If you could change one rule in any discipline

[CanadianSkaterGuy]

It depends. Polina Edmunds used to do a COE Biellmann - that's certainly extremely hard, and needs additional consideration, not less.

Similarly, holding a Biellmann for 8 revolutions is harder than holding a crossfoot upright for 8 revolutions.

You're right - they do need to get granular about what is harder and deserves extra feature/bullets/etc. and what is easier. I get right now though it's for ease that they just make each of those features one increase in level. But yeah - a COE Biellmann is hella scary and tough!

 

[CanadianSkaterGuy]

Make any fall count the element as 0.

I'm inclined to agree for step sequences and spins and death spirals.

But for jumps and twists/lifts, those are high-ticket items and you're going to see risk removed from programs if skaters are worried about falling - no point even trying a quad lutz for 13 points that could be 0 points, when a triple lutz is way easier/safer and bags you 7 points.

 

[4everchan]

I'm inclined to agree for step sequences and spins and death spirals.

But for jumps and twists/lifts, those are high-ticket items and you're going to see risk removed from programs if skaters are worried about falling - no point even trying a quad lutz for 13 points that could be 0 points when a triple lutz is way easier/safe and bags you 7 points.

hey, it's my rule change . I don't agree with your rule changes but I am not raining on your parade . So... let me be .

Okay . just teasing you. How have you been ?

I exactly want people to take risks where it's not a Hail Mary type of risk. I won't name anyone but there have been some people doing quads in programs, even at the Olympics, they have never landed... give them 0.

I respect skaters who aim to perform beautiful program which are clean. That's a huge amount of pressure to undertake. I feel that when skaters go for big BV, they have a cushion and can make mistakes, and do make mistakes and still win... why should they ? Is this sport about what you try or what you succeed to perform ?

 

[gkelly]

I'm inclined to agree for step sequences and spins and death spirals.

But for jumps and twists/lifts, those are high-ticket items and you're going to see risk removed from programs if skaters are worried about falling - no point even trying a quad lutz for 13 points that could be 0 points, when a triple lutz is way easier/safer and bags you 7 points.

A big part of the definition of executing a jump is landing it successfully (with penalties for only partial success). If a skater falls on a solo jump, they have not completed the element. (If they fall on the last jump of a jump combination or sequence, they have completed part of it, so the partial credit is more defensible.)

On the other hand, step sequences, spins, and death spirals (usually in that order) occur over a longer period of time and a fall at the end of the element does not invalidate everything that was successfully executed before the fall to the same degree that a fall on a solo jump landing does. You of course can make your own rule changes, but I personally would oppose a rule that penalizes a 99% or even a 50% successful element that is interrupted by a fall after it has accomplished the minimum requirements for the element.

Falls at the beginning of such an element already result in the element getting no value. Falls in the middle may or may not cancel what was already accomplished, depending how much was actually accomplished before the fall, in relation to the rules for what's required of that element in that program type. Personally, I think this is fairer.

In addition, especially with step sequences, what causes a fall is likely to be either 1) high attempted risk, e.g., in the depth or complexity of edgework of a step sequence or a difficult flying entrance in a flying spin; or 2) (especially for step sequences), a flaw in the ice such as a hole or foreign object most likely left by a previous skater, that is bad luck for the current skater but not a reflection on the quality of their own technique. For different reasons, I don't think either needs to be punished so harshly that all the successful parts of the element the skater accomplished before the fall receive no credit.

 

[TontoK]

Jumping on the zero-points-for-a-fall train.

I'll go further. If the jump does not take off or land on the prescribed edge, then it's really not that jump, is it?

As a flutz is not listed in the code of points, it should be considered, at best, a choreographic flourish.

It's always been one of life's mysteries to me that fans will go to battle over a point or two of PCS, but then dismiss edge calls. Rotation, too, for that matter. I mean even rooting for a "lax controller" so their favorite can "get by."

I always want the controller to be Shin Amano in a really bad mood.

 

[DizzyFrenchie]

Maybe one could remove the 1 point penalty for a fall. It made sense in the -3/+3 system because -30% for a fall wasn't enough, but with the +5/-5 system a completed jump is heavily penalised when followed by a fall, this additional penalty is too much. Of course, I mean if rules were respected, that is, if the jump was underrotated (or prerotated to a downgrade) it must be called before the input of the -50%; I'm not counting the PCS caps though, because it has never been called consistently, Alexander Lakernik even declaring that it was facultative; in my opinion this call should depend on the real disruption of the program, which it has never been.

 

[gkelly]

So then would you keep the fall deduction for falls outside elements, since there is no GOE penalty for those?

