Thread: Rules to help skating's popularity?

1. Originally Posted by gkelly
[Fans like]

*Come-from-behind stories (winners moving up in the free skate, and also larger-scale triumph over adversity stories involving off-ice aspects of the skaters' lives)

*Final results matching long program results, although they will understand a skater who bombed the earlier round (short program) not being able to pull up all the way to the top even by winning the free skate.
I think this is the trickiest part. The margins of victory in the short program are often so large that the show is over before the long program begins. IMHO the most pressing need is to adjust the scoring system in such a way that a skater needs to have two good programs to win.

In the short program at Worlds Denis Ten did 4T, 3A, and 3F+3T, he had all level fours on his non-jump elements, and positive GOEs on all 13 elements. He was still seven points behind Chan -- a margin which, as it turned out, proved insurmountable despite also delivering the best LP of the night while Patrick fell twice.

2. Originally Posted by Mathman
I think this is the trickiest part. The margins of victory in the short program are often so large that the show is over before the long program begins. IMHO the most pressing need is to adjust the scoring system in such a way that a skater needs to have two good programs to win.

In the short program at Worlds Denis Ten did 4T, 3A, and 3F+3T, he had all level fours on his non-jump elements, and positive GOEs on all 13 elements. He was still seven points behind Chan -- a margin which, as it turned out, proved insurmountable despite also delivering the best LP of the night while Patrick fell twice.
This is tricky. I think there are two issues here.

1) The short program was supposed to be worth approximately half as much as the free skate, one-third of the total score, but in practice it's often a bit more than that. For example, currently there are 7 scoring elements in the SP and 13 for men/12 for ladies in the FS.

So one solution could be to delete one of the elements from the SP -- I would suggest cutting down to two spins and either rotating the requirements from year to year or else defining them as "combination spin with one change of foot" and "spin in one basic position" and requiring a flying entry to one or the other but not both.

And/or add another element to the free program (for the ladies, add two more and add 15-30 more seconds). I would make these free choice for the skaters except I would not allow a 9th multi-revolution jump element for men. If they introduced my dream of a leveled small-jump sequence, that could be an option.

Maybe adjust the PCS factors so that the component scores in the free program are worth more than twice those in the short.

Or use only three components in the short [e.g., Skating Skills (includes the technical aspects of transitions), Program Construction (choreography including phrasing, and patterning aspects of transitions), and Performance/Execution (includes the execution of the musical interpretation)] with the 1.0 factor for men, and keep all five for the freeskate.

2) I think what fans object to most is skaters who fall or make other blatant mistakes on a couple of elements still scoring well on the strength of good completed elements and good components. I do think it is fair for skaters to earn high scores for things they do very well, and I think the ISU should do a better job of explaining to casual fans, or helping broadcasters to do so, the difference between acceptable, good, and exceptional basic skating and elements. It will help if fans have a clearer understanding of what's being measured besides successful completion of difficult elements and why falls per se are not considered an impediment to winning.

But there is also value to a clean performance, and skaters as well as fans expect penalties for major errors. So maybe adopt suggestions that have been made in other threads such as increasingly larger deductions for greater numbers of falls, and more explicit instructions to judges encouraging them to lower component scores appropriately for programs with visible errors.

3. Originally Posted by gkelly
Or use only three components in the short [e.g., Skating Skills (includes the technical aspects of transitions), Program Construction (choreography including phrasing, and patterning aspects of transitions), and Performance/Execution (includes the execution of the musical interpretation)] with the 1.0 factor for men, and keep all five for the freeskate.
I like it! This idea definitely merits further exploration.

It wouldn't be a bad idea for the long program, too, with appropriate factoring if necessary.

[quote]2) I think what fans object to most is skaters who fall or make other blatant mistakes on a couple of elements still scoring well on the strength of good completed elements and good components. I do think it is fair for skaters to earn high scores for things they do very well, and I think the ISU should do a better job of explaining to casual fans, or helping broadcasters to do so, the difference between acceptable, good, and exceptional basic skating and elements.[quote]

I am less optimistic about ever getting a satisfactory balance here. What to do with a Patrick Chan who has mind-boggling blade work, but gives it all back with performance-killing falls. It's like milking a cow. You get a squirt here, a squirt there, then just when the pail is three-qurters full the cow kicks it over and you have to start all over again.

