Mandatory PCS Reduction for Falls?

[Shayuki]

I think PCS shouldn't be influenced by the technical elements - that's not the idea at all.

This is something I've been thinking about, actually. In inline skating freestyle slalom for example, the artistic grade can only be within 10 points of the technical score(max tech is 60). I wonder how that kind of a system would work in figure skating.


It would make it much more difficult to pump up a skater who had a terrible performance. Though, it could make it more difficult to reward someone whose artistic ability is far stronger than their TES. On the other hand, it'd allow for some insane PCS scores if one did get crazy high TES. So if for Men you could in FS have your PCS within 20 points of your TES, if you happened to have a 115 TES skate you could have up to 135 PCS.

 

[Mathman]

I wonder if the ISU could acheive the same result without excessive rules. They have already come out with some sort of "advisory" to judges that they shouldn't give 10s to a program marred by a fall. Going forward, they could emphasize that point of view in their judges' training seminars, practice judging evaluations, and communications from the council, without trying to pre-judge what scores should be given out for every combination of technical strength and weaknesses.

But yeah, if you fall, the judges should not be cheering, by their PCSs, "what a superb skate!"

 

[andromache]

It would make it much more difficult to pump up a skater who had a terrible performance. Though, it could make it more difficult to reward someone whose artistic ability is far stronger than their TES. On the other hand, it'd allow for some insane PCS scores if one did get crazy high TES. So if for Men you could in FS have your PCS within 20 points of your TES, if you happened to have a 115 TES skate you could have up to 135 PCS.

Oh I don't like this at all. It would skaters like Samarin and Zhou wayyyyyyyy too much of an advantage over the Jason Browns and Misha Ges of the world.

Though at the same time, this is basically what the judges do anyway. Jason Brown is the exception in that he is getting the top-tier PCS without the top-tier technical content, and that's probably only because he comes from a big enough federation.

 

[gkelly]

As I recall, originally (back in the fall 2003 test version of IJS, and maybe in the 2004-05 season), judges were supposed to reduce the Performance/Execution component (by 1.0?) for each fall, but they didn't do it, at least not obviously and consistently. So the solution was to institute the fall deduction of 1.0 per fall subtracted from the total score. Just last year that was amended so that the third and subsequent falls got larger penalties.

If the idea is to put the penalties back into the program components, to the point that otherwise excellent programs that happen to involve butt (or both hands and both feet) to ice for a split second, need to lose significant points per fall, the question arises whether the significant penalty should apply to all falls, or whether the severity of the penalties should be proportional to the severity of the falls.

How many components would this apply to? Remember that freeskate components are factored, so that taking off one whole 1.0 per fall would actually subtract 2.0 from a men's freeskate or 1.6 from a lady's or pair's. (For short programs, though, the penalties would be half that.) Applied to multiple components, 1.0 from each would quickly add up to a lot of reduction. However, if the mandatory penalty is only 0.25 per component, and it doesn't automatically apply to all components (e.g., I have seen skaters continue to interpret music without missing a beat despite a fall), then the mandated penalty would be in line with the current fall deductions.

And of course judges are always free to reduce any or all component scores further for more or more disruptive falls.

It might be appropriate to put the responsibility for automatically penalizing falls back into the PCS, but if so, especially if those penalties are severe, then it should be instead of the current fall deductions, not in addition.

And you might end up with differences of opinions among judges about whether there had actually been a fall. One judge looked down to her notes or computer screen for three seconds, just long enough to miss it entirely. Another judge saw a stumble but didn't think there was any weight on any part of the body besides blades. In some cases, a judge might consider a drop to the ice as an intentional part of the choreography (and in some cases, at least before the current fall deduction rules, they might have been right).

So putting the determination of falls back on the judges would lead to more inconsistencies, and also give judges more to worry about. Fans who complain that judges already have too much to do in judging PCS surely wouldn't want to give them yet another task to keep track of -- right?

Better to leave it with the tech panel to determine what was or wasn't a fall. And then if you want the computer to subtract the penalties from one or more PCS instead of from the total score, it just becomes a matter of semantics, which also may be affected by the factoring or the size of the penalty and how many components it's applied to.

