Proposed tech requirement changes (ISU Congress June 2024)

[Karen-W]

@peibeck - my understanding is the proposal is to keep the SBS solo jumping pass in the SP and remove it from the FS while keeping the SBS combo jumping pass in the FS.

 

[kwanfan1818]

A bit like splitting the spins between the two programs. Here they'd split solo and combo, but without alternating them.

 

[Karen-W]

What I would like to see:

1) the Base Values for leveled spins and step sequences increased to incentivize skaters to put more effort into them, at the same time...

2) drop one jumping pass and one leveled spin from the FS and add in two more unleveled choreo elements in their place...

3) multiple unleveled choreo elements similar to dance with lots of options for the skaters and choreographers to select from - choreo spiral sequence, choreo spin (one position - layback, scratch, sit, camel, etc), choreo leap sequence (stag leap, falling leaf, Russian splits, aerials/twists, etc), choreo edge sequence (spread eagle, Ina Bauer, hydroblading, etc), choreo dance lift in pairs, choreo character sequence

4) bonus for landing at least one each of the following jumps: Axel (quad, triple or double), Lutz (quad or triple), Flip (quad or triple), Loop (quad or triple), Salchow (quad or triple) and Toe-Loop (quad or triple)

All of that would give skaters greater variety in their programs and dis-incentivize some of the problems others have pointed out could happen with the elimination of one jumping pass.

 

[gkelly]

This is for tech elements, but I do want to bring up scoring as well in this thread. Seeing someone do a huge 3A and then falling on a 3S< in combo but getting -4.00 GOE (-5 factored on the 3A's BV) makes no sense to me. It needs to be deducted on the 3S<'s BV. The 3A was fantastic. That wasn't the problem.

I understand your point, but it would be tricky to come up with a rule for scoring individual jumps in combinations/sequences that would apply appropriately in all cases. For example, often the first jump in a combination is landed with enough good qualities to deserve positive GOE if the skater had just stopped there and not done another jump, but the reason that second jump failed was because of some loss of speed or loss of balance on that first landing . . . or just getting too close to the boards. So how would you reflect that appropriately in those cases, differently from cases where the first jump was perfect and the second jump failed for some other reason, e.g., bad ice? Would the technical panel have to decide which jump the GOE reduction should be taken from based on their interpretation of why it went wrong?

The complexity of covering all cases would be similar to the complexity of trying to reflect the difficulty of specific combinations based on both (all three) jumps done in the combination and in which order. Under 6.0, judges could think to themselves that certain combinations were especially difficult and reward accordingly, but under IJS the base values are just added together. We've often had fan discussions about how combination difficulty should be reflected better in the scoring, but implementing such granular scoring would not be as simple as it might seem.

Not quite to the point, but in the example you mention above, if the skater fell on a 3S< then that jump alone racked up 7 or 8 points of negative GOE (for the underrotation and also the fall), but the maximum reduction is -5. However, if the 3S had been fully rotated before the fall and the 3A had had several positive qualities, the combination, scored as a unit, could have started at +2 for the first jump with just -5 taken off for the fall, to end up at -3. In which case the value of the reduction is higher based on the base value of the 3A, but the actual reduction (-3 instead of -5) benefits by being connected to the quality of the first jump.

 

[MacMadame]

And I would score the sbs jumps separately for each partner. And definitely keep them in pairs. They are the most exciting part including the throws IMO.

Like they do in dance with footwork. I like that. We'd see 2SM3SW instead of 2S.

4) bonus for landing the following jumps: Axel (quad, triple or double), Lutz (quad or triple), Flip (quad or triple), Loop (quad or triple), Salchow (quad or triple) and Toe-Loop (quad or triple)

I read this a few times thinking "That's all the jumps why get a bonus for doing one of them" but I realized you mean for doing one each of them in the FS (vs. repeating some of them).

I would be in favor of this but only if they make the Lutz and Flip the same jump. Like they did with the toe loop and the toe wally ages ago.

 

[Karen-W]

I read this a few times thinking "That's all the jumps why get a bonus for doing one of them" but I realized you mean for doing one each of them in the FS (vs. repeating some of them).

I would be in favor of this but only if they make the Lutz and Flip the same jump. Like they did with the toe loop and the toe wally ages ago.

I edited to say "one each of the following jumps."

I'm torn on making the Lutz and Flip the same jump. They aren't and I'd like to believe that the technology is there to make that determination - same with the technology to take the rotation assessments out of the hands of the tech panel. But if the sport isn't willing to make the investment in having that technology available at all competitions then perhaps it would be better to just consider them the same jump for judging purposes.

