Those Little Errors in Scoring

[Mathman]

Easy question: Should Sheldon Cooper be a single skater, pair skater, a tech guy or a judge? All wrong answers yield hilarious scenarios!

Sheldon (my hero Plus, he looks a little like Johnny Weir) = singles skater

Leonard and Penny = pairs skater

Raj = tech specialist

Howard = judge (he is so indecisive he never comes to a conclusion, so he has to ask his mother)

 

[Mathman]

Now what does the tech guys do once they have watched the feet? Announce to the judges that the Level is such and such because of his expertise in Musicality?

Joe, seriously, why are you pretending not to understand the difference between what the technical panel does and what the judges do?

Musicality = judges.

How many turns in the clockwise direction (one, two, three...) = technical panel.

Here are some judges’ marks from a recent competition for the same skater:

3A+2T (+1.14) 0 1 1 2 2 1 0 1 2
3A (-1.86) -1 -2 -2 -3 0 -2 -2 -2 -2
3F+3T (-0.10) -1 1 0 0 2 -1- 0 -1 0

Judge #5 is outside the corridor three times in a row. Ten more and he will have a lot of 'splainin' to do to the referee and the judges' oversight panel.

If you would name the skater, maybe we could guess the nationality of judge #5?

 

[seniorita]

Penny is kind of large for pair skating with Leonard
And Sheldon looks like Johnny Weir? H o W?

 

[Violet Bliss]

I watched the early part of Ice Princess for the first time yesterday. Apparently a Physicist can become an advanced skater in no time by applying the theory of Physics in rotating jumps. Makes Sheldon being a skater not so far fetched! It's true Physics rule everything!

Patrick said his least favorite subject was Physics. Good thing there is Kristy Krall and her Dartfish software as developed by our Ice Princess.

And Sheldon looks like Johnny Weir? H o W?

The same way Rachael's homie escort looked like Adam Rippon.

 

[seniorita]

Apparently a Physicist can become an advanced skater in no time by applying the theory of Physics in rotating jumps.

Something like if you know the electromagnetic laws you can light a bulb with your finger?
In the Princess movie, she lands also a 2axel in a week, passes the selectionals and takes the zamboni boy, also Santa Claus does not exist

 

[Joesitz (RIP)]

Joe, seriously, why are you pretending not to understand the difference between what the technical panel does and what the judges do?

Musicality = judges.

How many turns in the clockwise direction (one, two, three...) = technical panel.

MM. Seriously, why can't the judges or even the Tech Panel do both Musical Timing and Turns to the Left and Right? You did mention something that the Judges can not focus on both. What were those judges doing before CoP? I don't think you were involved with the 6.0system before Kwan. There has been no great change in straight line footwork or any other element because of the CoP. The combo jumps and spins have been around for a long long time. The judges saw them then and still see them. In fact CoP has basically eliminated scoring some elements, so the burden is less for the judges. It's difficult to believe that the judges lose focus and can not see the differences between musicality and clockwise turns? While it may be easier for them to be at the convention dinner, they have a job to do.

 

[Joesitz (RIP)]

To ALL in favor of the existence of the Tech Panel.

I am not against what they are assigned to do. My question is where was the Tech Panel since the Turn of the 20th Century? The Judges did it all and they produced World and Olympic Champions. What is the reason for change after more than 100 years? I don't see it as an improvement in scoring. If a judge can not focus on the competition, that judge should be replaced.

Maybe it's just to keep the fans busy with additional numbers to play with.

 

[Mathman]

MM. Seriously, why can't the judges or even the Tech Panel do both Musical Timing and Turns to the Left and Right?

They could. That would be OK with me.

I understood this thread to be about questions involving how the CoP scoring system works. The first post on the thread asked how are various errors are scored under the present system. I tried to look up the answers on the ISU web page, and to look at protocols of actual competitions to see how the rules played out in practice.

If instead the thread is about, should we abolish the CoP and either go back to 6.0 or try something else, well, that's a different story.

Mathematically speaking the only system that makes sense is unelaborated ordinals: this skater was best, that skater was second best, etc. It is the human factor -- who sez so?! -- that messes up the elegant mathematical simplicity of figure skating judging. (The world would be perfect if there were no people in it. )

 

[Mathman]

To ALL in favor of the existence of the Tech Panel....

What is the reason for change after more than 100 years?