Here's a performance with a fall from the 6.0 era that did in fact earn at least one 6.0 for the second mark, along with many 5.9s:
https://www.youtube.com/watch?v=1CoRH9tH_7I

Where do you think it should be capped if it were an IJS performance?

PCS : fall should reduce automatically perfomance and execution... but in some cases it does affect composition and transition... ... so it is hard to gage.... i would tell judges that any fall caps PCS at 9.50 but wouldn't dare to go much further than that.... because it does become subjective here

I could live with that, along with keeping the current fall deductions. However, for a truly great performance marred only by one brief fall, straight 9.5s should still be possible.

THOUGH : I'd like to see what would happen if skaters were told that fall = 0 point... Forget the deduction... just don't give value for an "invalid" element.. and to be valid the jump needs to be landed on your foot... (touch hand down/2 footed jump would be a major deduction) landing on butts, knees, belly or head = 0

Nope. What would happen in that case is that skaters would skate much more cautiously -- either water down their technical jump content only to include moves they have 99% confidence of executing successfully, or else slowing down and ignoring the music and performance qualities to concentrate on not falling.

If a judge is scoring a cautious, don't-fall-at-all-costs performance vs. a confidently aggressively skated performance that results in one fall, the latter should deserve higher PCS. But you can't build in mandatory reductions for skating cautiously.

Hardcore? Yeah... but maybe we would see cleaner programs with less subjective judging.

Not everyone is interested in a sport of careful technical execution with no risk.

Why not trust the judges to use their own judgment, and get out of the mindset that falling is somehow an insult to the audience. They're going to be penalized anyway, and we can debate exactly where and by how much, but falls aren't the only flaws that need penalty.

That’s completely reasonable but do you think that judges use PCS to make up for a fall ever and counter the lost points or do you think I’m wearing a foil cap here?

I very much doubt that many judges are looking at an otherwise good performance with a fall and thinking "I think that performance really deserved a 9 but it deserved to win, and the fall deduction is going to make the scoring close, so I'm going to give 9.5 instead to make sure it wins."

I think if they do give 9.5, they think the performance as a whole was excellent and might have deserved 10.0 without the fall.

If it was me, i would force judges to instead of giving a grade, to literally check bulletpoints for GOE (and let a computer compute final GOE), and evaluate each aspect of each pcs category (in a more rough scale, such as bad / average / good, for example, to be later added up and averaged by the computer). All the deductions for URs and falls and whatever else would be automatic too. All the info would be released with the protocols.
Advantage 1: you can, instead of just taking out the lowest and the highest result, check individually each aspect. For example, if out of 9 judges, only 1 said a jump had difficult arm variation, that surely does not applies. If all but 1 judge evaluated the purpose apsect of choreo as good, that means something too.
Advantage 2: full accountability of each judge, easy to track and contest or discuss. The system may still not be easy, but you can see where did this 9 in PCs or this +3 GOE on a jump came from.
Advantage 3: less reputation and so on scoring, since you have to evaluate lots of small specific aspects, not just give a single value and thats it. Kinda naturally forces some objectivity on it.

Yes, Advantage 3 is the biggest advantage IJS has over the old ordinal scoring. Breaking down the scores even more would magnify that advantage . . . while also putting more burden on the judges to make many many notes and divide their attention while watching the performance in real time. I think you'd need more judges, dividing the work, to capture all the details in the computer and not just in the judges' brains summarized per element or per component into one score. Could you create a system that allows you to input and record all these details in real time without taking your attention away from the ice?


I dont think we would even need to change the current values and deductions, just force judges to actually follow it.[/QUOTE]

 

[kenboy123]

Such scoring would really make the sport come down to good or bad luck and rid the PCS of its meaning.

I agree with this. There is a reason there is such thing as a PCS and that is to reward the actual components of the program. A fall or a underrotation or anything that the skater actually did wrong on the ice should be addressed execution part of the score, am i wrong here???...

 

[4everchan]

Here's a performance with a fall from the 6.0 era that did in fact earn at least one 6.0 for the second mark, along with many 5.9s:
https://www.youtube.com/watch?v=1CoRH9tH_7I

Where do you think it should be capped if it were an IJS performance?