 

[skatingguy]

Tennis has had electronic calling of some sort for quite a while, but since the pandemic they are moving towards no lines people, and all automatic calling. Quite simply the sport now is too fast for the human eye, and numerous errors cause delays in matches. The system they're using costs $100,000 dollars to install for each court to create a 3D computer generated image of the path of the ball. I don't think skating has the money to this for calls that, while important, are only one portion of the scoring system.

How Sony's Hawk-Eye electronic line-calling system transformed the U.S. Open

CNBC got a behind-the-scenes look at Sony's Hawk-Eye line-calling system to understand how the tech works in tennis and other major sports.

www.cnbc.com

 

[Karen-W]

Tennis has had electronic calling of some sort for quite a while, but since the pandemic they are moving towards no lines people, and all automatic calling. Quite simply the sport now is too fast for the human eye, and numerous errors cause delays in matches. The system they're using costs $100,000 dollars to install for each court to create a 3D computer generated image of the path of the ball. I don't think skating has the money to this for calls that, while important, are only one portion of the scoring system.

How Sony's Hawk-Eye electronic line-calling system transformed the U.S. Open

CNBC got a behind-the-scenes look at Sony's Hawk-Eye line-calling system to understand how the tech works in tennis and other major sports.

www.cnbc.com

I don't know that skating has that kind of money but it seems to me that it's something the ISU should at least have for major competitions like their sponsored series' - GP, JGP, Challengers - and the ISU Championships.

 

[MacMadame]

I'm torn on making the Lutz and Flip the same jump. They aren't and I'd like to believe that the technology is there to make that determination - same with the technology to take the rotation assessments out of the hands of the tech panel. But if the sport isn't willing to make the investment in having that technology available at all competitions then perhaps it would be better to just consider them the same jump for judging purposes.

The toe loop and the toe wally aren't the same jumps either. And the ISU has already made those two jumps have the same values.

I think avoiding someone doing 4 jumps all on the same edge -- which the system encourages since only the axel has more points -- would make for better programs overall.

Tennis has had electronic calling of some sort for quite a while, but since the pandemic they are moving towards no lines people, and all automatic calling. Quite simply the sport now is too fast for the human eye, and numerous errors cause delays in matches. The system they're using costs $100,000 dollars to install for each court to create a 3D computer generated image of the path of the ball. I don't think skating has the money to this for calls that, while important, are only one portion of the scoring system.

How Sony's Hawk-Eye electronic line-calling system transformed the U.S. Open

CNBC got a behind-the-scenes look at Sony's Hawk-Eye line-calling system to understand how the tech works in tennis and other major sports.

www.cnbc.com

We don't need all that though. We just need a few more cameras from other angles. Having more cameras would benefit the calling of step sequences too.

 

[skatingguy]

We don't need all that though. We just need a few more cameras from other angles. Having more cameras would benefit the calling of step sequences too.

If the calls for rotations, and edge calls were going to be taken out of the hands of the technical panel they would, but what you're mentioning would be a good first step.

 

[MacMadame]

If the calls for rotations, and edge calls were going to be taken out of the hands of the technical panel they would, but what you're mentioning would be a good first step.

Why would the calls taken out of the hands of the tech panel make a difference in how many angles the reviewers could look at?

 

[skatingguy]

Why would the calls taken out of the hands of the tech panel make a difference in how many angles the reviewers could look at?

If money wasn't an issue than the ISU could implement an electronic system that would allow the automatic calling of edges, and rotation, and take it away from the technical panel as @Karen-W mentioned. That's why posted the link about the tennis system, and the cost because it would seem to be cost prohibitive for the ISU.

 

[MacMadame]

If money wasn't an issue than the ISU could implement an electronic system that would allow the automatic calling of edges, and rotation, and take it away from the technical panel as @Karen-W mentioned. That's why posted the link about the tennis system, and the cost because it would seem to be cost prohibitive for the ISU.

Okay, so you are advocating for that extremely expensive system (that doesn't exist yet at least for skating) vs only having one camera. That makes more sense. I thought you were saying adding cameras would work only if they let the judges call those things and not the technical panel. Which makes no sense so I don't know why I read it that way.

 

[Theatregirl1122]

Personally, I hate the idea of calling jumps for pairs separately. I don't think it rewards or encourages anything that I would think should be rewarded in pairs.

 

[kwanfan1818]

Okay, so you are advocating for that extremely expensive system (that doesn't exist yet at least for skating)

Craig Buntin was already working on this soon after he left skating and got an MBA. (Or maybe it was his MBA project.)

https://www.sportlogiq.com/

According to the site, sportlogiq's tools are video based and used for a handful or sports, including youth sports. I believe his original plan was for putting sensors on skates, but it doesn't look like they're doing it for hockey.

So you'd have someone who knows skating and who is working on AI-based analytics from video.

