Men's Free Skate, Sat. 11/19 at 7:30 am EST

[Bluebonnet]

I see what you mean, but at the same time, I don't understand. We are told that PCS is, unlike 6.0, not used as a tool to rank skaters or reflect their reputations. These values, especially in categories like TR and SS, are supposed to be somewhat objective and quantifiable. Yet with this jump in PCS between the long and short programs, does this mean that skaters magically gain better SS within a 24-hour time period?

I chose Javier as an example because (although I do like his skating and am a fan), to me, he is an especially egregious example of this phenomenon. Last season (e.g. 2011 Worlds), Javi was a 6-range skater in PCS. After his short program at Skate Canada, Javi's score shot up to firmly in the 7-range. Now, he skated very well and I see how he has improved in many areas in his skating over the summer, like in CH. So the 7s seemed justified to me. But after the LP, BOOM! All 8s, except for TR. And he really didn't skate as well as he did in the SP either. Even as a fan, I didn't think the 8s were justified.

And yet we are told reputation, momentum, how you jumped, blah blah blah, aren't supposed to affect PCS like that under CoP. So what gives?

The more we go into detailed math, the more I'm confused about the number measuring method. It can't be absolutely accurate except probably BV in TES. Even BV has the trace of human brain processing in it with determinations on whether or not having under rotating, down grading, edge calls, level of spins and steps, and Credit for highlight distribution. SS and TR are absolutely not subjective-free. How do you measure a skill without using a ruler? It's a human brain processed result. It's not real math. Therefore, I think comparing and ranking are unavoidable. As long as the ranking is correct, it is correct. CoP and 6.0 are the same. The only difference is that CoP breaks down into details and see individual pieces rather than see the whole. As a result, CoP is more technical, less in relying on reputation, and more difficult to manipulate.

[SkateFiguring]

As I have expressed before, my thought is that PCS does resemble 6.0 in that there is a definitive ceiling on each mark, thus necessitating comparisons/ordinals. TES ceiling is more flexible and complicated to reach as it depends on elements executed, their BV, and GOE received. Judges are not deciding on the marks by comparing skaters on a holistic manner but are focusing on grading the performances of individual elements. With the tech panel making qualifying calls, the Total Scores are really not completely the judges' responsibilty or in their control. However, PCS are more than just ordinals in the PC but the marks themselves are important because they are added to form the Total Scores.

[Bluebonnet]

A bit too late for comments but I finally watched Rippon and Brezina's LPs.

Adam's skating was beautiful for sure. But his strokes were laboured. That has become part of his style. Didn't correct them at very young, now it's difficult to correct them. Oh, well, he and his fans might not think that was a problem. But to me, that makes his skating heavy, not as pretty as his look. Good that he has the guts to put in 4Lz every time.

Where was Brezina's quad that he blamed his coach for not letting him put in at SA? Generally, I like his skating style. He skated cautiously and some spins were very slow. The final spin was ugly.

[Mathman]

Don't.

I am surprised. But I think the ISU concurs.

IMO they are cutting off their nose to spite their face.

[wallylutz]

^^^ If you don't mind, wallylutz, I have a small question about PCS. I think some aspects of PCS should be somewhat fixed within a range--for instance, skating skills: Patrick Chan's skating skills marks should remain high even if he underperforms like he did this weekend.

Your understanding of SS as a part of PCS is incorrect. SS aren't "somewhat fixed" on the basis of the skater but rather contingent on the performance. A certain skater who has the potential to score in the 8s for SS provided he/she skates all out and pushed it to the edge may not get this score if the same skater holds back and skate below potential. It also seems to me you are confusing "underperforming" when it comes to elements such as jumps, which is what Patrick Chan mostly had issues with this past weekend. But a performance that is noted for technical errors such as stepping outs, falls and etc. on elements does not necessarily affect SS. Likewise, completing elements flawlessly is largely irrelevant as far as the criteria of SS are concerned. If a technical error is so severe that it disrupts part of the performance such as Oda's fall in his step sequence during last weekend's LP, that's another story and will rightfully affects his SS and other components as well. But a down and up fall will not have much of an effect. Another example I can give you is to look at Michelle Kwan's 1998 Olympics LP where she had no errors per say on her elements but because she was holding back and skated more cautiously, an argument can be made she displayed SS that was below her potential. Hence, if you feel that her maximum is around 9.00, then rightfully, such skate could be half a point lower. In other words, you will need to look at SS as though these single skaters are ice dancers.