About audience appreciation of basic skating skills, I can't decide if this is art or sport. Getting points for art (Sistine Chapel, 8.5, triple Lutz, 6.0, total, 14.5).) seems sort of weird to me. IMHO the idea that this is a sensible thing to do -- to quantify quality -- is what makes attempts at tweaking the point scale here and there seem to miss the mark.

4. Originally Posted by Mathman
I am less optimistic about ever getting a satisfactory balance here. What to do with a Patrick Chan who has mind-boggling blade work, but gives it all back with performance-killing falls.
I can think of one possibility that would be a huge change to the way competitions are run and therefore not likely to happen in response to one skater:

Separate singles competition into 1) a technical phase and 2) a performance phase.

The first phase would be similar to the current long program in length and number of elements, but without music. Or at least without expectations of choreography and interpretation. All non-jump elements would have levels, and Skating Skills and Transitions would still be scored. Maybe Performance/Execution as well. The required elements could be even stricter than the current long program; e.g., all six standard jump takeoffs must be demonstrated.

This is where skaters would be crying to cram in all their hardest stuff, where we would see the most mistakes, and where it would be hard to build up a lead on the strength of PCS. The leader would almost always be the skater who completed the most difficult elements successfully.

Second phase would also be long or long-ish program length (maybe 3:30 or 4:00, not 4:30 for men and pairs) with fewer jump passes (e.g., maximum 4-6 jump passes, maximum 1 or 2 combinations), no levels only GOE on non-jump elements (everything defaults to level 1, and judges can reward difficulty in their GOEs if they want but primarily should be judging quality), and all five components with Performance/Execution weighted most highly. Maybe don't even use the tech panel for this phase since there are no levels, and just let judges reflect underrotations and wrong-edge takeoffs in the GOE if they happen to see them.

This would be more similar to the way a number of pro competitions were structured. It would put the more subjective, more audience-friendly, more often cleaner-skated event as the final round.

I'm not sure this would be a good idea. In the context of an athletic competition, would it really be appropriate to put the less technical competition phase as the deciding round?

And there still might be the same problem if the skater with the best stamina who was able to land ~8 triples/quads with level-4 spins etc. has such a lead from the first program after everyone else stumbled and fell that the winner of the second program might well be too far behind to catch up.

However, if there is one outstanding athlete who wins by blowing away the field in the first program and holding on to protect his/her lead in the second, at least that second program is likely to be relatively clean even if not the most artistic.

Or if the winner of the less-technically-demanding second program had fallen and stumbled through the first one, s/he will likely pull up several places even if not all the way to gold, which provides the come-from-behind triumph narrative.

So more often the overall results would feel more satisfying to fans who watch only the second program.

IMHO the idea that this is a sensible thing to do -- to quantify quality -- is what makes attempts at tweaking the point scale here and there seem to miss the mark.
That's a paradox skating has always faced, even when they were primarily trying to quantify the quality of circles drawn on the ice.

And for very little of the sport's history would it even have been meaningful to characterize competitive freeskating programs as art -- and even then only the exceptional ones. Primarily it's an athletic contest that includes and rewards aesthetic qualities along with the technical. The place for skating-as-art is outside competition, or at least outside Olympic-style competition. The ISU is unlikely to hold art contests on ice, but they could, separate from the Olympic style events. Or they could go back to letting other promoters do so.

5. What about a slightly different line of attack ... adding a small series of "clean program bonuses" ... (a few extra points for not making any visible errors; more for not making any errors), instead of punishing mistakes more severely?

6. I was thinking about this in relation to GOEs. Yuna Kim got a whopping bonus of 16 GOE points for "going clean" (more than the base value of a quad Axel ). Every element was clean as a whistle and got positive GOE.

Nao Asada, in comparison, pretty much kept pace on the 8 element that she received positive GOE on. But on the four elements that were not clean, she received negative GOE which cancelled a good part of the positive.

So maybe this amounts to a "going clean bonus."

7. Thank you for creating this thread, gkelly. It is so much easier to bitch about the scoring system than to offer a solution/ideas.

This is something I have been thinking about for a while...please feel free to pick it apart...I am sure there is a lot of holes in this.

