Those Little Errors in Scoring

I presume they are not in the class of URs, WETs, or Falls.

They are definitely noticed by fans and judges alike.

How serious in scoring are they marked?

Jumps

Rotations - Is there any counter rotation in the air when it is required?

Landings - Those curlicue (my name) when the skater lands a wonky jump and make a semi loop headed in the wrong direction of outflow when it should have flow out?

- Those wonky jumps when a skater steps out of his landing?

- Those two foot landings?

- Those toe-pick landings?

- Other?

Spins

Those Travelling Spins?

Those less than Required rotations?

Those with special positions - are they correct? Which is correct - the head only or the torso?

Those 'combo' spins - do they have sufficient rotations for all parts of the combo?

Other?

I'm not sure what you're asking. How does the scoring system handle small errors on elements?

Most of the examples you listed would be reflected in the Grades of Execution.

For some of the spin examples you give, they could make the difference in whether the tech panel gives credit for an intended feature and calls a higher level or not.

Turnouts are usually -1
Step outs/Two-foots are usually -2 (unless the judge doesn't see the two-foot)
Bad landings are usually -GOE

Travelling spins are -GOE
Less than required rotations is deducted in levels, even down to a level 1
This was posted at FSU: Akiko Suzuki's flying camel combination spin was called a level 4 at Finlandia, but only level 1 at Cup of China.
http://www.youtube.com/watch?v=y3FPo1XQfiE (starting @ 1:12)
http://www.youtube.com/watch?v=Ybb6_1Qhhk4 (starting @ 1:31)
At CoC she didn't do enough revolutions in the cannonball position (required to be at least 2)

Those are good examples of the use of negative GoEs but we don't know exactly what the reason is for many a -GoE. One has to guess. Was it that step out of the landing on the 3Flip or was it that toe-pick landing on the 4Toe. All elements have definitions to show perfection.

Examples of Akiko above, do not show uniformity in judging; and definitely different Panels of Judges are never the same. This is extremely unfair to contestants.

Right, but Akiko didn't show uniformity in skating.

Joesitz said:

...we don't know exactly what the reason is for many a -GoE. One has to guess. Was it that step out of the landing on the 3Flip or was it that toe-pick landing on the 4Toe.

Click to expand...

Actually, I feel like I am getting better and better at making these guesses. If a skater does a 3Filp and it looks OK except for a step-out on the landing, and then I see in the protocols that all the judges gave -1 GOE on that flip, then I can pat myself on the back that I saw the same thing the judges saw.

It is a little harder when five of the judging give -1 to that flip and four judges give 0 GOE. This means that some of the judges saw something extra special in that particular jump (maybe superior height and distance and excellent air position?) that balanced out the error and brought the overall GOE on that jump up to 0 on that judge's score card.

Fortunately we can see how each judge scored each individual element.

Mathman said:

It is a little harder when five of the judging give -1 to that flip and four judges give 0 GOE. This means that some of the judges saw something extra special in that particular jump (maybe superior height and distance and excellent air position?) that balanced out the error and brought the overall GOE on that jump up to 0 on that judge's score card.

Click to expand...

I agree here. But what happens next? All judges who give -1 for downgraded flip , and other judges who see "something unique" about the flip (or finding ways possible to find something unique, don't care what it is, a little more height or distance than a another competitor's clean flip) give a +0 or +1 (which may be equal to what a clean competitor got) It is not a fair evaluation in my humble opinion. Too many of the variation of GOEs by individual judges, when judges are asked to defend their mark they find any way to justify it.

Mainly , I think the ISU should do a educational reform for all the judges. Some times, judges mark a downgrade as -negative GOE but others don't because they already assume its taken into account by the technical panel. There has to be a consistentcy and transparency in Technical panels and judging panels.

What kind of error are we talking about here?