 

[DizzyFrenchie]

So then would you keep the fall deduction for falls outside elements, since there is no GOE penalty for those?

What an interesting question! Actually it's the -1 penalty that makes more sense, rather than -5 in Grade of Execution which has such different impact depending on a Skater falling on a 3F or after the rotated 3A of a 4Lz + 3A + SEQ... I had never thought about this solution, but why not a GOE penalty for a fall, like, 3 to 5 points substracted to the GOE after bullets are added up and other calls deducted, and keeping the -1 fall deduction? It would often result in a -5 GOE but not always, wouldn't it be fairer should a jump ending on a fall get more points if there's just a slip after a beautiful jump than when there's an ineluctable fall after a botched jump? This wouldn't suppress the huge difference in sanction depending on the initial base value but it would diminish it a bit?

 

[gkelly]

It's not all about jumps. Or even all about elements.

There are at least two different things being punished in a fall on an element:
a technical failure on the element, and the disruption to the surface "cleanness" of the program as a whole as a result of the fall

In fact, even with the technical aspects of the element, there are falls that are incidental after the skater has completed good difficulty with otherwise good quality, and falls that represent a complete failure of the element such that nothing of value has been completed, and everything in between.
Sometimes the fall/failure ends up resulting in the skater completing a lesser element, such that they end up with fewer points for base value than they intended (for example, falling on the first jump of an intended jump combination, or in the middle of a spin such that they earn no levels for the features they didn't get to complete and maybe they don't get credit for an intended change of foot)
Some elements start out with high base value and some with lower base value. It's for this reason that the GOE penalties are based on a percentage of the base value rather than the same across the board for single jumps and quads, or level base vs. level 4 spins or steps.

I think it's a good thing to have flexibility in the GOE penalties to reward what is done well while also penalizing what was done badly and to scale those penalties to how many points the element starts out with, as actually executed.

Those penalties can be built into the GOEs, related to the base values, such that elements with falls are sufficiently penalized that an additional 1.0 deduction seems unnecessary, which is what I think you're proposing.

But then you have the "disruption to the program" issue. The places this can be penalized are in a separate deduction and in PCS.

There are now caps on maximum PCS with falls, but the caps only apply to skaters who were potentially going to earn scores in the 9s without any falls. The same principle, that falls should be reflected in PCS, applies to skaters starting in a lower PCS, but that's wrapped up in everything else that judges take into account when deciding whether a performance deserves 8s or 5s or 2s, and if so which 0.25 increment within those general ranges. It's not visible or enforceable at a granular level the way fall deductions are.

And then there are the separate fall deductions, which are specific and enforced and are the same regardless of whether the skater fell on something really difficult or fell on something simple like crossovers or their ending pose. (Or something difficult that earns no base value or GOE, such as a difficult transitional move.)

As of last season, in novice competition (advanced novice, intermediate novice, basic novice), the ISU is giving fall deductions for falls outside elements but none for falls inside elements, which I think is what you're asking for at junior and senior levels. That makes sense.

https://isu-d8g8b4b7ece7aphs.a03.azurefd.net/isudamcontainer/CMS/Corporate-Site/Governance/Transparency/ISU-Communications/2699-SP-Novice-Competitions-2025-2026-1744291808-1629.pdf

Note that the deductions at novice levels are 0.5 rather than the 1.0, in accordance that novice scores overall (base values and PCS ranges) tend to be lower to begin with.

 

[Jeanie19]

Jumping on the zero-points-for-a-fall train.

I'll go further. If the jump does not take off or land on the prescribed edge, then it's really not that jump, is it?

As a flutz is not listed in the code of points, it should be considered, at best, a choreographic flourish.

It's always been one of life's mysteries to me that fans will go to battle over a point or two of PCS, but then dismiss edge calls. Rotation, too, for that matter. I mean even rooting for a "lax controller" so their favorite can "get by."

I always want the controller to be Shin Amano in a really bad mood.

Not just the flutz but the lip, too. Many flips are lutzes.

 

[lileychristie]

First thing that pops to mind:

In singles, at least 1 mandatory non-jump choreo move in a choreo sequence and it must be held for at least 4s for it to count.
Not just spirals, but any non-jump choreo moves like ina bauer, spread eagles, hydroblades etc.

I love all those moves, but too often skaters barely hold them for 1 second, too quick to be appreciated fully

I specify "non-jump choreo moves" here cause there are also split jumps, russian split jumps etc and they can't be held for 4s obviously haha

 

[DiStSq4]

Not just the flutz but the lip, too. Many flips are lutzes.