But sometimes, I see big jumps in certain aspects of PCS within the same competition, and not just in more flexible things like PE and IN, which are more tethered to how a skater performed that day. I'm talking about things like Javier Fernandez at Skate Canada this year, when his SS mark jumped from 7.68 to a considerable 8.14 within the span of 24 hours from the SP to LP, yet I saw no clear improvements in this regard within the 24-hour timespan, and if I recall correctly, Fernandez actually skated with more speed and flow in the SP. I noticed this same phenomenon also happens when a lower-tiered skater unexpectedly makes the final group of a competition.

How is this justified?

I reviewed Javier Fernandez's SP and LP from Skate Canada 2011. First of all, I need to correct your perception that Fernandez skated with more speed in the SP vs. his LP, that turned out to be untrue. Here are the clips I used:

Javier Fernandez SC 2011 SP

Javier Fernandez SC 2011 FS

Right from the top, Fernandez flew across the ice in his FS, demonstrating far better flow and effortless glide than he did in the SP. More specifically, let's breakdown the various criteria of SS and see what are the differences between his SP and FS.

1) Balance/rhythm/precision: Pretty much the same between the two programs, this is an area where Javier needs to demonstrate a greater confidence in his ability to create beautiful patterns on ice. He comes across as somewhat tentative whenever he starts to execute complicate steps and slows down considerably and carries less flow. Precision of his skating should improve as he gains more mileage; right now, they look new and he looks cautious. Suggested Score: 6.75

2) Flow/Speed: Edge to his FS. The design of his SP, while entertaining to see, failed to showcase Fernandez's respectably powerful glide. Although well skated, it makes it hard from a judging standpoint to reward an aspect that was not clearly shown. Given his general cautiousness when executing intricate connecting steps, there were relatively few occasions to properly assess his flow & speed in the SP whereas his FS properly displayed his ability in this area beyond doubt. Suggested Score: SP = 6.50 LP = 7.75

3) Cleaness/Sureness/Edge Quality: Edge to his FS. It can be hard to dissociate technical errors on elements when evaluating PCS but it must. Despite the name cleaness/sureness, this has actually nothing to do with falling or stepping out of an element (e.g. jump). Rather, this criterion looks at the sureness and depth of the skater's edges. Javier Fernandez's edge quality still requires some improvement. Judging him as though he is an Ice Dancer, it is noted that his edge is not very secure at times and lacks a bit of depth. However, his FS was designed in a way that gave him more chances to demonstrate cleaner running edges than his SP. The fact that a FS is much longer than a SP often gives skaters more opportunity to demonstrate skills that they otherwise didn't have time to show in a packed SP, hence, quite often, the PCS will go up in the FS vs. SP. Suggested Score: SP = 6.75 LP = 7.25

4) Variety in power/acceleration: Edge to his FS. Javier showcased relatively few variety in his overall skating quality during his SP where the power and mostly, acceleration, remains at mostly same level. This is however considerably better in his LP where there were more clearly defined different levles throughout the different parts of his FS. Suggested Score: SP = 6.50 LP = 7.75

5) Multi directional skating: About the same between the two programs. There were some changes of direction in his skating but they tend to remain in the same direction most of the time. This is an area that Javier can significantly improve as well. While his performance are entertaining to watch, from the perspective of a sport as opposed to a show, he needs to show a greater ability to skate in unpredicted directions with more changes than he currently does. Suggested Score: 7.00

6) Skating on one foot: Edge to his SP. The composition of his SP is more intricate and showcased some very interesting use of turns and steps throughout his performance. His FS continues to demonstrate an above average use of connecting steps and skating on one foot but overall, slightly less intricate than his SP. Suggested Score: SP = 8.25 LP = 8.00

All things considered, I would give him Skating Skills score of 7.00 in the SP and 7.50 for his LP. The panel averaged to about 0.50 of difference as well which seems just about right to me.

