Ladies Free Skate and Results + GPF Finalists

[moyesii]

Joesitz,
Bloc judging actually becomes more potent under the CoP, just less transparent. Under CoP, just one judge can have a huge impact on the scores. If there is indeed a certain group of conspiring judges (or even just one corrupt judge) favoring certain skaters, the results will be easily manipulated without anybody knowing.

Keep in mind that statistically, the sum of scores has no real value, because it doesn't measure a truly representative value of a set of scores. Statistically, the sum is only a preliminary step in calculating a meaningful value in a data set. The sum itself is not meaningful because of the fact that just a single score can skew the entire data set. The ordinal system is more robust than the CoP, and more of a deterrant to cheating. Under CoP, the random count of judges' marks may or may not discourage cheating, but it definitely won't guarantee that it'll stop. And once any corrupted marks get into the CoP system, they will have a huge and unfair impact on the results. All it takes is one judge.

[chuckm]

I don't think one judge can have that much of an effect, moyesii. First of all, the judge has to be one of the ones selected. Then, the high and low scores get thrown out. If that judge was a 'skewer', his result gets eliminated.

What IS a danger is several judges (4 or more) getting together to pre-decide the results of a competition. The likelihood of some of them being selected is high, and even if a score gets thrown out, it is probable that one or more will be included.

Anyone who thinks this can't happen has his or her head in the sand.

An interesting thing is that although the same judges sit on every panel for the same discipline, their relative position changes. I think the 'seats' are selected electronically for inclusion, so whose scoring counts depends on where they sit for each segment.

[moyesii]

First of all, the judge has to be one of the ones selected.

Random count of judges can't be relied on to salvage the many faults in the code of points. If you say that there's a 50% chance that corrupted marks won't enter the system, then you can't ignore that there's a 50% chance that they will. Leaving things up to fate doesn't solve anything, doesn't make the competition any fairer.

The worst part of random selection is that in close competitions (for ex. see Trophee Lalique men's 3rd and 4th places) the podium results will be determined completely by chance. Due to the inherent variability and systematic error in the judges' marks (including subjectivity, bias, and human error), in a close competition just one judge can have a huge impact on the final result, even if the judge wasn't intentionally trying to corrupt the marks. The only way to avoid such a problem is by using ordinals, to allow each judge to place the skater according to their own calculations, and then by comparing each judges' marks within the panel. This produces a much more statistically meaningful result, and the final results will be much more representative of the marks of the judges. By using a randomized sum of the pooled scores, the results may or may not reflect the majority opinion of the judges; most likely they will be skewed so that the total scores are misleading; and the results will ALWAYS include random chance and error in their calcualtions.

The sum of the judges' scores will always lead to faulty and misleading results (sometimes significant errors, othertimes not), and there's no way to fix that so long as the sum is used to calculate the final results.

The CoP system simply makes cheating less transparent. At least in the ordinal system, each judge was accountable for their marks. Increasing the # of judges in the panel AND using the median to calculate the final results would make the ordinal system vastly superior to the CoP system.

[thvudragon]

Originally posted by moyesii
Yes, thvudragon, there's absolutely no such thing as bloc judging. It doesn't exist. We just say it does to explain all the 5-4 splits in the judging panels in the past.

Would you like to provide some examples of how there is an eastern bloc, but no western bloc? Such assumptions we are making today!

[moyesii]

Yeah. Get over it.

[Joesitz]

The bloc judging is cultural - not political - not conspiritorily. there are more slavic countries judging than there are celtic judges

Joe

[moyesii]

The bloc judging is cultural - not political - not conspiritorily.

Joesitz, Don't you think that the sociological foundation (theoretical) of bloc judging is irrelevant to its impact on skating results (empirical)? If judges are biased, showing favoritism -- these issues have to be addressed.

Of course judging is subjective, but there are rules and standards that they follow so that judging is not an arbitrary and arcane process. Rules and protocol govern both the ordinal and CoP systems. The CoP is more transparent, giving viewers a certain satisfaction of understanding the process, but the CoP is also more faulty and deceptive, with many weak links in its chain. A court judge is a perfect example of a less transparent but fair and legitimate system.

