Program Component Scores

[morninglight]

How true would you think the following statements are?

1. Score on Skating Skills follows the Technical Score.

2. Scores on Transitions, Performance/Execution, Choreography, and Interpretation follow the score on Skating Skills.

I suppose that there would always remain some levels of ambiguity in Program Component Scores. But how much clearer could judging of Program Component Scores be in the near future, and how? I think that the ISU definitions of these components are pretty clear, but judging does not appear to be so much clear to me.

Here's a link to:
the ISU definitions of Program Components

 

[Joesitz (RIP)]

The most important aspect of PCS scores is that they are extremely subjective. It's a case of one man's feast is another man's poison. Having so many judges of which some scores are not used in the competition, others are dropped because of high and low scores, we are still left with the human subtleties of: I much preferred Bambie to Gone With the Wind. (I am not writing about national prejudices, collusions, or cheating here. Just how humans behave.)

I kind of think much of the skating skills are covered in the Technical. One gets marked down in GoEs for lack of transitions into elements, for example. Choreography is mainly how well the hired hand develops the program.

I find the PCS definitions clearly written. However, I do find them ambiguous as they are often covered in the Technical. I would think a special study of PCS is needed. Perhaps combining those not so ambiguous into the technical and come up with one score per skater and not two.

Joe

 

[enlight78]

The most important aspect of PCS scores is that they are extremely subjective. It's a case of one man's feast is another man's poison. Having so many judges of which some scores are not used in the competition, others are dropped because of high and low scores, we are still left with the human subtleties of: I much preferred Bambie to Gone With the Wind. (I am not writing about national prejudices, collusions, or cheating here. Just how humans behave.)

I kind of think much of the skating skills are covered in the Technical. One gets marked down in GoEs for lack of transitions into elements, for example. Choreography is mainly how well the hired hand develops the program.

I find the PCS definitions clearly written. However, I do find them ambiguous as they are often covered in the Technical. I would think a special study of PCS is needed. Perhaps combining those not so ambiguous into the technical and come up with one score per skater and not two.

Joe

I agree with that. If a technical specialist can give levels to spins and step sequences. Why not have a seperate specialist give levels to transitions, skating skills, interpetaion, choreography. Then just have judges give GOEs. Then would only have one score. I don't thing we need performance execution, that is reflected in the GOE's and I don't believe you can reasonably deffine levels for P/E.

 

[Joesitz (RIP)]

I agree with that. If a technical specialist can give levels to spins and step sequences. Why not have a seperate specialist give levels to transitions, skating skills, interpetaion, choreography. Then just have judges give GOEs. Then would only have one score. I don't thing we need performance execution, that is reflected in the GOE's and I don't believe you can reasonably deffine levels for P/E.

I'll buy that. My only problem with the Technical or a possible PCS Specialist is questioning if they are fully qualified to give levels.

Secondary are they themselves favoring a particular skater, e.g., are Scott Davis and Viktor Petrenko thinking the same way? Remember there is really only one Specialist because any assistant has been completely numb as of today.

Joe

 

[debdelilah]

The most important aspect of PCS scores is that they are extremely subjective. It's a case of one man's feast is another man's poison. Having so many judges of which some scores are not used in the competition, others are dropped because of high and low scores, we are still left with the human subtleties of: I much preferred Bambie to Gone With the Wind. (I am not writing about national prejudices, collusions, or cheating here. Just how humans behave.)

I kind of think much of the skating skills are covered in the Technical. One gets marked down in GoEs for lack of transitions into elements, for example. Choreography is mainly how well the hired hand develops the program.

I find the PCS definitions clearly written. However, I do find them ambiguous as they are often covered in the Technical. I would think a special study of PCS is needed. Perhaps combining those not so ambiguous into the technical and come up with one score per skater and not two.

Joe

I'm not sure how you could get rid of the scoring ambiguity without getting rid of the art, which is inherently subjective. Skating is competitive above all, so if artistry(of whatever type) was no longer part of the score, it would probably begin to fade out of competitive skating.

I'm not at all sure that choreography is all the work of the hired hand. Nicolai Morozov has choreographed for many, many skaters--including Brian Joubert and Kimmie Miessner. I read an interview that he did in 2003, where he was asked whether he might be overextending himself, with his programs losing quality or becoming too similar. He said that he does recycle footwork sequences, etc, but that they don't come off the same because skaters perform them differently. He was criticized about his program for Joubert coming off as a poor copy of his program for Yagudin, and he said this was because Yagudin was a better performer, not because the choreography was worse. (You can see the interview translation here: http://p100.ezboard.com/falexeidiscussiongroupfrm24. The thread title "Nicolai Morozov interview"
is near the bottom of the page).

I actually really like the system of having professional choreographers--not every actor is a good playwright, but the two need each other to shine. The skaters are like actors performing a script, in a sense--anyone who can read a script can get through performing it in some way, but some will do it much better than others.