While we're talking about flip and lutz, can we talk about how the textbook flip used by ISU is "slight inside edge"? Whether someone's on a deep inside edge or they're flat, it's not "proper" all the same. I know there's the whole thing about deeper edge prerotates more, but every other jump also need prerotation! Why do we care about this specifically on the flip? Why do we want only a "slight" edge? And in pairs it doesn't even matter whether you're doing a throw flip or a throw lutz, they're valued all the same!

 

[FSFan1999]

While we're talking about flip and lutz, can we talk about how the textbook flip used by ISU is "slight inside edge"? Whether someone's on a deep inside edge or they're flat, it's not "proper" all the same. I know there's the whole thing about deeper edge prerotates more, but every other jump also need prerotation! Why do we care about this specifically on the flip? Why do we want only a "slight" edge? And in pairs it doesn't even matter whether you're doing a throw flip or a throw lutz, they're valued all the same!

Theres no specific rule that edge depth on flip and lutz is important, as long as it is clear. Fans & judges decided what they preferred in terms of jump edge depth.

With lutz, traditionally it jumped from one circle to the other, and generally speaking this "torque" used to create more edge (and flips were always done on relatively straight, square entries). But if it's clear outside it's lutz, clear inside it's flip. I dont know any documents that say more depth is necessary by jump definition.

 

[DiStSq4]

Theres no specific rule that edge depth on flip and lutz is important, as long as it is clear. Fans & judges decided what they preferred in terms of jump edge depth.

With lutz, traditionally it jumped from one circle to the other, and generally speaking this "torque" used to create more edge (and flips were always done on relatively straight, square entries). But if it's clear outside it's lutz, clear inside it's flip. I dont know any documents that say more depth is necessary by jump definition.

I believe there's an old ISU jumping video where it says too deep of an edge is an error of the jump. It's one from the old DVDs, but I don't remember the title.

 

[FSFan1999]

I believe there's an old ISU jumping video where it says too deep of an edge is an error of the jump. It's one from the old DVDs, but I don't remember the title.

if you can find the specific clip let me know. It could be there was another error contributing to it (dropping inside shoulder, edge spiraling deeper instead of staying stable, edge starting flat then going deeper).

 

[DizzyFrenchie]

It's not all about jumps. Or even all about elements.

There are at least two different things being punished in a fall on an element:
a technical failure on the element, and the disruption to the surface "cleanness" of the program as a whole as a result of the fall

In fact, even with the technical aspects of the element, there are falls that are incidental after the skater has completed good difficulty with otherwise good quality, and falls that represent a complete failure of the element such that nothing of value has been completed, and everything in between.
Sometimes the fall/failure ends up resulting in the skater completing a lesser element, such that they end up with fewer points for base value than they intended (for example, falling on the first jump of an intended jump combination, or in the middle of a spin such that they earn no levels for the features they didn't get to complete and maybe they don't get credit for an intended change of foot)
Some elements start out with high base value and some with lower base value. It's for this reason that the GOE penalties are based on a percentage of the base value rather than the same across the board for single jumps and quads, or level base vs. level 4 spins or steps.

I think it's a good thing to have flexibility in the GOE penalties to reward what is done well while also penalizing what was done badly and to scale those penalties to how many points the element starts out with, as actually executed.

Those penalties can be built into the GOEs, related to the base values, such that elements with falls are sufficiently penalized that an additional 1.0 deduction seems unnecessary, which is what I think you're proposing.

But then you have the "disruption to the program" issue. The places this can be penalized are in a separate deduction and in PCS.

There are now caps on maximum PCS with falls, but the caps only apply to skaters who were potentially going to earn scores in the 9s without any falls. The same principle, that falls should be reflected in PCS, applies to skaters starting in a lower PCS, but that's wrapped up in everything else that judges take into account when deciding whether a performance deserves 8s or 5s or 2s, and if so which 0.25 increment within those general ranges. It's not visible or enforceable at a granular level the way fall deductions are.

And then there are the separate fall deductions, which are specific and enforced and are the same regardless of whether the skater fell on something really difficult or fell on something simple like crossovers or their ending pose. (Or something difficult that earns no base value or GOE, such as a difficult transitional move.)

As of last season, in novice competition (advanced novice, intermediate novice, basic novice), the ISU is giving fall deductions for falls outside elements but none for falls inside elements, which I think is what you're asking for at junior and senior levels. That makes sense.

https://isu-d8g8b4b7ece7aphs.a03.azurefd.net/isudamcontainer/CMS/Corporate-Site/Governance/Transparency/ISU-Communications/2699-SP-Novice-Competitions-2025-2026-1744291808-1629.pdf

Note that the deductions at novice levels are 0.5 rather than the 1.0, in accordance that novice scores overall (base values and PCS ranges) tend to be lower to begin with.