[chuckm]

It's impossible to remove favoritism or unreasonably biased judging from the sport as long as a cloak of secrecy protects judges from scrutiny.

Under CoP, we can compare one judge's marks to the marks of other judges, and can plainly see when a different 'standard' is being applied. If the scores are way out of line with the other judges, then if that judge is selected, at least severely aberrant results will be thrown out. What is harder to detect is groups of judges acting in tandem, marking the same skaters up and/or down. Even if such patterns are detected, we don't know who the judges are because their identities are protected.

We can't rely on the ISU to police its own judges, because as we have seen in the past, nothing has ever been done until an enormous scandal erupted into public view, forcing the ISU to act. In those cases, the identities of the offending judges were known. Not any more.

[Joesitz]

moysesii - Much of what you are saying is blowing my mind. I need time to devour it. However, I do like it and this dialogue.

But my argument about cultural bias does exis, imo, and the slavic world has the edge on this form of non-intential bias. That was the reason for so many complaints about the old system. They were demanding that Eastern Europe should be limited to 2 judges in each competition. Western Europe would have 2 judges (much more diverse people). Asia would have 2 and N.America would have 1. However, this didn't happen and at 2002 Worlds, Irina received 7 out of 9 first places all 7 of which were from Slavic countries. There was no conspiracy here. It was a matter of communal taste in art (and sport). Even Dick when asked the question: Can MK win tonight? Reply was not with this line of judges.

I raise this example, not for discussion of fans' personal preference of Irina and Michelle but simple to show the theory of non intentional cultural bias. And if you go through the scores of other competitions which another poster had done sometime ago, where there are 2 or more slavic judges you will see that those judges agree with themselves. Again, I reiterate, this is not a conspiracy. It is simply the way people are brought up.

In the US and Canada, the populations are much more diverse and many of those people (and judges, for that matter) are close to their ethnic backgrounds and have specific different tastes in culture. Americans are not always in agreement with each other.

Now, conspiracies are a different kettle of fish and I do believe they exist and can cause more dislike for the sport especially in view of the SLC scandal(s). I'm not sure about the Dance Competition. I hae to study your theory on this. BTW, thank you for interest in this subject.

Joe

[euterpe]

It is true that 2002 Worlds had a panel for the Ladies SP and FS that was 100% European, and 5 of the 9 judges were Eastern European.

It is also true that the Russian and Belorussian judges gave Irina Slutskaya 6.0 for presentation in the SP.

But, having said that, Irina won that competition fair and square, and gave possibly the best pair of SP and FS performances she had done since the 2000 GPF. Kwan had stumbled on her 2A in the SP and finished 3rd, behind Irina and Fumie Suguri, so she wasn't in control of her destiny in the FS anyway. Even if she had won the FS (and it was close between Irina and Michelle, as they both skated very well), she couldn't have won gold.

The most egregious example of cultural bloc judging I have ever seen was the 2001 GPF final FS, where Irina stumbled throughout, completing only 4 clean triples, while Kwan completed 6 and Sarah Hughes 7. Because Eastern European judges dominated the 7-judge panel, Irina was placed 1st in a 4-3 decision.

[moyesii]

Originally posted by Joesitz
But my argument about cultural bias does exis, imo, and the slavic world has the edge on this form of non-intential bias. That was the reason for so many complaints about the old system.

Joesitz,
I understand what you're saying about cultural bias. If you are right, then "bloc judging" would be a misnomer, because the judges would be unwittingly forming groups among themselves. But there has been proof that judges have been working together. Remember the toe-tapping incident, and SLC of course. No one can suggest that these are isolated incidents. Also, if results were based on inbred cultural differences and not forged alliances, wouldn't make the effects of cultural bias any less harmful than bloc judging. Also, are you suggesting that this problem is peculiar only to the 6.0 system, and is not a factor in the CoP?

There was no conspiracy here. It was a matter of communal taste in art (and sport).