1. Simplify the components categories. Although I like IJS quite a bit, it is clear that the judges don't vary their component scores enough and never will. Part of this is due to "old habits die hard" but I think a more significant reason is that it is damned hard to separately score (and essentially, rank) 24 skaters on 5 different components. Let's make 3 components, off the top of my head:
a. Choreography / "the program as a whole"
b. Performance / execution
c. Skating skills / transition

Components are judged on a 6.0 scale. So essentially, judges are giving 3 presentation marks per skater. I think going back to 6.0 on the presentation/component scores , in combination with having only 3 components, help judges be able to better vary and mark accurately.

2. Emphasize "performance/execution" and slightly de-emphasize skating skills. I attempted to do this above by combining skating skills and transitions, which somewhat go hand-in-hand. I get skating skills are so important. I understand and love that Chan and Kostner are such superb skaters and they should be rewarded for it. But I feel like currently, the skaters with the very best SS have too much of a cushion. Feel free to discuss how else this could be achieved. Its such a buzz kill when a lower-ranked/skater with lesser skating skills has a hell of a performance but gets a disappointing score.

3. The judges new component score are delivered to the crowd like how they were in 6.0. To achieve this, the judges 3 component scores would be averaged together to get an overall component score for each skater. This would help bring back the iconic reading of the marks in the K&C instead of just the totals....which, let's face it, no matter how good the commentator, is anti-climatic. Oh yeah, judges aren't anonymous either. So for example, CAN judge scores Carolina Kostner's freeskate at a 5.9 for choreography/"the program as a whole", 5.7 for performance/execution, and 5.9 for SS/transitions. This averages to an overall component score of 5.83. So the arena announcer would announce,

"The technical score for Carolina Kostner is 65.83" (just as it is now),
" her component scores are:
5.83, 5.77, 5.90, 5.93, 5.68" etc

And the on display graphics will first show the TES score and then the component score from each judge. The thought here is to make the scoring more accessible/dramatic for the spectators but to also to encourage the judges to be bolder and vary their component scoring. A skater in the first group just skated a blinder and performed the hell out of it? Let's see a 5.6 component score on the board, then.

Quick, unrelated rant: Can the ISU PLEASE update their graphics package into the twenty first century!?! It screams 1998. And it's hard to read, especially on an internet feed. I'll develop one, free of charge. Every year, I hope they'll change it. I actually prefer what they were using in 2004-2006 to what they've had the past 6 years. /graphicsrant

Now I guess we don't have to go back to 6.0 and still use the current 10 scale, but I think going up in increments of .25 is quite ridiculous. What's the difference between a 8.00 and an 8.25? But between a 5.6 and a 5.7, that's a bit more defined, especially since judges tend to somewhat set in their ways.

Now in terms of overall scoring, we would have to factor the PCS to get an appropriate balance, but I am waaaaay too lazy to attempt to figure that out.

There are a few changes to TES and GOE I'd like to see, but they are relatively minor compared to overhauling the PCS.

8. gkelly, I'd like to see the SP count for 1/3 of the score and the LP 2/3.

Would it make any sense at all to keep IJS the same but, rather than just adding up all the points at the end of the competition to determine the winner, take one thing out of the 6.0 days: score the placements of the SP as 1st-0.5, 2nd-1.0, etc., and LP 1st-1.0, 2nd-2.0, etc., so that the results are truly SP, 1/3 of the result and LP, 2/3 of the result?

I think results could make more sense that way.

9. Originally Posted by skateboy
gkelly, I'd like to see the SP count for 1/3 of the score and the LP 2/3.

Would it make any sense at all to keep IJS the same but, rather than just adding up all the points at the end of the competition to determine the winner, take one thing out of the 6.0 days: score the placements of the SP as 1st-0.5, 2nd-1.0, etc., and LP 1st-1.0, 2nd-2.0, etc., so that the results are truly SP, 1/3 of the result and LP, 2/3 of the result?

I think results could make more sense that way.
It seemed like this was at least mentioned from 2003-2006 as an idea naturally appealing to us fans because it would carry over an aspect of the old system (factored placements). I think the factored placements can be fishy. They created scenarios where a skater "needed help from another skater" to win. I suspect your idea (a mix of IJS and factored placements) did not seem appealing or fresh to the ISU, and they wanted to eliminate judging scenarios like that from the free skate. Was this their thinking? It could have been tried, and it might have worked better, but my thinking is it reduces the internal logic of IJS.