I agree with you with what a cap really means, and it generated many protestations from fans of Yuzuru Hanyu (who anyway wasn't getting 10s since 2018 even for never seen perfection) when it was edicted, but for this cap specifically, I remember having read a few years ago (yet later than these protestations) from ISU that it had to be done also with lower scores, which actually amounted not to a cap, but to a deduction. Alas I don't remember where it was written, and now I wonder if it's still the rule? Scores seem to follow it though, where there's no big interest at stake.
I remember that I was a bit shocked by how judges seemed (and still seem?) to take GOE caps as deductions. If all bullets but one of the three first ones are there, and there's a -1 call to deduct, the result is 5-1 = 4 capped at 3, not 3-1 = 2!

 

[Isk]

Jumping on the zero-points-for-a-fall train.

I'll go further. If the jump does not take off or land on the prescribed edge, then it's really not that jump, is it?

As a flutz is not listed in the code of points, it should be considered, at best, a choreographic flourish.

It's always been one of life's mysteries to me that fans will go to battle over a point or two of PCS, but then dismiss edge calls. Rotation, too, for that matter. I mean even rooting for a "lax controller" so their favorite can "get by."

I always want the controller to be Shin Amano in a really bad mood.

Do you propose not to call a jump at all and consider it part of choreography?
The reason I am asking is while I am generally on board the no credit for fallen jump ship, because falls are unambiguous and too many falls disrupt the program, so do many bad landings.
But I am afraid whether to give a full credit or no credit for a wrong edge take-off is exacerbated by inconsistent calling and borderline cases, when the panel couldn't see well or the edge is not clear. This might lead to tiny edge variations causing big score variations. In math we call this an ill-posed or unstable problem and generally tend to avoid those and try to approximate them by less exact but more well-posed / stable ones to avoid exactly this: big errors in the solutions caused by tiny errors in the data. It's like a badly tuned water tap that you turn a tiny bit and feel the water instead of changing the temperature a tiny bit go from icy cold to scalding hot.

 

[DiStSq4]

if you can find the specific clip let me know. It could be there was another error contributing to it (dropping inside shoulder, edge spiraling deeper instead of staying stable, edge starting flat then going deeper).

It was the video where they had modeled the jump in 3D. For the flip, which I was referring to, they were comparing ice marks of takeoffs, and one of them is a deep curve, which results in over-rotation on takeoff. Does anyone else know of this specific video? Please do correct me if I'm wrong.

 

[yfan1]

if you can find the specific clip let me know. It could be there was another error contributing to it (dropping inside shoulder, edge spiraling deeper instead of staying stable, edge starting flat then going deeper).

It was the bolded part here.

 

[gkelly]

I remember that I was a bit shocked by how judges seemed (and still seem?) to take GOE caps as deductions. If all bullets but one of the three first ones are there, and there's a -1 call to deduct, the result is 5-1 = 4 capped at 3, not 3-1 = 2!

I'm not sure what you mean here.

First of all, GOE rules have changed at times. When the +5/-5 GOE scale first came in (2018?), the guidelines required all of the first three positive bullet points to be present, along with one or two (or all three) of the others, for the element to start at +4 or +5, before any necessary GOE reductions. (The official terminology for negative GOEs is "reductions," distinct from the "deductions" that are taken off the whole segment score.)

Originally there was no cap on the starting GOE based on the presence of errors.

Then a few years later it was specified "To establish the starting GOE Judges must take into consideration the bullets for each element. If in an element there is any kind of mistake that requires a reduction, the starting GOE cannot be higher than +3." Is that what you're referring to?

And also "Errors with a grey background must not have higher than +2 starting GOE." (The errors with a grey background were moderate to serious errors for jumps and spins, but also included "Does not correspond to the music" for choreographic and step sequences, which could have reductions as small as -1.)

These caps were in effect last year, 2025-26, and I think also 2024-25. But not before that.

This season (2026-27), the guidelines have eased up a bit.
For positive GOEs, only the first two bullet points are now mandatory for +4 or +5 starting GOE; the "effortless throughout" bullet is no longer mandatory.
Also, this year's guidelines now state: "If in an element there is any kind of mistake that requires a reduction, the starting GOE cannot be higher than +4."
I.e., as of this year, if you have 4 or 5 bullet points including the first 2 but not the 3rd (effortless), even if there is an error requiring a reduction the starting GOE is now capped at +4 before the reduction.
The "Errors with a grey background must not have the higher than +2 starting GOE." wording from the negative GOE chart does still apply.

So if you're going to complain about judges not following rules, make sure you're comparing to the rules for the same season that the judges were judging under.