A step out from a landing is supposed to get a -2 reduction in the GOE. So even if the jump was special enough in other ways to deserve +GOE before the reduction,

I don't think you'll find an example of a jump with a step out that gets 0 GOE from four judges. Maybe from one or two judges who happened to look down to record the mark as soon as the jump was landed and missed the step out.

This would be more likely with an easier-to-miss error such as a wrong edge takeoff or landing, light touchdown with the free foot, or minor underrotation.

Also, the jump just might not be very good, without outright errors. E.g., scratchy landing or ugly air position. In that case, I could certainly imagine judges disagreeing about whether to give 0 or -1.

If we're talking about downgraded jumps, then even if it looked good enough in real time for some judges to give 0 or +1, not only will the -GOE from the majority of judges likely take the GOE lower than what a clean competitor earned, but the base value of the jump will also be lowered.

"Downgraded" means that a triple is scored with the base mark of a double jump, so even without negative GOE the total score for the element will be significantly less than that for a clean triple.

This year, downgrading for a the lower-rotation jump, marked with << on the protocols and on the judges' screens, is only for jumps that are severely underrotated -- 180 degrees or more. It's extremely unlikely that 4 judges would miss that error and give +GOE.

New this year is the "underrotated" designation, marked with <, for jumps underrotated by more than 90 but less than 180 degrees. That lowers the base mark to 70% of its full value. It's possible that some judges not see that error or might think there were enough good qualities of the jump to balance out the GOE reduction. But the skater would still lose points in the base mark -- the element will still be worth less than a fully rotated triple.

Occasionally judges will make mistakes. But can you really find examples where there's a significant error, obvious error, like a step out or downgrade and almost half the judges give scores better than -1?

On jump combinations it's a little more likely, if the first jump is really good and the second jump is good in most of its phases, or vice versa.

chillyfranz said:

I agree here. But what happens next?

Click to expand...

Well, what happens next is that they take the average of all the judges' scores. If 5 judges give -1 GOE and four judges give 0 GOE, then after throwing out highest and lowest that averages out to -0.57 GOE for the element.

This is not alarming to me, any more than it was alarming under 6ordinal judging for one judge to give a 5.6 and another judge give a 5.7.

Joesitz has offered the example of competitive diving. The judges score each dive from 0 to 10, there is a mark for difficulty, then the to marks are combined (by multiplying, in the case of diving.) I look at the element scores in figure skating the same way. The judges grade the quality of each element from -3 (terrible) to +3 (outstanding). This is added to the "difficulty score" (base value) for the total score for the element.

I believe this was the original idea of the CoP.

However, what happened, I think, was that the judges were not always in agreement as to what constitutes excellent quality. In particular, judges were letting skaters get away with bad mistakes without penalty. So little by little the ISU has whittled away at the autonomy of the judges, writing increasingly detailed rules about just what faults should result in negative GOEs and just what features should count on the positive side of the scale.

Too many of the variation of GOEs by individual judges, when judges are asked to defend their mark they find any way to justify it.

Click to expand...

At first when I read this, I thought that to minimize the mistake of the judges who might have given +1 instead of -1 GOE even by accident, a good thing would be the CoP software to give the judge a restricted GOE option, once the element is called, also his execution circumstances would be called. Like if it is an obvious step out, travel spin etc, and the judge might have missed it, someone would make sure that it didnt matter if judge was looking elsewhere. Someone would call not only the element but also the range of Goe this element could get, so the software depending on the error of the element could give them the option of 0, -1 etc Goe but not the option of +3 (for example), like they can pick from narrowed down options the GOE depending on the element execution. Or the oposite in a great element the judge wouldn't put by accident 0, or -1.

But then I thought the one who would call the quality or the error of the element to narrow down the GOE options is also open to human error, and it would be a plus time consumption, and it is better to have 9 judges with one mistake possibility than one caller with the same error possibility(like what happens now with UR and edge calls), cause at the end judges would end up just pushing buttons.

^ I agree heartily with your second paragraph.