The judges in Salt Lake City were all over the place in their rankings of Kwan, Slutskaya, Hughes, and Cohen. It did not look good to people in sport who took objectivity or fairness seriously. The idea of "needing help to win" is strange in an individual sport, and it was a cute way of saying that free skate judging was somewhat dependent on how judges played the cards.

Audiences could and did live with this aspect of factored placements. It was made clear to everybody watching that the SP was high pressure and there was no margin for error. One fall or stumble in the SP, and the he skater could end up ranked where they "needed help from another skater to move up." It was okay with us fans and perhaps made the SP competition seem high stakes and more exciting.

However, I wonder if one ISU goal in the 2000s was to make the SP more technically difficult. The ISU did not want to continue to see 20+ ladies SP programs with 3Lz-2T. Because skating a clean short was so necessary to get the right placement, the factored placements perhaps encouraged 20+ ladies to do 3Lz-2Ts. This is just my theory, but I suspect the ISU viewed IJS as a "cure" for everyone doing the same jumps in the SP.

Now the irony is everything else besides the jumps in the SP often looks exactly the same!

10. About the "needing help from another skater to win," the problem now is that if you are too far back, sometimes you can't win period, other skaters notwithstanding.

For instance, under factored ordinals suppose the SP goes

Kwan
Slutskaya
Cohen
Hughes

Hughes must win the LP and hope that Slutskaya beats Kwan.

IJS

Kwan 60
Skutskaya 59
Cohen 58
Hughes 40

Sarah is out of luck.

11. Well if Sarah were really that far behind on points after the short, we wouldn't expect her to pull up to a medal unless we want what the skaters actually do during the short program to be irrelevant to their final results.

However, if the scores were
Kwan 60
Skutskaya 59
Cohen 58
Hughes 56

then she could easily win just by skating clearly the best, regardless of what order the other skaters finish in.

With these particular skaters, we also have to consider how heavily the system penalizes things like underrotation and wrong-edge takeoffs, which would make a big difference in how far real-life 2002 Sarah Hughes might have been behind in the short program and how far she could have outscored the other medal contenders in the long.

As a hypothetical example, if we're talking about four or five "clean" short programs with comparable tech content, then we would expect the SP totals to be close together, differentiated primarily by which skaters had the most strengths in basic skating, presentation, and positive grades of execution.

But if we're talking about three skaters who all skated well, and the next best skater and everyone below either completely bombed and/or were in a completely different league on quality and difficulty, then we would expect the top three to stay ahead of the pack and fight it out for the medals among themselves.

12. "
Originally Posted by Mathman
About the "needing help from another skater to win," the problem now is that if you are too far back, sometimes you can't win period, other skaters notwithstanding.

For instance, under factored ordinals suppose the SP goes

Kwan
Slutskaya
Cohen
Hughes

Hughes must win the LP and hope that Slutskaya beats Kwan.

IJS

Kwan 60
Skutskaya 59
Cohen 58
Hughes 40

Sarah is out of luck.
Talk about choosing data to make a point
With COP, you can be in 7th after the short and still win, or not. There is a lot more movement possibility in general with COP than with 6.0.

As for training fans, I think they did an excellent job at Worlds. There were dozens of short segments shown on the jumbotron before and during the competition. They explained the scoring, GOE's, under rotations, downgrades, spin levels, PCS, and more. I think one of the biggest problems is that watching skating on TV or a computer is completely different than watching it live. To the point where you really can't compare what you see through live streaming to what the judges score. IMHO, often what looks the best on TV doesn't really look that good live. That is not to say that every mark and every score is perfectly done. I do think that the judges and technical people do need to be identified by name and country affiliation. I think that the protocols should indicate which elements were reviewed by the tech panel and what the outcome was, ie. was it unanimous or did the controller have to break a tie. Finally, while I don't think all falls are equal, I do think that a fall should be an automatic -3 GOE across the board. In fact, the tech panel should enter the fall and the computer should be programmed to automatically enter the -3 so there aren't and data entry mistakes made by judges.

