59th ISU Congress: Watch and Discuss

[gkelly]

How far apart (by percentage of the higher score?) would TES and PCS have to be from each other to qualify as the system not balancing them well against each other.

Or be evidence of a particular performance not being well balanced -- e.g., if a strong-PCS skater who is usually able to execute high difficulty loses base value or whole elements due to nondisruptive errors, or intentionally chooses to water down difficulty when a program is new or when suffering from health issues, etc.

Or, for that matter, a skater who is usually able to earn higher PCS skating slowly/cautiously and without projection to the audience when adding a few higher difficulty elements . . . and succeeding in doing so.

 

[Mathman]

To the extent that TES and PCS do often end up in the same general range, that means that the PCS factors have been well chosen to balance the likely TES for a given skill level.

I think that it is kind of a chicken and egg thing. I assume that the PSC factors were chosen by some sort of process like this: Lokk at many performances at various levels and keep track of how many TES points skaters were earning. Then set the PCS factor as appropriate to match that number on the average.

The problem is that over the years the total TES that can be achieved has grown, but the factors stayed the same. If they are still in balance under the new conditions, this is evidence that the judges are compensating by being more and more generous with PCS, at least for the top TES guys.

How far apart (by percentage of the higher score?) would TES and PCS have to be from each other to qualify as the system not balancing them well against each other.

I don't know what exact range would be considered satisfactory. maybe something like 40% to 60% of the performances should have PCS<TES and 60-40% go the other way. In the competitons that I looked at (2024 worlds) and 2013 worlds men's LP), with the factor set a 100 the split was pretty close to 50-50. With a factor of 120, in both competitions the split was 18 skaters with PCS>TES and 2 with TES>PSC.

If the factor were changed to 120, I would expect that before long the judges would be adjusting their PCSs downward to bring the two back into balance despite the factors. (Although their work would be cut out for them. They would have to give a mark of 8.0 instead of 9.50 to make it all come out right.)

 

[Mathman]

Or be evidence of a particular performance not being well balanced -- e.g., if a strong-PCS skater who is usually able to execute high difficulty loses base value or whole elements due to nondisruptive errors, or intentionally chooses to water down difficulty when a program is new or when suffering from health issues, etc.

I don;t think that any set of rules or scales of vaules can take such things into account. That is why we can only go by averages. Presumably the skater who made uncharacteristic flubs would be averaged out by the skater for whom eveything came together in a skate of a lifetime.

 

[gkelly]

I think that it is kind of a chicken and egg thing. I assume that the PSC factors were chosen by some sort of process like this: Lokk at many performances at various levels and keep track of how many TES points skaters were earning. Then set the PCS factor as appropriate to match that number on the average.

The problem is that over the years the total TES that can be achieved has grown, but the factors stayed the same.

True.

If they are still in balance under the new conditions, this is evidence that the judges are compensating by being more and more generous with PCS, at least for the top TES guys.

It seems that the shift in PCS judges toward encouraging scores above 9 happened ca. 2013-14 in the lead-up to Sochi Olympics, and the shift toward inclusion of quadruple lutzes, flips, and loops (and axels!), and multiple of the above, becoming normal at the top ranks started a few years later than that.

Of course, once 9+ component scores had become the norm, it was easier for judges to award them to skaters whose basic skating or performance quality wouldn't have earned those scores ca. 2013 but now are including difficult quads.

I don't know what exact range would be considered satisfactory. maybe something like 40% to 60% of the performances should have PCS<TES and 60-40% go the other way.

I'm not so much talking about the number of skaters who earn higher scores in TES vs. PCS, but the actual size of the differences.

Are performances that earn 80 for one and 85 for the other showing balance between TES and PCS, regardless of which one is higher? What about 80 and 90? What about 95 and 125? (Well, we know which score is which in that last example. And in all of them for women.)

If the factor were changed to 120, I would expect that before long the judges would be adjusting their PCSs downward to bring the two back into balance despite the factors. (Although their work would be cut out for them. They would have to give a mark of 8.0 instead of 9.50 to make it all come out right.)

I suspect that would only happen if the ISU technical committee started advising them to do so, if the tech committee or other powers-that-be wanted to keep the numbers for each part of score similar.

I don't think the judges would do it on their own initiative because their job isn't to keep track of the likely TES for each performance.

And readjusting their mental PCS benchmarks back down to pre-2014 levels would mean some major rethinking of what constitutes "Good" or "Very Good" vs. "Excellent" or "Outstanding."

Even taking account the changes in component definitions with the switch from five to three components, how would the actual performances of the early IJS era compare to those of the 20-teens and 2020s in terms of today's PCS criteria, or the early IJS criteria for that matter?

