2024 GP France: Women's Thoughts?

[lariko]

I think that is where the problem lies. It is absurd to think that anyone can tell whether the jump is "exactly" on the quarter, using the ordinary common-sense definition of what "exactly" means.

If the rule were something like, 0 to 45 degrees, no call; 45 to 90 (q), reduce DOE by -1; 90 to 180 (<),, reduce GOE by -2, 180 to 360 (<<), downgrade -- at least that would have the advantage of being well-defined, although it would still be up to the judgment of the tech panel.

Still, overall I am satisfied with the current conventions, compared to earlier attemps to incorporate these considerations into the IJS. (Especially the part about, all jumps except the Lutz are automatically under-rotated because of the mechanics of the jump.)

No matter what number of degrees it is, I don't think it's something actually measurable while the rule gives a false impression that tech panel actually has a number to go on like from a map with a protractor or a digital equivalent

 

[Mathman]

No matter what number of degrees it is, I don't think it's something actually measurable while the rule gives a false impression that tech panel actually has a number to go on like from a map with a protractor or a digital equivalent

I am not so sure about that. If I were a tech caller I would throw up my hands in despair if I were required to assert that the jump was exactly 45 degrees, not 44 or 46.

But if my task is to judge whether the jump flls in the range 45 to 90, I would have a lot more confidence in making that detrmination. I would expect to be "clearly correct" most of the time, although there will still be borderline cases were different adjudicators see things differently.

 

[lariko]

I am not so sure about that. If I were a tech caller I would throw up my hands in despair if I were required to assert that the jump was exactly 45 degrees, not 44 or 46.

But if my task is to judge whether the jump flls in the range 45 to 90, I would have a lot more confidence in making that detrmination. I would expect to be "clearly correct" most of the time, although there will still be borderline cases were different adjudicators see things differently.

In my profession, parts per million above a guidleine is above guideline and must be addressed. Qualitative judgment is "looks rotated or looks slightly under". But when you state it hits between 45 and 90 it becomes a quantitiave statement that means that the jump had been landed in 46 to 89 degrees range and there is no actual measurement attached to it that says that, indeed, it was 52 degrees so correctly ended up as a q. It is no different than saying that the jump was landed at 90 degrees from a fully rotated position. Because you still have to tell the difference between 45 and 46 degrees. The funny thing, theoretically the number of skaters called if it was 45 to 90 versus exactly 90 is not going to change, but it should, since theoretically pretty much everyone who doesn't go over underrotates by an unknown number of degrees.

So, we don't measure, but we give an impression that we do, then we penalize without any proof.

 

[lariko]

Anyway, I am watching Kagiyama land ugly/very pronounced step-outs witn -2 or less GoE and the penalty for basically invisible q at -2 GOE minimum feels egregious.

Also, it looks like the marks judges put into the table are prorated because for many TES, final GOE is not a simple average of judges' GoE marks anyway.

 

[Mathman]

I think that we need to be careful when it comes to measuring. If we possessed really, really, really accurate measuring instruments (parts in a million), figure skating might end up being embarrassed when it was "discivered" that, for instance, women's "quads" actually achieve barely more that three complete revolutions in the air (1080 degrees). The ISJ would have to embark on a public relations push to explain that, yes of course akate jumps to not actually have the amount of in-air revolutions that their name implies ("triples," etc.) That's nbecause of, um. well, um -- OK, you know nothing about skating technique if you don't understand that.

Unfortunately, the alternative to measuring is to give those lyin', cheatin' judges free rein to do whatever they want to achieve their nefarious ore-determined and politically motivated.outcome.

The ISU is between a rock and a hard place -- if not between the Devil and the deep blue sea.

 

[Mathman]

Also, it looks like the marks judges put into the table are prorated because for many TES, final GOE is not a simple average of judges' GoE marks anyway.

Yes. +5 gets 50% of base value, +4 gets 40%, etc.

