Page 1 of 2 12 LastLast
Results 1 to 20 of 26
  1. #1

    Join Date
    Jul 2007
    Posts
    2,023
    vCash
    5550
    Rep Power
    7820

    How much should errors affect the performance & execution mark?

    This discussion from the Europeans forum was interesting and although it's been discussed before after certain competitions with controversial results, I thought warranted its own thread. Since I like doing mathematical experiments to translate skating into specific numbers on the PCS, I came up with a system below.

    Sorry that I went overboard with the quotes. I hope that doesn't violate a board rule but there were a lot of relevant posts.

    Quote Originally Posted by bek
    V/T and S/K had similar base values within a point. I can sort of understand them winning overall. But S/K only won that free skate by a point. The reason is V/T were given humungous PCS. Its not right IMO. Can I mention that V/T may be better skaters with bigger elements, but I actually do like S/K's choreography better?

    These type of PCS for messy messy isn't right.

    How V/T got 9.25 for P/E and S/K got much lower than that I'll never understand. Part of the fun of watching sports is to see who wins. For example in basketball you have to play well to win the game, and sometimes the lesser team wins because they were better that night. That's what makes it fun-you don't know the outcome.

    Now the outcome is crystal clear....

    I don't think one fall if everything else is superior should could a team out.. But multiple falls?

    When the audience can tell the major errors its ridiculous. And the thing is that S/K's program was extremely well executed according to the protocol sheet. Its not like you had underrotations and all kinds of crazy things. They should have won that free by a landside, if you were judging the skating and not the reputations.
    Quote Originally Posted by kosjenka
    It was overall fair that V&T won, but I will never understand how can PCS be 9s and high 8s with so many major mistakes...
    It is almost like we can have a prefixed official PCS that a skater or team will receive that season and just add up the TES from each competition. I dont think falls should be more seriously punished in TES than they are now, but how on earth can Performance / Execution mark be this high? Isnt performance something unique? Looking at PCS for V&T you would never say this was one of their worst performances ever...
    Quote Originally Posted by miffy
    ITA with the majority here, the biggest issue seems to be the PCS being barely affected by what happened.
    Quote Originally Posted by Corianna
    I don't understand why fans expect PCBs to be much affected by falls. If I understand correctly, falls get docked with the mandatory deduction and negative GOE - generally costing the skater 4 points. Usually falls don't affect edge quality, ice coverage, interpretation etc, and I don't think skating skills are defined by staying upright. Some would probably hurt choreography if the skater were down for a long time. This system was put in place so that falls would not be the end of the world, but that playing it safe would be, yet a lot of people continue to insist that falls are the only thing that count. Also under the old system, winning the free meant winning the competition. Now it's a two programme event. The conspiracy theorists tend to ignore the results of the short.

    The skating does look different now. Personally I like the current style, but I can understand why some people may not, and want to go back to the old system, but I don't understand why they cry" foul" when the new rules are followed. Agitating for a return to 6.0 is logical for these people, but the venting of outrage and repeated cries of foul are tiresome.
    Quote Originally Posted by bek
    Once again S/K had the same base value as V/T. Just about. So this was not an issue of one team playing it safe, and another team not playing it safe. Its an issue of one team being a mess and another not being a mess.

    And if you want to argue skating skills shouldn't be affected well okay, what about Performance/Execution. What about interpertation. Once again, why should one half of the mark be preset no matter what the skaters actually put on the ice.

    Not to mention one could argue in a lot of ways difficulty is not rewarded, who can affected Lepisto's bronze-given to her by PCS.

    I think its great skating skills matter more now than they did before. And I'm fine with that. But the actual performane they put on the ice should matter too.

    Nobody once again is saying they want one fall-and your out. But people don't want multiple falls either. And when you have two skaters doing a quad, and the person who falls on the quad getting 6 points, that means the person who landed it isn't getting as much advantage as they should. And if they are facing Patrick Chan, its little to no advantage because Chan's guaranteed once again huge PCS.

    Its not fun watching people fall multiple times in a program, and it shouldn't be rewarded. Nobody would claim if V/T fell once and still got high P/E marks etc. Its the multiple ocassions that people are complaining about.

    And falls can affect intepretation.

    There's also the issue of perception. I think you can explain to folks that one fall but the other skater is better. But one you get into multiple glaring errors- there is the perception that the competition is fixed.
    Quote Originally Posted by Hedwig
    Performance mark has to go down with many falls - it hurts the general performance even if (what as good as never happens) that skater let not affect the program by it.
    But usually they skate more dejected or more cautious after falls as well so the PE mark has to go down even further.

    IF they skate dejected as well, they usually skate slower and more cautious which in turn affects SS (but not by much, I agree there). IN mark can go down as well - again if the skaters skate differently after a fall.

    But this is what many argue - why not award 9.5 for skating skills but 5 for performance in a case like pairs?
    Or why not award a program like Amodio's Morozov programs with 8 for SS and 4 for interpretation? (I say IN and not CH because CH actually stand more for things like ice coverage, using the ice surface evenly etc and not for choreography as we use it)
    Quote Originally Posted by alchemy void
    Greatly prefer COP to 6.0. I think TES has been implemented better then I ever would have imagined.