13. If there is a difference of less than, say 1.5 points, between the scores of the medal contenders, the skater with the least number of falls triumps. This addresses the probability of random error somewhat. If they have the same number of falls, then the skater with the highest tes receives the higher medal.

There was one year when the skaters in the GP (the final I guess) had to skate two long programs. I would say do that and drop the short program. In stead of adding the the scores together though, the average or median of the two scores would be used to determine ranking. A second long program would not inordinately add to the length of the competition.

14. [QUOTE=mag;3882723]As for training fans,...[/qupote]

I think that's one of the problems. Sports fans do not attend competitions in order to be trained.

Originally Posted by Iceman
Instead of adding the the scores together though, the average or median of the two scores would be used to determine ranking.
Averaging produces exactly the same result as adding the two scores (you just divide everyones total by two). For two scores, the median is the same as the mean (average).

15. Thanks for the clarification, Math. Do you remember what year it was that two longs were required? I just remember Irina and Michelle dukeing it out.

16. ^ It was the 2000 Grand Prix Final. There was a short program and a long program and then first and second had a skate-off (another long program) for first, third and fourth had a skate-off for the bronze medal, and , and fifth and sixth had a skate-off for fifth.

Irina beat Michelle in all three segments. Michelle did the Red Violin as her first LP, then reprised the previous year's Ariane in the skate-off. This format did not go over well, and the ISU never used it again.

Personally, I am fine with the SP and LP format. If anything, I would rather have two short programs and no long -- four minutes taxes my attention span )

IMHO, one of the strengths of 6.0 was that the skaters knew going in that if they wanted to win they had to skate two strong programs. Once in a while something unusual would occur, but by and large the deal was, top three in the short then win the long = gold medal.

17. I think ISU should fix the judging system. They also should stop redefining things, ISU seems call somethings an Elephant and when you go into details provided it sounds more like an ant. So, Performance/Execution should mean performance and execution of the routine and not something else. A fall is hugely disruptive and should be deducted from this score. When after 2 falls and a near fall on another an athlete scores 9.00 for P/E from a judges, then the whole scoring system reputation is compromised.

Falls should be penalised more in general, I was reading somewhere of relative deduction for a fall in FS. At present an athlete loses a point for a fall. If it happens on jump an addition 2.1 to 3 points are taken out (assuming a fall happens on a triple or a quad). So for a men's program where an athlete could score say 160 points, the maximum deduction in program for a fall is 4/160 roughly 2.5%. I think this is way too low. More fall there are in the program more disruptive they are, so they variant a larger deductions. I think first falls should get a 2 point deduction, 2nd fall an additional 3 point deduction (so a total of 5 point deductions for 2 falls), 3rd a 4 point additional deduction and so on...

A fall should also incur deductions in all the components to a varying degree. P/E should suffer the most, in between a 0.75 - 1.00 automatic deduction per fall. Interpretation and S/S should also suffer 0.5 point deduction per fall . A fall is not a great choreography nor is a beautiful transition they both should also bring in deductions for a fall.

I understand that ISU may have to increase the value of jumps a little with these rules, so that athletes actually take risks. However they really do need to do something about the rules.

18. Originally Posted by Domshabfan
I think first falls should get a 2 point deduction, 2nd fall an additional 3 point deduction (so a total of 5 point deductions for 2 falls), 3rd a 4 point additional deduction and so on...
I actually kind of like this idea, accumulating deductions for # of falls. Under the current system, perhaps not so harsh and start with the first fall getting 1 point deduction, then the 2nd fall getting 2, and the 3rd getting 3.

19. I've often wondered why skating doesn't award individual event medals and an all around medal the way other sports do. It might make the sport more understandable if the skaters who have the best skate in either the long or the short get the event medals, and then the skaters who accrue the most points overall get the overall medal.

I know there is a non-public "small" medal ceremony, but that is not the same for either the skaters or the fans.

20. Originally Posted by aliceanne
I know there is a non-public "small" medal ceremony, but that is not the same for either the skaters or the fans.
Re-posting from the 2013 Worlds videos thread in Kiss and Cry...

Free program small medals public ceremony videos:
Pairs (full ceremony)
Men (full ceremony)
Dance (comments by the medalists plus fun moments with a late arriving Soloviev!)