Th Scale of Values document (ISU documant 2656) spells it out and makes the calculation for us in a couple of hundred lines covering every possible jump from 1T,<< to 5Lz, plus for every possible level of every possible spin and every possible GOE.

 

[gkelly]

They can tell if it is 90 degrees under specifcially? How?

If you had to estimate an angle that someone showed you, whether with lines on paper or a circular arc on paper or something in the real world, the easiest numbers to assess by eyeballing would be 360, 180, and 90.

We know what a right angle looks like. We don't have as much of a sense of what intermediate amounts look like.

Still, I would have no objection if the ISU decided to get rid of the q call and go back to the simple << vs. < distinction, with no call if the tech panel decides the underrotation may have been exactly 90 degrees but was not more than 90.

Also, I swear, I saw q'd jumps have running score of over 2, but they went way under after q was called.

Yes, the judges can only go by what they see with their own eyes in real time while the program is in progress.
The tech panel reviews happen after the end of the program. The judges also have an option to review video of selected elements after the program.
So any reductions for q or ! or even < calls are likely not to be applied during the performance, unless the error was so obvious that there was an open possibility in the judge's mind that the jump might even deserve << or e.

Also, if you're talking about the numbers in the scoring box onscreen during the performance, remember that the value of the GOE that shows there is the prorated value based on the base value of the element. Each + is worth one-tenth of the element's base value, and each - is worth one-tenth of the base value.

There is no element that has a base value of exactly 10.0 points, so even if all judges agreed that the element deserved the same GOE, let's say +2, the averaged positive GOE that you see in the scoring box would not be 2.00 but rather two tenths of the element base value. If the element is a quad loop, flip, lutz, or axel (base values of 10.5 to 12.5 for these quads with no rotation or edge calls), then all judges awarded +2 would average out to higher than 2.0. For any other element, the GOE will be less than 1.0 for +1 and less than 2.0 for +2, so if the scoring box is showing a GOE higher than 2.0, that would mean that at least a few judges awarded +3 or higher.

So if you see a GOE of higher than 1.00 for a solo double axel (base value 3.3), that means that at least a couple judges awarded at least +4. Because if all judges awarded +3, the final positive GOE would be 0.99. And if only one awarded +4, it would have been thrown out when trimming the high and low scores.

Other relevant asides:
If the jump is in the second half, then the base value gets another 10% bonus, but GOE values are not affected by that bonus. They are affected by reductions of base value for rotation, edge, and +REP calls.

For jump combinations and sequences, the GOE is pegged to the base value of the highest value jump in the combo/sequence.

And -2 sounds terribly high, since it's 2x penalty for a fall.

Not true.

First of all, the fall deduction is exactly 1.00, subtracted from the total score for the whole program, in addition to any GOE penalty for the element itself assuming the fall occurred on an element. Most elements with falls end up earning -5 GOE in addition to the fall deduction.
And if there are multiple falls in a program, the deduction for the third and fourth falls is 2.00 off the total score, in addition to any GOE penalty on the elements. IIRC, the fall deduction for a fifth fall, or beyond that, is 3.00, but I'd have to look it up.

Second, the -2 GOE for all elements except for the more difficult quads adds up to less than 2.00. For rotated triple salchow and lower, and for triples up to lutz with underrotation and/or e calls, the base value is less than 5.0, so the value of +2 GOE would be less than 1.00.

Also, it looks like the marks judges put into the table are prorated because for many TES, final GOE is not a simple average of judges' GoE marks anyway.

Yes. See above.

For further details, see the singles/pairs Scale of Values and Rule 353 in the Singles & Pairs/Ice Dance rulebook for principles of calculation.

It's just arithmetic, but quite complicated. Lots of factoring and averaging. Which is why we really can't expect judges to keep track in their heads of what all their scores for one skater will add up to compared to a previous skater's total, especially since they get no information about what levels the tech panel awards for spins and steps (and other leveled elements in other disciplines).

 

[Magill]

The ISU is between a rock and a hard place -- if not between the Devil and the deep blue sea.