 

[Mathman]

I suspect that would only happen if the ISU technical committee started advising them to do so, if the tech committee or other powers-that-be wanted to keep the numbers for each part of score similar.

I agree with this, but I suspect that the technical committee would in turn be under pressure to issue such advice. The ISU, having advertised and promoted a scoing sytem that provides scoring balance, would be happiest if they could then point to the results and say, "Look how wonderfully our scoring sytem is working out."

 

[gkelly]

I agree with this, but I suspect that the technical committee would in turn be under pressure to issue such advice. The ISU, having advertised and promoted a scoing sytem that provides scoring balance, would be happiest if they could then point to the results and say, "Look how wonderfully our scoring sytem us working out."

And yet, they have not made any move to increase the PCS factoring to better match the TES in recent years. TES significantly higher than PCS does not seem to bother them.

 

[Mathman]

And yet, they have not made any move to increase the PCS factoring to better match the TES in recent years. TES significantly higher than PCS does not seem to bother them.

It doesn't seem to bother them at the level of the individual skater. There will alsways be some skaters who get big tech and others who are more impressive on the performance side. As long as the balance works out OK on the average, I don't think that there is anything much that the ISU can do about it.

Would Jason Brown benefit by an increase in the PCS factor from 100 to 120? At 2024 worlds LP he had the highest PCS, beating Kagiyama by 0.37 points, Sia Him Fa by 1.71 points, Ilia Malinin by by 3.37 points. and Shoma Uno by 5.14 points. (Uno's PCS were lower than usual because of jump mistakes).

Now raise the factor to 120. Brown's margins of victory rise to 0.44 versus Kagiyama, 2.05 against Sia Him Fa, 4.04 versus Malinin and 6.47 versus Uno. These tiny bonuses would not cause Brown fans to jump up and down in delight nor provide the ISU with anything to feel smug and satisfied about one way or the other.

 

[4everchan]

It doesn't seem to bother them at the level of the individual skater. There will alsways be some skaters who get big tech and others who are more impressive on the performance side. As long as the balance works out OK on the average, I don't think that there is anything much that the ISU can do about it.

Would Jason Brown benefit by an increase in the PCS factor from 100 to 120? At 2024 worlds LP he had the highest PCS, beating Kagiyama by 0.37 points, Sia Him Fa by 1.71 points, Ilia Malinin by by 3.37 points. and Shoma Uno by 5.14 points. (Uno's PCS were lower than usual because of jump mistakes).

Now raise the factor to 120. Brown's margins of victory rise to 0.44 versus Kagiyama, 2.05 against Sia Him Fa, 4.04 versus Malinin and 6.47 versus Uno. These tiny bonuses would not cause Brown fans to jump up and down in delight nor provide the ISU with anything to feel smug and satisfied about one way or the other.

The factoring alone cannot work if the judges think that Brown is only 3.37 over Malinin in PCS. Both, an adjustment in the factoring and in HOW judges are rewarding skaters need to be done. If a skate, an expert in quads can get 30 points in TES over many guys, why can't a expert PCS skater can only get 3.37 points over another skater ?

 

[lariko]

I think in the current world, where individual opinion is valued very much, making PCS a deciding factor will be a mistake, because it will hugely increase the frustration when someone wins/doesn’t win and the anser is basically Crom, like in Conan the Barbarian said whenever he faced anything inexplicable. The PCS marks are too nonsensical and esoteric to lead to agreement. While some stalwart can stubbornly argue that Malinin or Chen are crap at PCS, wast majority comprehend and agrees with their wins, specifically because they see clear tech and perfectly fine skating that looks great for anyone who is not inane.

Where the confusions comes now for me is in the tiers of skaters who have similar content, the majority in the competition, because there is absolutely no way at all to explain the ranking of the guys from 5 to 15-ish for men and pretty much everyone in women. Unless someone has a terrible no good day in a very obvious way, the difference in skating is just too insignificant and choreography is a matter of prefernce for results to be anything but random/bias.

Once the rules change, I think I won't be able to tell anything at all, like it's in women rn. And if I can't see why, figure skating is just a show, not a competition, and there are more interesting shows out there with plot that is more exciting.

 

[Mathman]

The factoring alone cannot work if the judges think that Brown is only 3.37 over Malinin in PCS.

It just seems like an impossible task to work it all out (easy though it is it is to complain about it). I could live with, with, say 8.75s for Malinin and 9.75s for Brown. But then there are others that also deserve 9.75s on the same scale, like Kagiyama and Uno (not to mention some not quite so prominent contenders like Cha and Vasilievs) and you would have to give 10.0s to Hanyu. So everybody is still bunched together at the top and the PCSs play only a minimal role in terms of distinguinshing one performance from another.