This is interesting for various reasons

1) the fall happens on a spin... right now, that spin would have counted as 0.. right? and the fall deduction - 1
2) I have a hard time dealing with comparing a 6.0 to a COP 10.0... because back then, these were ordinals... of course, the judges were supposed to deduct... but and ordinal is an ordinal... and a 10. is an absolute mark... (in principle)... However, in here, that fall was very disruptive since it emptied the program... the 6.0 judge was a nutjob even if this skating is so pure and beautiful and very well thought... I would have capped performance and execution at 9.5, musical interpretation to me isn't affected here, composition is... and this is harsh but falling on this spin at such an important point would probably be capped at 9.0 or 9.25.... transitions were not affected by this IMHO nor were skating skills...
3) this debate is made more interesting because of the overall quality and purity of the skater... but that's the issue.... we need to reward general quality overall with giving higher PCS mark and higher GOE but when ONE element fails, it needs to be penalized... the issue for me is that judges haven't quite yet learned how to judge both GLOBALLY and LOCALLY . Under 6.0 they judged very well globally... elements in one bunch... and "overall" artistic impression.... in one bunch...

Right now, judges need to do both... assign GOE on EACH element, yet assign PCS in 5 categories, not just an umbrella number and on top of that, there will always b a sense of overall quality that needs to be reflected.... to me a skater like Lynn could fall a couple times and still get high scores because her skating is so pure... however, in a performance based sport, how does one can account for a lesser quality skater, getting 100% of their job done...versus a top notch skater getting 85% of their job done....

that's one of the reasons judges have a hard time, and fans are always calling 911 on robbery...

4) this was also interesting to me for the orchestral version of the Grieg Notturno which i only know in piano version... it was very nice.

Nope. What would happen in that case is that skaters would skate much more cautiously -- either water down their technical jump content only to include moves they have 99% confidence of executing successfully, or else slowing down and ignoring the music and performance qualities to concentrate on not falling.

If a judge is scoring a cautious, don't-fall-at-all-costs performance vs. a confidently aggressively skated performance that results in one fall, the latter should deserve higher PCS. But you can't build in mandatory reductions for skating cautiously.



Not everyone is interested in a sport of careful technical execution with no risk.

Why not trust the judges to use their own judgment, and get out of the mindset that falling is somehow an insult to the audience. They're going to be penalized anyway, and we can debate exactly where and by how much, but falls aren't the only flaws that need penalty.

I see where you are coming from but your explanation about being more careful and slowing down is one way of avoiding falls but not necessarily the only way.
Also, in the example you brought in, the skater would receive 0 for that spin... like some do for pair spins or aborted lifts etc... so "failure leading to a score of 0" is already in place and not just in jumping rules (zayak or double/triple jump requirements)

If we compare to other disciplines that are performance based, it is quite normal to trim down a program or making it safer in order to retain a pure and fluid execution without big mistakes... musicians do it all the time, dancers too... because actually, nobody pays to go see a concert to hear singers crack or musicians play wrong notes... same with dance.... I have heard multiple times things like "we are doing the simplified version tonight, my ankle is too sore" or whatever. Lifts are often downgraded too when things are not 100% mastered or dancers aren't fit... All of this because it's better for the greater public to see a performance that is done with confidence....

Also, skaters already slow down into quads and triple axels... and still fall on them... as a performer, I don't go out in public when i haven't mastered my content 100% ... how many times have Adam and Jason landed rotated quads (lutz and toe respectively) yet they include them in every single program.... It is becoming a planned fall because for that jump they receive more than a clean 2a.... this is what Jeffrey Buttle (olympics 2006) did before he decided to go for a clean performance in 2008 when he won worlds...

do we want to encourage planned falls? Do we accept that a failed quad jump brings more points than a good double axel or even some triples?

I don't think it's good for the sport to have skaters fall multiple times and win the top prize based on higher base value... though, I am as excited as everyone to see the sport develop and i hope it keeps going ... it's a special era in figure skating for sure.

 

[Mathman]

Here's a performance with a fall from the 6.0 era that did in fact earn at least one 6.0 for the second mark, along with many 5.9s:

https://www.youtube.com/watch?v=1CoRH9tH_7I

Where do you think it should be capped if it were an IJS performance?