No. ISU's choice is between using 19th/20th century measurement tool known in the literature as the "doctor's eye" (the judge's eye, in this case), or slightly more 21st-century-fitting measurement tools / technologies.
For unknown reasons ISU is choosing to stick with these19th/20th century tools while most of the audience has 21st century expectations.
But it was discussed here so many times... sigh

 

[gkelly]

No. ISU's choice is between using 19th/20th century measurement tool known in the literature as the "doctor's eye" (the judge's eye, in this case), or slightly more 21st-century-fitting measurement tools / technologies.
For unknown reasons ISU is choosing to stick with these19th/20th century tools while most of the audience has 21st century expectations.
But it was discussed here so many times... sigh

I have no doubt that sometime in the 21st century there will be technology developed and adopted that can be more precise than the human eye to measure jump rotation.

Maybe by the 2030s, assuming it can be cost effective at least for major events.

But then there still needs to be a decision about how to score each jump once its exact amount of rotation is determined, taking into account both rotation and also every other aspect that could be measured by technology (e.g., height and distance) or only evaluated by the human eye and brain.

Maybe the measurable aspects could be incorporated into the base value somehow. Any suggestions for how that might work?

But other qualities will still be evaluated by humans.

 

[Skating91]

I think that we need to be careful when it comes to measuring. If we possessed really, really, really accurate measuring instruments (parts in a million), figure skating might end up being embarrassed when it was "discivered" that, for instance, women's "quads" actually achieve barely more that three complete revolutions in the air (1080 degrees). The ISJ would have to embark on a public relations push to explain that, yes of course akate jumps to not actually have the amount of in-air revolutions that their name implies ("triples," etc.) That's nbecause of, um. well, um -- OK, you know nothing about skating technique if you don't understand that.

Unfortunately, the alternative to measuring is to give those lyin', cheatin' judges free rein to do whatever they want to achieve their nefarious ore-determined and politically motivated.outcome.

The ISU is between a rock and a hard place -- if not between the Devil and the deep blue sea.

You will find a lot of top women (top 5 in the world) are at times barely rotating 1.75 times jumping a supposed triple (almost 225 degrees prerotation and then almost on the q with the landing but not called depending on which region they hail from). This is especially the case with the second jump on combos. It has gotten out of hand and skaters have taken liberties since the end of the last Olympics.

All this is easy to measure but the ISU officials actively ignore it, and judges marks are almost always consistent with one another as if it's all been discussed in advance that this skater is worth that much, this skater should receive this and so on.

 

[Magill]

I have no doubt that sometime in the 21st century there will be technology developed and adopted that can be more precise than the human eye to measure jump rotation.

Maybe by the 2030s, assuming it can be cost effective at least for major events.

But then there still needs to be a decision about how to score each jump once its exact amount of rotation is determined, taking into account both rotation and also every other aspect that could be measured by technology (e.g., height and distance) or only evaluated by teh human eye and brain.

Maybe the measurable aspects could be incorporated into the base value somehow. Any suggestions for how that might work?

But other qualities will still be evaluated by humans.

Yes, sure. But maybe if ISU stops pretending 21st century never started yet and starts to objectify the scoring process, maybe, just maybe something would start to improve. For right now the 21st century audience is left with an impression that ISU / feds just refuse to use objective instruments to avoid limiting the human judges' control over scores and placements. Delegating even part of this process to machines takes this amount of power from them, however big or small it may be. When you see people discussing all over the internet that there is this blind spot on the rink which skaters use to place their jumps to avoid edge/UR calls, and everyone and their little brother knows of this spot, and still ISU does nothing about it for years, not a slightest move to rectify this situation, for years, you naturally conclude ISU/feds have no problem with this, or even choose to leave this spot for skaters to use.
I'm not even saying this is true or not, I'm just saying this is just a natural conclusion when obvious deficits of the evaluation process are not being addressed properly and TPTB rather than take all the effort to make scoring as fair and objective as possible choose to pretend these problems do not exist, or are just some fan fantasies or obsessions.
And this is just one of many examples.