I could see Daisuke maybe wanting to do choreography himself later in life--he choreographed his Secret Garden exhibition himself a couple of years ago, and said that doing that revived his interest in skating.

 

[enlight78]

I'll buy that. My only problem with the Technical or a possible PCS Specialist is questioning if they are fully qualified to give levels.

Secondary are they themselves favoring a particular skater, e.g., are Scott Davis and Viktor Petrenko thinking the same way? Remember there is really only one Specialist because any assistant has been completely numb as of today.

Joe

I agree that those would be the issues. I also don't think the outcome would be a whole lot worse. I think the scores would be easier to identify or at least encourage skaters to put more transitions and choregraphy in their program to get the high level what ever it may be. As long as the PCS specialist is not Dick
Button I'll have my hopes up.

 

[Hsuhs]

How true would you think the following statements are?

1. Score on Skating Skills follows the Technical Score.

2. Scores on Transitions, Performance/Execution, Choreography, and Interpretation follow the score on Skating Skills.

I could calculate correlations between 2 sets of scores, to find it out. But what scores should I use? From what competition?

 

[gkelly]

I could calculate correlations between 2 sets of scores, to find it out. But what scores should I use? From what competition?

I would suggest Junior Worlds, because it's a large data set and most of the skaters are not well known so reputation wouldn't be as significant a factor.

 

[Joesitz (RIP)]

I agree that those would be the issues. I also don't think the outcome would be a whole lot worse. I think the scores would be easier to identify or at least encourage skaters to put more transitions and choregraphy in their program to get the high level what ever it may be. As long as the PCS specialist is not Dick
Button I'll have my hopes up.

Why would you worry about Button? Have you heard that he will be acting as a Technical Specialist? If not why bring it up? There must be some reason you wanted to.

If Dick did express an interest in being a Tech Spec., I can't imagine him being biased. He loves the Sport and everyone in it. Like most Americans, he is taught that the best man should win - not the Country.

I think most posters should leave the Dean of Figure Skating alone. He's forgotten more than anyone else knows.

Joe

 

[morninglight]

statistics

I could calculate correlations between 2 sets of scores, to find it out. But what scores should I use? From what competition?

1. Technical Score vs. Skating Skills

I am not quite sure if a bivariate correlation between TS and SS is the one that helps us. But they would surely be highly correlated, given that the skaters who are technically advanced in elements tend to have greater skating skills. But still, it is not always true. So the question that I am asking is whether skating skills are evaluated by itself, regardless of the difficulty of elements, or difficulty of elements is doubly rewarded .

2. Skating Skills vs. Transitions, Performance/Execution, Choreography, and Interpretation

I saw graphs that plot SS vs. (TR+PE+CH+IN)-4×SS.

The reference (In Japanese):
http://www.geocities.co.jp/Athlete-Athene/9074/index.html
Figure skate shiryo-shitsu. Under the column, "Sougou Shiryo 1", "Kakuzuke Memo."

What it tells is that whereas about half of ice-dancers get higher scores on TR, PE, CH, and IN than SS, the vast majority of the single and pair skaters get lower scores on TR, PE, CH, and IN than SS. Very few single skaters such as Yukina Ota, Alissa Cziny, Sasha Cohen, and Jeff Buttle get higher or nearly equal scores on TR, PE, CH, and IN compared to SS. The issue here appears to be that a single skater, on average, gets lower TR, PE, CH, and IN scores than SS score. So it looks like TR, PE, CH, and IN follow SS; and SS may be doubly rewarded.

In some cases, it may end up that athletically more advanced skaters get relatively high credit for their artistry than artistically more advanced skaters who are trained in ballet and dancing and spend lots of energy and time to enhance presentation skills.

So the question again is whether it is possible to evaluate Transitions, Performance/Execution, Choreography, and Interpretation by themselves, apart from Skating Skills .

 

[Hsuhs]

I would suggest Junior Worlds, because it's a large data set and most of the skaters are not well known so reputation wouldn't be as significant a factor.

I think it's a good idea. I'll do that.

I am not quite sure if a bivariate correlation between TS and SS is the one that helps us.

It could be a beginning.

The reference (In Japanese): http://www.geocities.co.jp/Athlete-A...074/index.html Under the column of "Kakuzuke Memo."

I can't see any graphs there.

So the question again is whether it is possible to evaluate Transitions, Performance/Execution, Choreography, and Interpretation by themselves, apart from Skating Skills .

I don't know, but maybe a factor/ principal components analysis could produce an answer.

 

[morninglight]

I can't see any graphs there.