Basically the tech panel make verifiable yes or no calls, with video aid if necessary, though video use is specifically limited by ISU rules. The judges make human judgment on quality, with guidelines and bullet points. Having 9 of them make individual errors or biases less influential on the eventual marks .

Judges are highly trained and I don't see them as inherently evil and/or incompetent. They don't need to be robots in straight jacket just because many people think their own judgement is the absolute truth.

I don't think figure skating as a sport should be a popularity based reality show type competition.

As for

Too many of the variation of GOEs by individual judges, when judges are asked to defend their mark they find any way to justify it.

Click to expand...

just because a statement gets written and quoted, it doesn't automatically become a fact .

SkateFiguring said:

just because a statement gets written and quoted, it doesn't automatically become a fact .

Click to expand...

I think for FS singles, there should be understanding for individual judges on how to award a goe vs. what the technical panel has punished. If some judge score an error as -2 and some score as +1 on the same element of the same skater there is a problem here. It might mean the judge who gave +1 did because he/she thought the technical pannel already punished the skater for the error whereas the judge who score -2 on the same element did not take into account the technical panel who already punished the skater for the erorr in the base value. I know all judges have the opinions, but a -2 GOEand a +1 on the same element is something strange. I don't understand your post

chillyfranz, what I have issue with are

1. your originally stated that

Too many of the variation of GOEs by individual judges, when judges are asked to defend their mark they find any way to justify it.

Click to expand...

There are two claims here that weren't backed up, especially that judges find any way to justify and defend their marks, because I really don't see how you could know this as a fact. I am open to you demonstating these to be true but they have not been established yet to me.

2. Senoirita then proceeded with this statement as a matter of fact to formulate thoughts and resolutions. No logic or solutions are ever valid based on a false premise, or one that is still questionable in need of supporting facts.

As to the possibility of a double penalty with a tech call, I believe one of our very knowledgeable posters here have explained the procedure how the tech panel makes a firm call and how they immediately flag one to be verified. Judges are not kept in the dark.

Overall, I think the ISU officials are not particularly stupid so as to have missed most of what fans would have thought of and have concerns with. What would occur easily to a fan would have easily occured to them already. Sure sometimes rules have to be changed and tweaked after real life application. Unintended consequences have to be dealt with and new intended consequences are to be worked into the rules. It is a on going process that the IJS allows.

Judges are human, so a varietion can happen either because they do have different values or even because one misses or sees something other don't. That's why there are guidelines and there is a panel of 9 from different countries.

Judging - not perfect, nor is it hopelessly stupid and flawed in every way. When such base asumption is used as a premise of all discussions, it's hard to come up with useful and constructive contributions to improve the system, assuming that is the objective.

chillyfranz said:

I think for FS singles, there should be understanding for individual judges on how to award a goe vs. what the technical panel has punished. If some judge score an error as -2 and some score as +1 on the same element of the same skater there is a problem here. It might mean the judge who gave +1 did because he/she thought the technical pannel already punished the skater for the error whereas the judge who score -2 on the same element did not take into account the technical panel who already punished the skater for the erorr in the base value.

Click to expand...

I don't think it works like that. The technical specialist can't really punish anyone. If the tech sepcialist gives an "e" for wrong edge, then the judges are supposed to take that into account (to the tune of -2 GOE) in their scores. In the case of an under-rotation call, the base value of a jump is reduce AND the judges are expected to give negative GOE on top of that. The judges would never be in a position to say, oh dear, this skater has already been punished enough, I will give him a break with a couple of extra GOE points.

For the kinds of errors that this thread is about -- two-footed landing, step-out of landing, etc., the tech specialist has no role to play and each judge must score what he sees.