I am neither in favor of nor opposed to anyone's existence, but if you agree not to jump on me for giving my opinion about the question that you asked, here it is.

Although the idea of the CoP had been worked on earlier, it was the Salt Lake City judging scandal that served as the spur to rush the new judging system into service. Skating received a black eye, fans either became irate or lost interest, and the International Olympic Committee was on Cinquanta's case like white on rice.

The crux of the scandal (at least the ISU spun it this way) was that those lyin', cheatin. schemin' national federation officials routinely coerced their judges to wave the national flag, to inflate the scores of skaters from their country, and to enter enthusiastically into deals with other federations to their mutual gain.

What the tech panel does is take half of the power out of the hands of the judges (those biased homies) and put it safely in the hands of objective and heroic experts who are beholden to the ISU alone.

 

[gkelly]

anel will not necessarily be the same for each skater across a season because there are times when skaters get ahead of themselves and don't skate the step sequence as well as they did the competition before (due to nerves or excitement or whatever) and the complexity/variety of turns changes because instead of doing a counter, a skater rotates, changes edge and does a three turn instead or they put their foot down when approaching a turn or immediately after as a balance check and the tech panel can't count it as it was intended. Also, perhaps they are a little stiff and what WAS full body movement becomes modest body movement.

All true.

Also, if a skater falls in a step sequence, I believe the level has to be dropped.

Sorry, I can't easily check the rules on this from the computer I'm posting from tonight, but I don't think that's necessarily true.

If a skater falls in a step sequence, the judges need to drop the GOE from what it would have been otherwise.

The technical panel is not required to drop the level from what the skater earned before and/or after the fall. If he completed three features according to the requirements, he earns level 3, fall or no fall.

It's quite likely that a skater who was trying to achieve level 3 on his step sequence but fell in the middle will not successfully complete all three intended features. Just leaving out one of the planned steps and turns could disqualify it for the "variety" feature.

So it's very likely that a skater who falls on a step sequence will earn a lower level for the sequence that day than he earned at a different competition when he didn't fall. But it depends on what he actually succeeds in executing. There's no rule that requires that he not get full credit for what he did execute.

MM. Seriously, why can't the judges or even the Tech Panel do both Musical Timing and Turns to the Left and Right? You did mention something that the Judges can not focus on both. What were those judges doing before CoP?

They were trying to evaluate all the aspects of the step sequences, but there were no official rules or standards about what should earn more credit -- just general guidelines and judges' own opinions. There was no separate score for the step sequence so no one knew whether any judge had taken it into account at all or scored only on, say, jumps, speed, and presentation.

The very best judges of step sequences might have been able to take into account all the details that are now considered by tech panel and judges and weigh them consistently for each and every skater. But there still was no way to let anyone know everything they took into account, and once the program was over and the score entered, the judge would have forgotten most of the reasons for how her evaluation of the step sequence figured into the score for the program as a whole.

Most judges would probably focus on only their favorite few aspects of step sequences, or on whatever was most obvious about each sequence (which would make it hard to be consistent from one skater to the next).

I don't think you were involved with the 6.0system before Kwan. There has been no great change in straight line footwork or any other element because of the CoP.

Actually there have been significant changes in step sequences, especially straight line sequences, because of the CoP. For better and for worse. In general, the sequences tend to be more complicated, to show deeper edges, and to curve or wander back and forth across the ice rather than sticking to a pretty much straight line. They tend to take significantly more time to travel from end to end of the ice and to use fewer small jumps, toe steps, lunges, quick steps on toes or flats, and other such moves that don't gain points, while using more of whatever the rules for level features reward each year. On average, CoP sequences are probably less musical than 6.0 sequences, although there's always a range under both judging systems.

Mathematically speaking the only system that makes sense is unelaborated ordinals: this skater was best, that skater was second best, etc. It is the human factor -- who sez so?! -- that messes up the elegant mathematical simplicity of figure skating judging. (The world would be perfect if there were no people in it. )

Well, ordinals are very simple when there's only one judge judging, which would put the whole event under the power of that judge's biases.

Once you get, say, 24 skaters and 9 judges with different preferences and pet peeves, the numbers can get very messy indeed.

 

[Mathman]

Actually there have been significant changes in step sequences, especially straight line sequences, because of the CoP...

On average, CoP sequences are probably less musical than 6.0 sequences.
although there's always a range under both judging systems.

I think this is the point of the Choreography Footwork Sequence, is this right?