We could look at it like this. There are 40 possible marks that can be given for each component. It would seem reasonable to suppose that a mark of 0.25 means the skater is in the lowest 1/40th of all skaters. If you are in the top 1/40th, you get a 10. Well, every skater at the world championships -- every skater we ever see on TV -- is in the top 1/40th of all skaters. So I guess they all deserve 10s. across the board.

On the other side of the ledger, we have been encouraging skaters for years: Do a quad Lutz -- we'll give you 11.50 points if you do. Do a quad Axel -- I double-dog dare you -- I;ll give you 12.5 points.

Someone takes them up on it and all of a sudden it's, ha, ha, just kidding. That's too many poimts, It throws the scoring sytem out of line. We'll take away a jumping pass and restore order and ballance.

OK. It would also be OK if they didn't.

 

[4everchan]

It just seems like an impossible task to work it all out (easy though it is it is to complain about it). I could live with, with, say 8.75s for Malinin and 9.75s for Brown. But then there are others that also deserve 9.75s on the same scale, like Kagiyama and Uno (not to mention some not quite so prominent contenders like Cha and Vasilievs) and you would have to give 10.0s to Hanyu.

Hanyu is retired. No need to give him 10s.

So everybody is still bunched together at the top and the PCSs play only a minimal role in terms of distinguinshing one performs]ance from another.

So we have a problem

We could look at it like this. There are 40 possible marks that can be given for each component. It would seem reasonable to suppose that a mark of 0.25 means the skater is in the lowest 1/40th of all skaters. If you are in the top 1/40th, you get a 10. Well, every skater at the world championships -- every skater we ever see on TV -- is in the top 1/40th of all skaters. So I guess they all deserve 10s. across the board.

On the other side of the ledger, we have been encouraging skaters for years: Do a quad Lutz -- we'll give you 11.50 points if you do. Do a quad Axel -- I double-dog dare you -- I;ll give you 12.5 points.

Someone takes them up on it and all of a sudden it's, ha, ha, just kidding. That's too many poimts, It throws the scoring sytem out of line. We'll take away a jumping pass and restore order and ballance.

See it the other way around.
At the beginning of IJS, PCS was low. Very low. Also, they had to do 2 step sequences, which counted for a bit more than now (scroll back to @Arigato 's list of best scoring Steps and you see that rules have changed concerning the GOE factor applied to the BV of steps).
So, they nixed the second steps and included a bunch of silly rules about how to get level 4... which now looks like a marching band drill instead of cool steps... and they lowered the maximum points on steps... All along, the ISU was sending a message : you better jump big because the other elements are not going to get you there... and btw.. don't worry, it's over the times when Jeff Buttle would score 78 in PCS and probably be on top of the food chain with that score. We give everyone and @el henry 's little cousins from 8.5 to 9.5 nowadays

So of course, some people, (I am not alone... ) are not happy.

OK. It would also be OK if they didn't.

Here is what I'd do;

Technical program instead of SP. Counts for about 1/3. 6/7 jumping passes, 2/3 combos. 4 minutes. 1 leveled step sequence, 3 leveled spins... yeah....(almost what we get as a LP)

Free program instead of LP. Counts for about 2/3. 3/4 jumping passes. 1 combo. 4 minutes. 2 leveled step sequences. 1 choreo sequence. 2 leveled spins. 1 choreo spin.

Would this favour good jumpers... probably because getting a good score in the technical program would mean skating later for the free, and still probably having a huge lead... but the factoring of programs would compensate. Also, the programs would be a bit more different from one another. The skaters would have room to play in the free program.

you better be an interesting skater with good skating skills and presentation to skate my free program

The ISU needs to revamp the concept of the two programs, one short and one long that are pretty much copy /paste from one another.

 

[Mathman]

I'm not so much talking about the number of skaters who earn higher scores in TES vs. PCS, but the actual size of the differences.

Are performances that earn 80 for one and 85 for the other showing balance between TES and PCS, regardless of which one is higher? What about 80 and 90? What about 95 and 125? (Well, we know which score is which in that last example. And in all of them for women.)

Personally, I am quite comfortable with a difference of 10 points. If a skater gets 80 on TES and 90 on PCS, that individual skater did not present a very well balanced program, at least in that performance. I do not see anything wrong with that per se. The skater receives points for the elements that he does and he also recievs points for the quality of his skating overall. All is well.

Even if there is a large gap there is usually an entirely satisfactory explanation: the skater got lower TES than expected because he fell on his big jumps. His music choices this season suck -- what was he thinking of? These considerations do not requrie any "corrections" to the rules.