 

[Foolhardy Ham Lint]

The only way I want to see a jump repeated a 3rd time is if there is a significant and visible variation on one of the 3. For example, put the flip and lutz in the same box, but give a skater credit for a 3rd version if one of the versions has a long BOE glide with a lutz takeoff.

Or combining the element with a split jump, stag jump, Ina Bauer or spread eagle.

At this rate, competition is going to look like 80s skating before the Zayak Rule was implemented.

 

[MacMadame]

Craig Buntin was already working on this soon after he left skating and got an MBA. (Or maybe it was his MBA project.)

https://www.sportlogiq.com/

According to the site, sportlogiq's tools are video based and used for a handful or sports, including youth sports. I believe his original plan was for putting sensors on skates, but it doesn't look like they're doing it for hockey.

So you'd have someone who knows skating and who is working on AI-based analytics from video.

It is not being used to call figure skating edges and rotations right now. Note that I didn't say it wasn't possible. I said it doesn't exist yet. Which it doesn't.

 

[kwanfan1818]

It is not being used to call figure skating edges and rotations right now. Note that I didn't say it wasn't possible. I said it doesn't exist yet. Which it doesn't.

It might not be used right now, but not only is the video technology available, Buntin had the sensor-based technology that, IIRC, could detect edges and revolutions, developed to at least a prototype. It doesn't have to be developed from scratch, and the video-based use makes it more than a one-off.

This is not impossible, and, at the championship level, is portable and unlikely to be wholly unaffordable if youth leagues are using it.

 

[On My Own]

I understand your point, but it would be tricky to come up with a rule for scoring individual jumps in combinations/sequences that would apply appropriately in all cases.

I would say the judges should be able to use their discretion better in the cases you've mentioned. Just give them the option to mark the GOE based off the jump they feel has been done best and make deductions based off the jump that was done the worst.


About lutz/flip and toe loop/toe walley - I wouldn't say they're the same case at all. And I wouldn't want flip and lutz merged just because some can't do both of them properly.

 

[carriecmu0503]

I can 100% have it both ways

If a combo is well done overall, then it makes perfect sense to score it based off the highest BV element in the combo.

If you fall on part of the combo or make some other error, deduct GOE from that part of the combo.

Once again, combos are graded as an entire element, based off the total value of the element. Period.

 

[On My Own]

Once again, combos are graded as an entire element, based off the total value of the element. Period.

I'm very curious about the quality of the points you wish to make on this forum. Surely, someone who has a PhD would be able to tell that I mean that I hope to see elements graded differently in the future?

 

[carriecmu0503]

I'm very curious about the quality of the points you wish to make on this forum. Surely, someone who has a PhD would be able to tell that I mean that I hope to see elements graded differently in the future?

The point is that you clearly do not understand the judging system, and that it is already hard enough to judge elements consistently without getting all willy nilly with the rules. GOE is assigned to elements based on the total value of the element. It is very simple, and coaches and skaters understand this very well.

 

[On My Own]

The point is that you clearly do not understand the judging system,

Sweetheart, I understand the judging system, and I imagine much else of the world, better than you ever will.

I hope we're done with this now? I need to go back to doing research for my own PhD.

 

[Sk8swan]

For the pairs I do agree with getting rid of 1 of the sbs jumps in the FS, it's enough for testing the skaters jumping in synch (which imo is far more important than the skaters single jumping ability). Why not getting both the pair and sbs spin instead of having 2 sets of sbs jumps I never understood that. For sure the shorter length of the FS cut to 4 min does not help having both sbs and pair spins and you could argue you see one in the SP and one in the FP but still... 3 solo jumps overall in the pairs programs never made sense to me.
I love pairs lifts (perform by high level pairs ) and I would prefer not to see them cut off in number... Why not give the pairs more freedom to chose the elements they want to perform in the FP? Within a range obviously. I think it will help see more variety in the programs construction and make the discipline more interesting to watch.

 

[Private Citizen]

What exactly would be the point of a combo or sequence if you're going to grade the elements separately? Not that many judges do this, but the GOE of the combo or sequence should also reflect what goes on between the jumps -- e.g., edge pulls (Eteri students), obvious reset for the second jump, long pause, standstill take-off, hopped euler, etc.

I agree that it would be silly to judge pair jumps separately, too. It may even create weird incentives to do mismatched jumps -- e.g., for partner 1 to deliberately do a triple (or quad) while partner 2 does a double because they have no ability to rotate more. I like the idea of a Gordeeva and Grinkov (or Marina Kielmann)-style jump sequence of up to five single and double jumps, with a capped base value, maybe 5.00. Perhaps the GOE on this element could be set very high, so that pairs could get an additional 5.00 for a +5 and lose everything for a disaster.

