "machine". So the machine was right or was there something wrong while doing pairings or did Hou experience a statistical anomaly. I'm confused.

Interesting case.

(i) Let the organizers make their pairing code public, and people (including me) will check.

(ii) 1.g4 is still playable. The decisive mistake (or protest) was 2.f3. It is a pitty that she gave her king some fresh air by 3.d3. 3.h3 instead would have been more in style.

@Queeg
Yes.
It would be nice to know what pairing program was used. I have read that, at one time, Swiss Perfect was known to produce unusual pairings.

There are detailed rules governing Swiss pairings in the FIDE handbook. One of these rules states (C.04.2.A4):

"The FIDE Swiss Rules pair the players in an objective and impartial way, and different arbiters or software programs following the pairing rules should arrive at identical pairings."

So it is possible to examine if the pairings were fixed or not. I suggest to await the result of such an investigation instead of indulging in speculations.

I realized this also from the very start until round 4, all of her opponents were women. No other participant in the event faced more women than she did. Hope they would be able to find a good conclusion and avoid the blaming part which is not beneficial to all involved.

You protest your way; Yifan has the right to do it her way.

pairings are deterministic, and not random. saying that there is a 0.03% chance or whatever to play 7/10 games against women is irrelevant because that is not how tournament pairings work. please learn how tournaments are organized before you abuse probability online.

@abdekker -- ....quite happy to answer you.

First, you really should not use Feynman (the license plate example) without crediting him...at least explain what he was attempting to say, if you can. The point is, which you failed to state, is that one should talk about the probability before the event occurs. After it has happened, it is 1. Einstein said something similar, but I will not post that anecdote.

Regarding your "fair coin"...you have assumed a fair coin. Nice. And you know it was fair how? If you take a Bayesian view, you don't know a priori, you perform the experiment, and look at the data, and alter your confidence in fairness depending on the data you get.

--- On a different topic, the claim is the pairings were not manually done. Well maybe that needs to be relaxed...so that if the situation of 7 woman pairings is going to happen, the pairings can be changed manually (consistent with the tournament format)

Statistically speaking...IF the field had been split even between men and women, an 11.7% chance of drawing a female opponent 7 out of 10 rounds.
With this playing field, less than a .000002% chance.

The question here..Are the pairings given to Hou Yifan correct or wrong? It will be investigated and checked by other pairing experts to find out the truth.

I totally support Hou. How else could she attract media attention to her protest if not this way?!?! Even chessbase wouldn't bother mentioning strange accidency of many female pairings if she would just win her game. She is in the position in which she can trigger media discussions by such behavior as today. It is not all about chess, especially for female players chess is means to prove that they are not second class human beings.

Here is one glaring example of bias and manipulation, IMHO, ...Rd 7 Hou Yifan defeats Borya Ider on table 20 and moves up to only table 15 for Rd 8 against Wenjun Ju. The other 8 players in Rd 7 who also had 4 pts and won their game, all playing on tables below her, moved up an average of over 10 table spots for Rd 8. Gopal actually moved up 15 spots, from table 24 to table 9, to play Caruana. Deac moved up from table 25 to table 10. Also, the opponents they defeated in Rd 7 were rated 100 pts less than Yifan's, and their rating is also 100 pts less than hers. She was also rated above them in the overall standings after Rd 7.

There are other examples as well, including Rd 5, dropping from table 4 to 19 to play Pia Cramling, where Fressinet only dropped from table 8 to 17 with the same points. Statistically impossible for any algorithm to pair W-W 70% of the time with this cross table. Human intervention with some of the pairings is obvious. The "why" is the unknown.

I commend her courage to forfeit the last game and raise awareness.

abdekker - talking about statistics. You start with the Swiss system pairing rules. You then evaluate if there is some latitude in making the choices in following these rules. You then evaluate the probability that there be intra-female pairings in that proportion in that latitude if it exists. Then you ask the organisers, if there was some latitude following the rules, on what basis they made their paring choices. Then you conclude.

Intuitively, the proportion of intra-female is so high and improbable at first sight that it looks fishy at first sight - the onus of proving otherwise is thus on the shoulders of the negationist - which we can still consider, only based on a rationale process as mentioned above.

You express yourself with a lot of authority even though I have strong doubts that you went through the process mentioned above. Which attains strongly the credibility of your thesis.

back room boys a bit to blatant with their fiddling. and with her moves she got what she wanted, a lot of attention. england developes more and more in the direction of the congo.

Organisers should share what version of which pairing program they used and Chessbase should try to reproduce the pairing. That could prove who is right. If that does not happen we should assume this information is not shared to hide the dodgy pairing made by hand...

Before we continue the debate of fair or unfair, can any experienced administrator or tournament director tell us the following: If player X faces several potential opponents with identical points and colors in previous rounds, does the pairing program randomly picks an opponent for X, or is there a specified algorithm which considers their seedings? If it is the latter, then the pairing algorithm can now be run again to check if there was any flaw.