It is the same fascinating dilemma -- sport or art?

In a sport you want everyone to do exactly the same thing and you give the prize to the one that does it best (runs fastest, jumps highest, lifts the heaviest weight).

In art, originality and harmonious composition are highly prized.

So now a skater can do the regular step sequence, squeezing in as many twists and turns as possible to achieve a high level. Then he can do his ChStep1 or ChSpiral1 and concentrate on musical expression and compatibility with he theme of the program.

They can have their cake and eat it, too...just not both at the same time.

Well, ordinals are very simple when there's only one judge judging, which would put the whole event under the power of that judge's biases.

Once you get, say, 24 skaters and 9 judges with different preferences and pet peeves, the numbers can get very messy indeed.

There is actually a mathematical theorem called "Arrow's Impossibility Theorem" (Arrow won the Nobel prize for economics in 1972) that says it is impossible to design a ranked voting system for three or more skaters which

(a) is not a dictatorship;

(b) has the feature that if every individual judge prefers A to B, then the judgment of the group will always rank A ahead of B;

(c) does not allow flip-flips. (This was the bane of both majority-of-ordinals and OBO).

http://en.wikipedia.org/wiki/Arrow's_impossibility_theorem

Edited to add: Sorry, I can't help myself. (b) is called the Pareto condition. In ecomonics it says that if a change in public policy advances the fortune of every individual member of society, then it is good for the commonweal as a whole. Unfortunately, this is not always true, as in the example of heavy borrowing by the government.

 

[Mathman]

This is why there are three people on the tech panel - each one has certain bullets he/she looks for in the step sequence to determine the level of the sequence. They decide before the event who is looking for which bullets - TC may decide he/she is looking for at least 1/3 of the turns in each direction and body movement while the TS is looking for variety/complexity of turns and step sequence at least 1/2 on one foot and the ATS is looking for number of rotational turns in each direction and quick change of directions.

Thank you for this post. A ray of light amid all the heat. This is very useful information to know. :thumbsup:

 

[mskater93]

If all three were looking at all bullets, it would be hard to pinpoint #of features. I've seen some of the scratchy notes from an event I participated in where they called and negated features for spins. For example, I had a COE in one spin but it was short rotation and the TS put a line through it one way on her sheet and another where they ended up debating "clear increase of speed" but ended up not calling it after lengthy discussion, this got a line in the other direction on her short hand notes since there was a tech and judge critique to the event. Back entry got circled with her short hand notes and 8 revs got circled and I ended up with a L2.

As for how things were evaluated under each juding system, 6.0 was a holistic approach by the judges. The entire program was evaluated as a whole and one tech and one presentation score on a relative scale was presented for each program and the reasons for each score were quickly forgotten other than some sketchy notes. IJS is a very piece meal approach to scoring in that each piece gets points and they all add up to a number on the end and the biggest number wins. I think this is why so many viewers dislike IJS - because everything is about gaining points, programs that win tend to have a lot of "stuff" but holistically speaking don't make sense. Kinda like getting up close to a Monet where all those little dots don't make sense unless you back away.

 

[Mathman]

IJS is a very piece meal approach to scoring in that each piece gets points and they all add up to a number on the end and the biggest number wins. I think this is why so many viewers dislike IJS - because everything is about gaining points, programs that win tend to have a lot of "stuff" but holistically speaking don't make sense. Kinda like getting up close to a Monet where all those little dots don't make sense unless you back away.

I think the ISU was under a lot of pressure to make figure skating "more like a real sport." What is "unsportly" about skating is that it is judged. In a "real sport" there is a score-keeper to count how many runs or baskets or goals you score. You don't win the game just because someone is impressed with the aesthetic feel of your performance taken as a whole.

Hence the technical panel.

---------------

Thanks for sharing the information about the tech specialist's short-hand notations on your score card. That is very interesting!

 

[mskater93]

Thanks for sharing the information about the tech specialist's short-hand notations on your score card. That is very interesting!

You're welcome. If it sheds some light on how things are done and stops some of the ideas that the TP is clueless, I am more than willing to share it...

 

[Violet Bliss]

I looked for the protocol of the GPF and saw it was not up in the ISU site. Then I recalled Mathman mentioned marks on Chan's 4T in SP and I even responded to that with mention of marks of Oda and Vener.

Where can one find those marks again?

eta Never mind. Found it.