I'm curious how you'd explain what criteria specifically are involved in cultural bias (non-skating-related judgments) that discriminate two skaters such as Michelle Kwan and Irina Slutskaya in the eyes of the judges. For example, skating-related factors that sometimes put Slutskaya rightly ahead of Kwan were speed, spins, and jumps. But what cultural aspects of these skaters' skating would you say made one skater more favorable to Eastern European judges? I felt that Kwan's nuanced skating and choreography to Scheherazade were much more in the classical European style than Slutskaya's Tosca. Aside from the fact that Slutskaya is from Eastern Europe, how would the judges "unwittingly" identify more with Slutskaya's skating than with Kwan's skating?

[Mathman]

I have wondered about that, too, Moyesii. We think of the European and especially the Russian style as favoring grace over klutziness (Oksana v. Nancy). Yet when it was Kwan against Slutskaya and Butyrskaya, suddenly they did a hundred and eighty.

I think the so-called Eastern Bloc has been more consistent in pairs. Sale and Pelletier, skating like a man and a woman, were not as well received by Eastern European judges as were the more ornate Berezhnaya and Sikharudlize, skating like two artists.

Mathman

[Mathman]

Pardon the double post, but I wanted to remark on some earlier posts on this thread.

The more I read arguments on the statistical merits and demerits of the CoP versus ordinal systems, both from casual fans and from serious observers with some exposure to statistical modeling, the more I have come to appreciate Cinquanta's cleverness. Like Bill Murray in Groundhog's Day, we can hardly click on a figure skating URL without reading the same rehash about means and medians, standard deviations and ordinals, analysis parametric and non. (BTW, whatever we might feel about this argument, there is no reason for us to use discourteous language to each other.)

Speedy has thrown out so many red herrings, statistical and otherwise, that we have forgotten that this was supposed to have something to do with corruption, scandal and pre-judging via deal-making in smoke-filled back rooms.

But I love the CoP. All those delicious numbers for us to look at! Speedy has co-opted me, too.

Mathman

[giseledepkat]

I really enjoyed your last post, Mathman! It was so finely nuanced, even for you!

I'm reminded of the old children's bedtime story, The Emperor's New Clothes. I think that a lot of the people so upset by CoP must feel like the little boy in that story, watching the emperor parade by and shouting, But can't you see -- he's naked!

Speaking for myself, I'm a fan of the new system, and I find so much to like in it -- but every once in a while I wonder if I'm just enraptured with my own image of the finery which has been described for us.

[moyesii]

Speedy has thrown out so many red herrings, statistical and otherwise, that we have forgotten that this was supposed to have something to do with corruption, scandal and pre-judging via deal-making in smoke-filled back rooms.

Mathman, I haven't forgotten. Just only so much at a time that we can deal with, and very little it feels like we can do (as spectators).

By the way, sorry if my posts sound like rehash. If you find yourself enamored with CoP and the ISU, perhaps it'd be better just to ignore my threads about the CoP and its demerits. As for me, I find the statistical faults in the CoP to be glaring, and very dangerous, particularly because they mislead the general public to thinking that results are being computed fairly and systematically without error. I'm reminded of something as well, giseledepkat. In my stats course in college, one of the things I learned very early is that statistics explains many phenomena counterintuitively. Therefore, I think the emphasis that people like Sandra Loosemore and George Rossano have placed on statistics in their methodology. Unfortunately, the majority of people will not be savvy to stats, and this makes it easy for us to be hoodwinked by the ISU into thinking that we have a legitimate, working system. As I said, one of the particularly preposterous aspects of the CoP is that many competitions' medal placements' will be determined by the random count of judges' marks, particularly in close competitions. Would it be right for an Olympic medal to be decided this way? I can picture it now: Another close event a la Kerrigan/Baiul. Cinquanta says, "Ah, but it's not the judges' fault. The competition was decided by fate! No one is to blame!"

If you haven't become too weary of a CoP discussion, perhaps we can have a discussion where you state one or two things that you like about the CoP and then we can analyze those, and then move on to another aspect of the CoP and analyze that, and so forth. Or, if you're way ahead of me (since I'm still learning), maybe you can either state your counterarguments to either Loosemore's or Rossano's findings on the statistical flaws in the CoP, or steer me towards your previous writings on the matter. This way we can tackle the CoP one step at a time, and you guys can help me understand and learn some things as well.