To me, this question is kind of moot, for the reason that we will never see a program in the IJS era that features this kind of investment in every physical and choreographic detail.

I do have an off-topic question, though. In the jump sequence just preceding the element that she fell on, how would that be scored? It is something like a 1A + 1/2 loop + ? (2S?) + some turns into a double something. Would this be scored as two separate passes?

 

[Mathman]

Such scoring would really make the sport come down to good or bad luck and rid the PCS of its meaning.

I hope you don't mean that falling on a jump is a matter of good or bad luck. It is a matter of good or bad execution of technique.

Even if a skater does 99 jumps in practice without a fall, if she falls on that 100th one, there is a reason that she fell on that hundredth time.

 

[gkelly]

Such scoring would really make the sport come down to good or bad luck and rid the PCS of its meaning.

It is true that sometimes falls are a result of bad luck (a bad patch of ice -- more likely later in the draw, which is also a matter of luck) rather than bad technique.

 

[chillgil]

but what if having multiple falls actually fits with the program? lol stay with me here but there ARE some pieces of music and some programs that could still be sold fantastically even with major errors. The one that comes to mind is Yuzu's 2014 cup of china FS. Who knows . . . I might be the only one who loves that free skate on that day in particular objectively as a fan of performing arts, not as a fan of hanyu because . . . you know that comp sucked for him for every reason imaginable. But he fell on almost every jumping pass and it worked for the program and it fit the character of the phantom. Yeah obviously his SS and TR score shouldn't be so high but I wouldn't hesitate on giving that performance 9s or higher in the right areas

sometimes a fall works with the program especially if coordinated with the music..

 

[4everchan]

 

[chillgil]

..........https://www.youtube.com/watch?v=HkE83Zu8z58 4everchan

 

[4everchan]

 

[Viiktoruu]

I hope you don't mean that falling on a jump is a matter of good or bad luck. It is a matter of good or bad execution of technique.
Even if a skater does 99 jumps in practice without a fall, if she falls on that 100th one, there is a reason that she fell on that hundredth time.

It is true that sometimes falls are a result of bad luck (a bad patch of ice -- more likely later in the draw, which is also a matter of luck) rather than bad technique.

Of course it's not only luck, but it certainly is a factor too. If you fall once out of 100 jumps, the luck is on when it happens as well. If the falls would be penalised to the point we're talking here - this would lead to drastic reduction of the program difficulty, focusing entirely on the technical side (not to fall and then be punished through the PCS as well), front-loading, and finally - a bit of luck: is the one fall in 100 jumps at a big competition, and not even PCS will be high (even though the skater might have hit every beat of the music and had wonderful transitions and edges etc.) so that they lose to an overall worse skater, who just didn't have that 1 fall in 100 jumps that day.

9,5 average PCS --> 8,5 or 9,0 (different users suggested different caps) are at least 4 points less than deserved (ladies and pairs), and only due to one jump. And let's say the jump was a 2A - you get a bigger deduction than the base value of the jump on the level of PCS, additionally all -3 GOE (-1,5 points) and a -1 for deduction. That is about -6,5 points for a jump with a BV 3,3. Now that's just not right.

 

[gkelly]

This is interesting for various reasons

1) the fall happens on a spin... right now, that spin would have counted as 0.. right? and the fall deduction - 1

Yes, that spin would get 0 points -- no base value, and no GOE.

If a skater gets at least 3 revolutions of spinning in a basic position before she falls, then she gets some points for the spin.

A fall at the end of a level 4 spin with all the features accomplished could still leave the a respectable base value. And if it was really good up until the fall, then the final GOE might not be -3.

But a fall at the beginning of an element pretty much invalidates the element. Just as falling on the takeoff of a jump results in a total score of -1.0, whereas falling on a landing gives base value for the jump as executed, -3 GOE, -1.0 fall deduction. For a rotated triple axel or quad, that still leaves quite a lot of positive score left. For a single or double or downgraded triple jump, the element score will be less than 1 so the net score after fall deduction is negative.

2) I have a hard time dealing with comparing a 6.0 to a COP 10.0... because back then, these were ordinals... of course, the judges were supposed to deduct...