The page brings you to the homepage of "Figure Skate Shiryo-shitsu" (In Japanese). Unfortunately there is no direct link to the page that displays the graphs that I was referring to. You have to find a title "Kakuzuke Memo" under the column of "Sougou Shiryo 1." Unless you read Japanese, it may be difficult. But I wanted to at least cite the source to backup. If the link is not working properly, you can go to Yahoo Japan and go to Sport / Figure Skate / Related Sites / "Figure Skate Shiryo-shitsu".

Yet, these graphs just backup what we see in competitions. If a single skater gets 7.0 on SS, for example, she/he would have similar scores on the other program components, which are on average a little bit lower than SS. If she/he gets 6.0, 5.5, or 5.0... on SS, the same rule applies.

But again, not necessarily these numbers or statistical analyses themselves, I am more interested in how judges evaluate these components.

 

[Mathman]

I don't know, but maybe a factor/ principal components analysis could produce an answer.

:yes: I hope you'll find the time to do it. It would be very interesting.

However, I agree with Morninglight's last statement. I do not believe that any statistical analysis will be able to address the most interesting question -- all this correlation that we are seeing between various component scores, is it because the judges are not applying the criteria discriminately? Or is it rather simply a reflection that, on the average, the better skaters in one category tend also to be the better skaters in all?

 

[Hsuhs]

You have to find a title "Kakuzuke Memo" under the column of "Sougou Shiryo 1." Unless you read Japanese, it may be difficult. But I wanted to at least cite the source to backup.

The problem is, my comp doesn't recognize Japanese characters. I can read a word or two if it's kana. I trust you on this, still I can't see it.

The good news is there is a moderate correlation between the Skating Skills scores and the TESs, for these numbers:

http://www.isufs.org/results/wjc2007/wjc07_JuniorMen_FS_Scores.pdf

it's ,775 (p<0.01, 2-tailed)

 

[Hsuhs]

I hope you'll find the time to do it. It would be very interesting.

I was hoping you'd do it!
The trial version of SPSS is available online. It's a fun program, I promise.

 

[iloveaxel]

1. Technical Score vs. Skating Skills

I am not quite sure if a bivariate correlation between TS and SS is the one that helps us. But they would surely be highly correlated, given that the skaters who are technically advanced in elements tend to have greater skating skills. But still, it is not always true. So the question that I am asking is whether skating skills are evaluated by itself, regardless of the difficulty of elements, or difficulty of elements is doubly rewarded .

Of course technical scores are highly correlated with skating skills. One important aspect of SS is to evaluate skaters' ability to implement difficult elements at high speed. Speed is really a key to separate skaters apart. If you are slow, it'll be much easier to do the jumps, throws & lifts etc. If you can only do a bunch of doubles or some simple triples, no matter how stretched your line is, it's very difficult to give you a high SS score.

The idea to separate technical scores and PCS is just laughable. Again, the same example, you can have the most complicated transition in the world, but if your program only includes a bunch of double jumps, what's the use for those transitions? You still should not get a high score in transition because there's no way for judges to evaluate whether you can still do triple jumps while keeping those difficult transitions.

All these stuff are inter-connected.

 

[morninglight]

all this correlation that we are seeing between various component scores, is it because the judges are not applying the criteria discriminately? Or is it rather simply a reflection that, on the average, the better skaters in one category tend also to be the better skaters in all?

Thank you for the good summary. I am somewhat doubtful if the judges are really applying the criteria discriminately. Choreography, interpretation, transition, and performance scores seem to me to be, more or less, automatically inflated for skaters who have greater speed on the ice. If a single skater gets 7.5 for skating skills, for example, she/he would perhaps never get 6.0 on choreography or interpretation.

 

[Hsuhs]

I wish there was a way to find out if it's true, b/c I'm somewhat uncomfortable with conspiracy as the only explanation.

The correlation between the Skating Skills scores and the sum scores (TR+PE+CH+IN), for the same data, is .749

 

[enlight78]

Why would you worry about Button? Have you heard that he will be acting as a Technical Specialist? If not why bring it up? There must be some reason you wanted to.

If Dick did express an interest in being a Tech Spec., I can't imagine him being biased. He loves the Sport and everyone in it. Like most Americans, he is taught that the best man should win - not the Country.

I think most posters should leave the Dean of Figure Skating alone. He's forgotten more than anyone else knows.

Joe

I was just joking based on some of his commentary.

 

[enlight78]

The idea to separate technical scores and PCS is just laughable. Again, the same example, you can have the most complicated transition in the world, but if your program only includes a bunch of double jumps, what's the use for those transitions? You still should not get a high score in transition because there's no way for judges to evaluate whether you can still do triple jumps while keeping those difficult transitions.

All these stuff are inter-connected.

I disagree. According to my interpitation of the rules. A skater should be able to get a resonable high score in P/E, transitions, and interpitaion. Transitions is about connecting the elements period. It has nothing to do about how difficulty the element is. Skaters use transitions between spins, footwork, and jumps. If I follow your logic. Skater must do nothing but triple jumps after transitions to get credit for them.