Even if there is no particular reason -- this skater is just better at one aspect of skating than another, well, that's what makes the (skating) world go 'round.

The virtue of simply counting up the number of skaters with PCS>TES versus the number with TES>PCS as a measure of whetherthe factoing is about right or not overall, is that it is easy to do at a glance and does not require any computation -- just counting -- or statistical savvy.

 

[el henry]

My little cousin's head is spinning

But I do enjoy the back and forth on how to even out scores and how to make certain that all avenues of excellence, which even my little cousin can understand, are rewarded.

 

[Mathman]

Here is what I'd do;

Technical program instead of SP. Counts for about 1/3. 6/7 jumping passes, 2/3 combos. 4 minutes. 1 leveled step sequence, 3 leveled spins... yeah....(almost what we get as a LP)

Free program instead of LP. Counts for about 2/3. 3/4 jumping passes. 1 combo. 4 minutes. 2 leveled step sequences. 1 choreo sequence. 2 leveled spins. 1 choreo spin.

Sounds good to me (although what we have now ais also pretty good, IMHO ).

Didn't the ISU run something like this up the flag pole a couple of years ago? Only it was two programs of equal length and weight with separate medals given for Techy Program, Pretty Program, and then somehow these would be conbined for Overall. General fan reaction was NO! what we like about figure skating is everything and the kiitchen sink thrown together in one program.

 

[gkelly]

Sounds good to me (although what we have now ais also pretty good, IMHO ).

Didn't the ISU run something like this up the flag pole a couple of years ago? Only it was two programs of equal length and weight with separate medals given for Techy Program, Pretty Program, and then somehow these would be conbined for Overall. General fan reaction was NO! what we like about figure skating is everything and the ku=itchen sink thrown together in one program.

Well, we never got the ISU's actual proposal about this because it would have been for the 2020 congress, which was cancelled. So we could only complain (or approve) about what we guessed they might mean.

 

[Mathman]

Well, we never got the ISU's actual proposal about this because it would have been for the 2020 congress, which was cancelled. So we could only complain (or approve) about what we guessed they might mean.

It was a very bad year all the way around

I remember when this format was used in a lot oft pro competitions back in the day. A Tech Program and then an Aetistic program. At the pro level the Artistic Program dominated. Compared to ISU competitions, the pro performers were past their technical peak but bursting forth with program ideas on the entertainment side.

 

[Mathman]

But I do enjoy the back and forth on how to even out scores and how to make certain that all avenues of excellence, which even my little cousin can understand, are rewarded.

In all seriousness, though, this aspect of "what do audiences understand and appreciate" has kind of fallen by the way side. Otherwise, spins would be more highly valued. Everybody can recognize a fast, well-centered, beautifully positioned spin. The viewers can even tell twhen the spinner changes position, though they are not too much interested in "levels."

 

[4everchan]

It was a vey bad year all the way around

I remember when this format was used in a lot oft pro competitions back in the day. A Tech Program and then an Aetistic program. At the pro level the Artistic Program dominated. Compared to ISU competitions, the pro performers were past their technical peak but bursting forth with program ideas on the entertainment side.

just to be clear... i am not talking about an artistic or aesthetic program... it would still have serious tech content BUT much less and more of the stuff that requires skaters to stay on the ice with their blades

BTW... in many many cases, the crowd cheered the loudest for step sequences and choreo sequences at worlds... Almost only Ilia manages to get uproar with his jumps

So of course, let's add more steps

 

[gkelly]

In addition to voting for your personal favorite format at https://www.goldenskate.com/forum/threads/your-ideal-program-format.98563/, please feel free to use that thread for discussing what you like or dislike about the various formats, or how you would define some of them differently than I did.

 

[TallyT]

Hanyu is retired. No need to give him 10s.

Sit them all in front of RE_PRAY and tell them if they can do anything like that, even if only for 4 minutes, they might get 10s

The ISU needs to revamp the concept of the two programs, one short and one long that are pretty much copy /paste from one another.

Because I am blithely ignorant of the technical and political minutiae. I can go further - keep the SP but ban any jumps whatsoever from it, make them work to make the rest both superlative and exciting (and before anyone says 'can't be done', after Pyeongchang, Yuzu performed in Continues with Wings while still healing from that ankle injury and did a selection of his competitive programs shortened and reworked to take the jumps out. Audiences still found them enthralling. It ain't easy but it can be done and at elite level 'ain't easy but can be done' should be a good challenge.)

(And yes, I am tongue superglued in cheek here so jump fans, please no conniptions at the mere thought)