Not really. In short programs (which didn't exist for singles until a few months after that performance), judges were supposed to deduct for errors or rule violations.

For freeskates, they only deducted for rule violations. Otherwise, each mark was basically the judge's composite opinion of all the good, mediocre, and bad aspects of the performance, as applicable. Similar to each component now. The technical mark probably started at a benchmark score for that level of program, adding value for elements completed or partially completed and
and then adjusted up or down for quality and for missed elements.

I would have capped performance and execution at 9.5, musical interpretation to me isn't affected here, composition is... and this is harsh but falling on this spin at such an important point would probably be capped at 9.0 or 9.25.... transitions were not affected by this IMHO nor were skating skills...

Makes sense.
Of course that's your personal evaluation. Another judge might be harsher in their opinions, and a third judge more lenient. So it would be hard to agree on official caps that apply to all judges evaluations of all possible performances by top-level skaters with one fall.

Right now, judges need to do both... assign GOE on EACH element, yet assign PCS in 5 categories, not just an umbrella number and on top of that, there will always b a sense of overall quality that needs to be reflected.... to me a skater like Lynn could fall a couple times and still get high scores because her skating is so pure... however, in a performance based sport, how does one can account for a lesser quality skater, getting 100% of their job done...versus a top notch skater getting 85% of their job done....

Each judge needs to evaluate and weight all of those aspects. In 6.0 they had to do it all into two marks, or one placement using the two marks as mnemonics to keep the skaters in the order they thought they deserved. (Back in the 70s, total points for all the figures and freeskating, and short program once relevant, was one of the tiebreakers, so the margin of victory in each score did make a difference in a way that was no longer the case after 1980, but that's a digression.)

In IJS judges can separate their opinions about failed elements from their opinions about the successful elements, and both from their opinions of the skating skill or performance or interpretation as a whole. And they're not called on to evaluate the difficulty of the technical content at all, just its quality.

Some fans complain when judges don't keep the different scores separate and seem to allow skating skill or technical content to dictate other components. Other fans complain when judges do keep them separate and don't reflect failed elements in the PCS. Or maybe sometimes it's the same fans who will complain no matter what the judges do.

Also, in the example you brought in, the skater would receive 0 for that spin... like some do for pair spins or aborted lifts etc... so "failure leading to a score of 0" is already in place and not just in jumping rules (zayak or double/triple jump requirements)

Yes, for failures at the beginning of an element, or especially in short programs for errors that cause the element not to meet the requirements for that required element slot.

If we compare to other disciplines that are performance based, it is quite normal to trim down a program or making it safer in order to retain a pure and fluid execution without big mistakes... musicians do it all the time, dancers too... because actually, nobody pays to go see a concert to hear singers crack or musicians play wrong notes... same with dance....

But sports fans pay to see athletes take risks. Faster stronger higher. And competitive skating is primarily a sport. On thin blades on a slippery surface where falls are always possible if the skater's balance is off by the tiniest fraction of an inch. Or even if the skater doesn't actually make any mistake at all but happens to hit a rut or a stray sequin left by the previous skater at just the wrong moment.

For show skating, the comparison to performing arts values would make more sense.

do we want to encourage planned falls? Do we accept that a failed quad jump brings more points than a good double axel or even some triples?

I would prefer not. And I think the answer there is to make the value of the negative GOEs on those high-value jumps much larger, so that -3 (or -5 depending how rules change next year or whenever) would take away more than half of the base value.

But I do think that it makes sense for a rotated quad with a fall to be worth more than a downgraded quad with a fall, or than a rotated triple with a fall. Many more skaters are capable of achieving the latter than the former.

I do have an off-topic question, though. In the jump sequence just preceding the element that she fell on, how would that be scored? It is something like a 1A + 1/2 loop + ? (2S?) + some turns into a double something. Would this be scored as two separate passes?

Good question.

What I see is 1A (landed on the opposite foot, i.e., one-foot axel), half-flip, 2A+1Lo< (landed on the opposite foot, i.e., half-loop), half-flip, inside axel (nonlisted jump)+2Lo<<.