 

[gkelly]

Well, the sight lines/blind spots will vary from rink to rink depending on the architecture of each. And will also vary for clockwise vs. counterclockwise jumpers.

Adding another camera angle could help. Should be feasible for the most important events held in big arenas. Probably not for all JGPs or even challenger events.

 

[Magill]

Well, the sight lines/blind spots will vary from rink to rink depending on the architecture of each. And will also vary for clockwise vs. Counterclockwise jumpers.

Adding another camera angle could help. Should be feasible for the most important events held in big arenas. Probably not for all JGPs or even challenger events.

Yeah, everyone knows more cameras would do. For years. Still, they are not there and it seems judges are not really vocal about needing them. It seems they are perfectly fine and happy with this one camera and its quality. Still, I hear that what we are shown as slo mo by TV is better quality than what judges and the other members of the evaluation team use for actual judging. How come?
And sorry, I am not prepared to believe it is a question of the lack of money cause this is a kind of equipment many people have at home.... nowadays it's cheap.
IMHO it is most of all the lack of political will to introduce changes, with the energy that could be used to introduce improvements used actually to defend the status quo instead.

 

[Mathman]

About the "doctor's eye." I think that is the whole point. The doctor's eye cannnot say that a point on a circle is 137 degrees from vertical rather than 138 -- nor does anyone care.

The doctor's eye can, however, discern, in most cases with great confidence, that the angle is somewhere between 90 and 180. That is why I would have no problem with categories of under-rotation like this: 0 to 90 degrees = no call by the tech panel, judges give GOEs based on what they saw,

90 t0 180 = tech panel calls <, judges reduce GOE by 1, 180 t0 360 = tech panel calls <<, judges reduce GOE by 2. More than 360 = tech panel calls "d" and the base value is lowered accordingly. I do not think that 21st century robot eye would necessarily bring any major improvement to figure skating, either as sport or as entertainment. Nor would it prevent bad actors in the judging and oversight ranks from acting bad if they are inclined to do so.

 

[Magill]

About the "robot's eye" - I agree you do not need a robot. At this stage, some decent cameras, a computer and a dedicated software would help enormously to make a difference and improve. Objectifying scoring and evaluation, and employing machines instead of corruptible humans to do the job, or at least a part of it, helped many corrupt systems go way cleaner, if not completely clean. See no reason why FS should be any different.
As for bad actors in judging and officialdom, they just need to be removed. Such removals also help to strengthen morale of the remaining lot while doing nothing usually lowers the morale and ethical expectations even of those who are taking the effort to keep their integrity and do their job properly.

 

[lariko]

I think that we need to be careful when it comes to measuring. If we possessed really, really, really accurate measuring instruments (parts in a million), figure skating might end up being embarrassed when it was "discivered" that, for instance, women's "quads" actually achieve barely more that three complete revolutions in the air (1080 degrees). The ISJ would have to embark on a public relations push to explain that, yes of course akate jumps to not actually have the amount of in-air revolutions that their name implies ("triples," etc.) That's nbecause of, um. well, um -- OK, you know nothing about skating technique if you don't understand that.

Unfortunately, the alternative to measuring is to give those lyin', cheatin' judges free rein to do whatever they want to achieve their nefarious ore-determined and politically motivated.outcome.

The ISU is between a rock and a hard place -- if not between the Devil and the deep blue sea.

They will also find it across the board and accurately, so the changes in scoring will be easily explained and consistent from there on. Until this level of a accuracy people can insinuate all they want that women didn’t actually rotate quads, but current women who jump triples actually rotate triples simply because they believe women can jump triples but not quads. The main thing is it will be consistent no matter which skater it is and who sits on tech panel.

With precise measurements the judging can only improve because speed, rotation, momentum, coverage all of it is physical motion. Not to measure it is like saying that we don't need photo finishes, let some volunteer guys who know a lot about running decide who won the race.