I hope this helps.

[evangeline]

Thank you for taking the time to break that down, wallylutz, I really, really appreciate it. Just a wondering about something though. You mentioned that the panel averaged to about 0.50 of difference to your scores (which, to be honest, I personally agreed with more than the actual averaged marks given to Javier), and that was OK. So I was guess if Javier averaged 6.50 in the SP and 7.00 in the LP for SS by the judging panel, that would be ok too. This seems a bit strange to me as this a rather broad range of acceptable marks, especially since it refers to numbers that are averaged.

Also, I was wondering about people like judges #3 and #7, who gave Javier a 8.75, or judge #1, who gave a 8.50, all considerably above the 0.50 difference. Do they get written up after the competition?

[wallylutz]

Thank you for taking the time to break that down, wallylutz, I really, really appreciate it. Just a wondering about something though. You mentioned that the panel averaged to about 0.50 of difference to your scores (which, to be honest, I personally agreed with more than the actual averaged marks given to Javier), and that was OK. So I was guess if Javier averaged 6.50 in the SP and 7.00 in the LP for SS by the judging panel, that would be ok too. This seems a bit strange to me as this a rather broad range of acceptable marks, especially since it refers to numbers that are averaged.

Also, I was wondering about people like judges #3 and #7, who gave Javier a 8.75, or judge #1, who gave a 8.50, all considerably above the 0.50 difference. Do they get written up after the competition?

The question you asked was why was the SS score from the SP lower than his FS? After reviewing both performances, I concluded that it was justifiable based on the breakdown of SS criteria. Marks are relative numbers and each person has his/her own zone for the scores. Some are more lenient and others, more strict. The important thing is consistency, not the absolute number. This philosophy was carried over from the 6.0 era when ordinal ranking was king and a judge's 5.5 may be the same as another's 5.8 Javier Fernandez may have scored an average of 7.6 for his SS but relatively, Patrick Chan got an average that was almost 9.0 Although the ISU tries to keep a certain range by training the judges to be as consistent with each other as possible, in reality, when you have 9 people out there, you are going to have different standards just as 9 MLB umpires will have unique strike zones for each of them. Don't be surprised by the ranges of the spread in marks as it is usually difficult to determine whether there is an outlier by simply looking at the scores. Someone who gave Fernandez 8.75 for SS may not necessarily be wrong, for all we know, he could have given Chan 9.5 and Takahashi. 9.25, and therefore, really placing Fernandez as 3rd best in the SS category. It's all relative. Then, there are judges who gave out noticeably low scores such as Plushenko's 5s for TR during the Olympic Games to make a point and those are obviously prepared to defend their decision, which is fine as well if they can justify it.

[Blades of Passion]

I reviewed Javier Fernandez's SP and LP from Skate Canada 2011. First of all, I need to correct your perception that Fernandez skated with more speed in the SP vs. his LP, that turned out to be untrue. Here are the clips I used:

Javier Fernandez SC 2011 SP

Javier Fernandez SC 2011 FS

Right from the top, Fernandez flew across the ice in his FS, demonstrating far better flow and effortless glide than he did in the SP. More specifically, let's breakdown the various criteria of SS and see what are the differences between his SP and FS.

A very bad and incorrect analysis. You really are grasping to defend poor judging and the precious ISU, aren't ya?

Just because he starts his Long Program with quicker movements, that doesn't mean he is showing superior ability to create flow and effortless glide. If you pay attention, he creates very good flow and effortless glide after the opening stationary steps in his Short Program.

Cleaness/Sureness/Edge Quality: Edge to his FS. It can be hard to dissociate technical errors on elements when evaluating PCS but it must. Despite the name cleaness/sureness, this has actually nothing to do with falling or stepping out of an element (e.g. jump).