There are no steps or turns between the jumps, only nonlisted jumps and the toe-assist on the landing of forward-landing jumps (half-flips), so this should meet the current IJS definition of a jump sequence.

The scoring for a jump sequence is the total of the two highest-value listed jumps, times 0.8.

The double axel is obviously the highest-value jump in this sequence. Which is the second highest? Single axel is currently worth 1.1; double loop is 1.8, underrotated it's 1.3, downgraded it's 0.5. I'm not sure whether they base the determination of highest value on the intended element or the element as executed. In this instance, which jump counts could depend on whether that 2Lo was called as under or down.

Of course, under the current rules it wouldn't be worth putting all those jumps into the sequence. She'd have been better off at least stopping at the inside axel, if not at the double axel.

 

[JSM]

I would be interested in a discussion about a base value deduction for a fall (similar to the deduction you receive for an under rotation). In both instances, it is a failure to complete the jump successfully.

I'm not a fan of penalizing jump falls in PCS - it's a technical error and should be penalized as such. Though I do agree that 10.00s for a performance with a fall is ridiculous.

 

[college pat]

actually i agree with the mandatory PC falls deduction.

they did it for figures, under 6's 0

in most cases it does hurt the flow, choreography, interpretation of the program.

in a few cases it doesn't but majority of time it does; even for half a second.

 

[4everchan]

thank you gkelly for your answers. It's an interesting debate indeed... I am just hoping for cleaner programs but I do appreciate risky elements... I think though that sometimes, everything is sooooo risky that it's tough to enjoy a performance...for instance when a skater invests so much on his first jump.... and it fails... and then every other jump becomes a source of worry...

After the 2002, I started to hate pairs.... not only there was the scandal.. but also, i had been watching them for a decade at last and they were always doing the same things... 2a, 3t jumps.... 3s, 3loop throws etc... it is only when D/R started pushing the content with their lutz throws and SBS that I got excited again... other teams brought up their levels where they could as well, whether it was on the twist or the SBS 3-3 or the 3a throw.... however, except for one team that for sure we know the throws may be risky, in general, D/R land their lutzes, S/H have a great quad twist, S/K used to land their 3-3 very well... and other teams have all raised their content.... Z/E last year had puny content... this year, they have picked up elements that weren't done just a few years ago... But... as I said, I do feel that
1) falls on quads are very much more punished in pairs because the bv isn't high... even with sbs 3lutz, i would raise the bv because it's so rare in the field that i find it weird that it's just 1.5 ish higher than the salchow .... which is landed by many teams.... (compared to single skating where the 3a is given much higher base value than the lutz or other jumps... I would give salchows and toes the same BV here.... let's say 4.0, I'd gives the loop 5.0, the flip 6 and the lutz 7. or something along those lines..

2)as mentioned : teams here can compete and get points with their own strength, twist, and throws offer the possibility of the quads, the SBS jumps can really make an impact if harder jumps are landed well.... in singles, it's just jumps... so the risky elements are everywhere... I know that when I watch Pair A they will have great throw.... but Pair B may have better jumps...so in the end, the viewer is not left with a rush of adrenaline after the first missed element... I hope i make sense here.. i am falling asleep as I type... big dinner lol
2)

 

[moriel]

1 fall = capped at 9.

2 falls = capped at 8.5

3 falls = capped at 8

4 falls = capped at 7.5

And so on.

Nobody and I mean NOBODY should be getting 10s with a fall. Nobody should be getting 90+ for programs with major errors.

Something similar for pops please. NOBODY should be getting 10s with pops.

 

[Mathman]

Of course it's not only luck, but it certainly is a factor too. If you fall once out of 100 jumps, the luck is on when it happens as well. ...

I understand the point, but I do not think that this is how a professional or Olympic athlete looks at it. I think that if an athlete hits 99 in a row and then fails on the 100th, he will say that he did the 99 correctly and he did the 100th incorrectly. In other words, it is not bad luck that the miss occurred when it did, it was faulty technique on that attempt.

In general, I think it is not in the athlete's best interest to tell him that he had bad luck.

(The part mentioned by gkelly about hitting a spot of bad ice is well taken, though. And there could be other factors beyond the athlete's control, such as getting struck by a meteorite.)