Then again...In the short program at the Grand Prix Final, one judge gave Patrick Chan's 4T< a +2 GOE. The others gave mostly -2's and -3's. A spread of -3 to +2 on the same element is a little strange. Maybe the +2 was a key-punch error. In any case it would have been thrown out in the trimming of the mean. (We do not know the nationalities of the individual judges. )

SkateFiguring said:

2. Senoirita then proceeded with this statement as a matter of fact to formulate thoughts and resolutions. No logic or solutions are ever valid based on a false premise, or one that is still questionable in need of supporting facts

Click to expand...

actually i didnt take the comment as a fact, I just quoted it (basically the first part of the sentence) to connect to what i wanted to say, that in case of big variations in Goe marks (for example like the Chan example above) what would have been better at first thought-to reduce the choices of a judge and spread the decisions to other people- it is not a good idea afterall.

I believe we have to take into account that the judges are human, and human error will happen.

I believe also that the number of judges' individual GoEs used in a competition would add up to be a proper consensus for the element. It would then follow that all the consensus scores when added together to the base values would give the grand total of the Tech for each contestant.

Still I feel that the GoEs are a matter of opinion as are the Levels given since the guidelines while notworthy are still bogdowned with the human element.

Thanks MM for bringing up my suggestion to use the scale of 0-10 for each element multiplied by its base value. It uncomplicates the scoring system for judging and fans and it also does away with the Tech Panel. judges for Senior Competitions are quite aware of errors in elements. I think we woudl have a better full Technical Result which imo, was quantifiable with little question.

Mathman said:

Then again...In the short program at the Grand Prix Final, one judge gave Patrick Chan's 4T< a +2 GOE. The others gave mostly -2's and -3's. A spread of -3 to +2 on the same element is a little strange. Maybe the +2 was a key-punch error. In any case it would have been thrown out in the trimming of the mean. (We do not know the nationalities of the individual judges. )

Click to expand...

I never looked at these sets of marks before but this sounded curious. Indeed the GOEs for Chan's 4T are all over the place. There is a big spread even if the +2 (which I'm inclined to think of as a punching error) is taken out - from 0 to -3. There was some controversy about this UR call. Patrick himself seemed unconvinced and his coach said there were discussions about it, presumably among the officials, and even fans had arguments over it in some forums. So maybe the diverse opinions among the judges got reflated in the GOEs given, guidelines notwithstanding (just because I'm not 100% sure, and too lazy to look up right now, what the guidelines say about this. )

OTOH, Verner's and Amodio's popped 3A not only got downgraded to 1A but received -3 across the board as well. OUCH! Just goes to show how severely punished popped jumps are.

And I mentioned Oda's 2A fall in the LP before. He landed it well and then fell for whatever reason. The big difference maker is that the fall was attributed and attatched to the jump and he got dinged with -3s. If somehow the fall could be deemed separate from the jump element, he would have received only the obligatory -1 in the total score. The scores were correct if the fall was a technical error on Oda's jump. But if it were caused by a rut on the ice, for example, then he was extremely unlucky that it not only happened but happened right after a jump! OUCH!

OK OK. I realize I'm OT now for this thread. Consider it a minor thread drift please.

Joesitz said:

...it also does away with the Tech Panel. Judges for Senior Competitions are quite aware of errors in elements.

Click to expand...

I think the tech panel would still have a role to play in identifying the elements and in assigning level of difficulty to spins and step sequences. It would be quite embarrassing for one judge to say, that was a pretty good triple Salchow; I'll give it a +1 GOE. Then the next judge says, no, that was triple loop.

Of course a qualified judge knows the difference between a loop and a Salchow. But in a given competition a judge might be required to identify and evaluate 500 or more individual elements. A person cannot maintain lazer concentration hour after hour, and never miss one.

As for the levels, here I think the rules for getting a certain level of difficulty are spelled out pretty objectively. To get a "4" in level of difficulty on a spin you have to change edges so many times, present so many different body position, so many changes of direction, so many revolutions on each foot, etc. It is the job of the tech specialist to count these out. The judges are looking at other things, like centering, speed, extension, etc.