NO. No, no, no, no, no. Every edge and every movement a skater makes is part of their Choreography and their display of Skating Skills. The program and the skater's ability does not suddenly stop when a jump or another technical element happens. Everything is part of a whole. If a skater shows poor edging as a result of a bad jump exit, then it must be accounted for within the assessment.

[skatinginbc]

The fact that a FS is much longer than a SP often gives skaters more opportunity to demonstrate skills that they otherwise didn't have time to show in a packed SP, hence, quite often, the PCS will go up in the FS vs. SP.

Interesting theory. To test the hypothesis, I used all the data from this year's ISU GP competitions so far and the results are as follows:
Men's short program skating skills: Mean = 6.98514, Standard Deviation = 0.82103
Men's long program skating skills: Mean = 7.02405, Standard Deviation = 0.84173
p = 0.35025

Women's short program skating skills: Mean = 6.41128, Standard Deviation = 0.70831
Women's long program skating skills: Mean = 6.57821, Standard Deviation = 0.75841
p = 0.01736

Conclusions: The difference in SS scores between men's short and long is statistically insignificant, although it is significant for women (about 0.16 difference).

[wallylutz]

Interesting theory. To test the hypothesis, I used all the data from this year's ISU GP competitions so far and the results are as follows:
Men's short program skating skills: Mean = 6.98514, Standard Deviation = 0.82103
Men's long program skating skills: Mean = 7.02405, Standard Deviation = 0.84173
p = 0.35025

Women's short program skating skills: Mean = 6.41128, Standard Deviation = 0.70831
Women's long program skating skills: Mean = 6.57821, Standard Deviation = 0.75841
p = 0.01736

Conclusions: The difference in SS scores between men's short and long is statistically insignificant, although it is significant for women (about 0.16 difference).

Perhaps you should expand your pool of data to include the GP results from the last 3 years and see what gives. Do a separate one for ISU Championships (GPF Final, Europeans, 4CC and Worlds but exclude Olympics) and compare. Please let us know what you find.

[skatinginbc]

Perhaps you should expand your pool of data to include the GP results from the last 3 years and see what gives. Do a separate one for ISU Championships (GPF Final, Europeans, 4CC and Worlds but exclude Olympics) and compare. Please let us know what you find.

The beauty of hypothesis testing with a t-statistic is that I don't have to know the population parameters. If the preliminary result looks promising, then there might be an incentive for me to expand the sample size. But the result did not appear to have either a practical significance or a statistical significance. So, well, I'll leave it for others who are willing to spend a great deal of their precious time on this.

[SkateFiguring]

A trivia: All three Men on TEB podium were born in the same year.

Brezina - March 30,1990
Song - August 9, 1990
Chan - December 31, 1990

Also, all TEB 2011 champions are exactly the same as at SC.

[Art&Sport]

A trivia: All three Men on TEB podium were born in the same year.

Brezina - March 30,1990
Song - August 9, 1990
Chan - December 31, 1990

Also, all TEB 2011 champions are exactly the same as at SC.

Thanks for those interesting dates/ numbers SkateFiguring -- a lot more interesting than the eye-glazing stuff the posters before you have been spewing out. Wow, Brezina is older than Song by 4 months and about 8 or 9 days (pls help me out math whizzes) and older than Chan by 9 months and 1 day (correct?). I wonder if that gives Brezina any kind of edge with the judges? But, oh I forgot, Chan is the one with the holier than thou edges.

[jettasian]

Thanks for those interesting dates/ numbers SkateFiguring -- a lot more interesting than the eye-glazing stuff the posters before you have been spewing out. Wow, Brezina is older than Song by 4 months and about 8 or 9 days (pls help me out math whizzes) and older than Chan by 9 months and 1 day (correct?). I wonder if that gives Brezina any kind of edge with the judges? But, oh I forgot, Chan is the one with the holier than thou edges.

We all know you don't like Chan. No need every opportunity to make your point cross.

[skatinginbc]

I agree with Jettasian. Bringing up Chan's "holier than thou edges" sounds sarcastic--simply a bad joke.