Adding to sk8swan's comments, I would also favor a system that aligns base values of different types of elements so pairs can choose what type of elements to perform. Anything that encourages variety is good with me.

 

[On My Own]

What exactly would be the point of a combo or sequence if you're going to grade the elements separately?

I did not say "separately". I said if someone falls on 3A+1Eu+3S<, then deduction should be on 3S< since they fell on the 3S being too underrotated. That is one raw GOE for one combo, scaled according to 3S BV.

If someone feels a 3Lz was landed too close to the boards which then messed up the rest of the combo as gkelly said, they can deduct raw GOE on the 3Lz. That is also one raw GOE for one combo, scaled according to 3Lz BV. This is letting it go to the judges own discretion.

 

[zebobes]

I did not say "separately". I said if someone falls on 3A+1Eu+3S<, then deduction should be on 3S< since they fell on the 3S being too underrotated. That is one raw GOE for one combo, scaled according to 3S BV.

If someone feels a 3Lz was landed too close to the boards which then messed up the rest of the combo as gkelly said, they can deduct raw GOE on the 3Lz. That is also one raw GOE for one combo, scaled according to 3Lz BV. This is letting it go to the judges own discretion.

I understand your goal, but that could lead to people who have tacked on a single toe and falling, to getting away with having less deductions because the fell on the single toe.

 

[On My Own]

I understand your goal, but that could lead to people who have tacked on a single toe and falling, to getting away with having less deductions because the fell on the single toe.

It might need more adjustment, but it also will help when we need to deal out positive GOE. As an example, Yulia Lipnitskaya's 3Lz+3T and 2A+3T combos. The preceding jump was always small, but on the other hand she could do a fantastic +3T. Instead of going +1 and scaling by 3Lz in that first case, it would make more sense to scale by 3T's base value.

 

[gkelly]

There is no mechanism for judges to affect which base value to use for the GOE reductions.

You would need to completely redesign the scoring system to allow for judges to make that decision and input it into the system.

Also, remember that <, <<, and e calls change the base value of an element and therefore also the value of the GOEs. As well as being in the bonus period, but that affects all jumps in the combination.

(Here is the current Scale of Values. See how complicated it already is for jumps. https://www.isu.org/figure-skating/rules/fsk-communications/28337-isu-communication-2475/file. We can't expect judges, or even tech panels, to memorize all the possible base values.)

How would your system work for judges to indicate to the computer which base value to apply to their positive or negative GOEs, falls aside?

What about elements with errors on both jumps?

Can judges change their mind about which jump's GOE values to apply after they see a tech panel < or << call?

E.g., in your example, 3A + 1Eu + 3S<F, maybe a judge thinks "The 3A and the euler looked fine but he fell on the 3S. Therefore he should be penalized -5 for the value of a 3S," which the judge may or may not remember is -2.15 on a 3S base value of 4.3, which they think is more relevant than the -4.00 reduction if scaled to the 3A value.

But then during the reviews they see (by their own rewatching and/or by the tech panel calls) that the 3A that looked good in real time was really 3Aq (which doesn't change the base value of the axel but does affect their opinion that the axel deserves no deduction).

Can they now decide that because there were multiple errors on the element affecting two or all three of the jumps in the combo, that they do want to use the GOE value of the higher value element after all?

(And in this case, it's also likely that there were problems with the euler that led to the fall. For relative simplicity, to make your case you'd be better off choosing an example with 3T at the end so there are only two jumps involved.)

So let's say there's a 3F+3T combination with a fall on the 3T, but both looked rotated in real time. So the judge chooses to go with the reduction value on the 3T because that's the jump the skater fell on and why penalize them more for the value of the 3F?

And then in review it turns out that the 3F actually gets both a < and an e call. So now there are multiple errors to penalize even though the maximum reduction is -5, and there's more reason to believe that problems with the 3F landing affected the skater's ability to land the 3T. More to your point, though, now the base value of the 3Fe< is now less than the value of the rotated 3T. Should judges take that into account when deciding which base jump's base value to apply to the reduction?

How would your system handle all these complexities? And does the judge need to do the math in their head about which jump has the highest base value if it might not be obvious?

 

[rfisher]

With the new removal of the SBS jump in pairs, we are well on our way to the merger of dance and pairs into one discipline!

• Dance lifts have been relaxed to allow lifts above the shoulders.
• Spins in dance & pairs are now indistinguishable.
• Jumps are being introduced into dance with the assisted jump.
• Compulsory dances are all but history.
• Deep edges, pointed toes, and neat feet are no longer rewarded in dance.

The edges are really starting to blur between the 2 disciplines!

No pairs would mark the end of my skating fandom.