    The problem is PCS. Skating skills are way over-valued because they always dictate the rest of the PCS. While I get SS are important, it's the peeformance that gets really under valued. If the skater has amazing skating skills, they are getting to get amazing PCS even with poor performances, empty programs, no expression. Likewise a skater with lesser skating skills but has a fabulous PERFORMANCE and execute really well will get screwed because their second rate skating skills put them into a PCS corridor.

    The paradox is that, to an average crowd, the performance and execution is the most obvious and exciting thing to see... when a skater absolutely kills it, the crowd goes wild.... but this is the least rewarded part of PCS. Skating skills are the least obvious and least exciting to see and it is by far the highest rewarded category. This makes for more WTF moments and much less incentive for skaters to perform and have interesting choreography.

    PCS categories need to be reworked, skating skills need to be part of TES to reduce their influence on PCS, and the PCS needs to be calculated on a 6.0 scale so the judge ends up giving one PCS score that can be announced like it in the 6.0 das. This will also force the judge to look at the whole of the program and compare the skaters' PCS against each other, instead of using a corridor that goes up by 1.5 points for each warm up group.
    Quote Originally Posted by kosjenka
    Going back to 6.0 is not an option. Not in any way.

    The conspiracy theory comes in when you study PSC and see that no matter how a skater or teams skates the all 5 scores are very much in synch.

    Once again - deductions for falls with mandatory -1 and negative GOE is perfectly fine but Interpretation, Performance/Execution need to be danged by multiple major errors. Falls interrupt the flow of the program. They damage the performance, falls mean that the execution of the program is poor. You cannot translate that with high 8s and 9s.

    From what i remember under 6.0 if you were in 4th place - no matter how close it was - you cannot win gold with winning LP unless the winner of SP does not end up in 3rd in LP.
    Quote Originally Posted by Corianna
    These are certainly legitimate arguments. I question though that relative unanimity across a panel, and multiple panels, is proof of something crooked or incompetent. I would take it as proof that the criteria as they exist are being interpreted correctly. The arguments for changing the criteria would only become stronger, surely. I have no problem with the way FS is judged now, but a clear, reasoned discussion such as many of the posters above have evinced, can only benefit the sport.
    Quote Originally Posted by bek
    But part of the reason why there's unamity is the corridor. Also, I could point out how much reputation based PCS are, its a problem when 1/2 of the score is preset. At that point, one could ask why are we even bothering have a competiton. Just video tape everyone in practice to get an idea of their skating skills-and call it a day.

    The problem becomes too how much it hurts let's say a no name skater. If you have a skater with absolutely no reputation with Patrick Chan skating skills-do you think the judges would automatically award said skater (especially if from a non skater country) 9s in PCS. NO. It would take quite a while.

    I realize this is a judged sport and so yes you have to earn a reputation. However, that's also why its important to at the very least start mandating mandatory deductions in P/E etc for messy programs.

    So at the very least your not going to have the double whammy of non rep skaters not getting the PCS they deserves, and also losing to messy programs from skaters with great reputations.
    Quote Originally Posted by topaz
    I strongly support CoP over the 6.0 system. It offers a clearer, consise idea of a skaters skills. I do think the PCS is manipulated by the judges, I think if we knew the names of judges would definitely help somewhat but in a way it may protect some judges who actually judge the skaters by the elements and not the skater accomplish ments.

    On the other hand, countries will still push and politic harder for the skaters they want the judges to recognize and appreciate. The judges are part of countries federation and they know which skaters their countries want to push internationally. Also, I think many of the judges are "old school", they have perferences for the type of programs they want to see and the music they feel is acceptable for the skaters.

    The falls from V/T did effect their program scores. I stated this in the pairs final thread, V/T received 2 points on an element they have received 7 points for when they perform it correctly. S/K may have simiar base value in TES, the execution of the elements(GOE) are not comparable to V/T.

    IJS does not mention in detail that the falls or the failure to execute TES elements matter in the juding P/E, IN. In some ways, I think some are still looking at P/E as an extension of the old 6.0 system in regards to performance.
    Quote Originally Posted by alchemy void
    Yeah there are some bad, old, corrupt, senile judges out there....but I think the main reason we get these PCS 'corridors' is due to the 5 PCS categories themselves for 24+ skaters. Even when we've done "You be the judge" exercises here on FSU, I have found myself using corridors just because it's damn near impossible to accurately judge 5 different PCS categories accurately and comparatively think about how you ranked Skater X in transitions vs. Skater Y, not to mention all of the responsibilities with GOE.
    Quote Originally Posted by Aussie Willy
    Yep is all very well to sit back in judgement of the judges but I think people's attitudes would change if they were actually sitting in those seats and having to make those decisions.
    Quote Originally Posted by peibeck
    I respect your opionion and know you have worked hard to become a judge, Aussie Willy. However, I do wonder whether you feel the marks are accurate/merited for V/T's free program and, if so, why? (I realize dance is your speciality.) What made their PCS "quality" (other than perhaps skating skills, which I can understand) garner a score 8% higher than their nearest competitors for a very, very weak skate, because it completely baffles me.
    Quote Originally Posted by Cherub721
    Here's the thing. S&K already have better CH, IN, and TR than V&T in the LP. S&K have lots of interesting transitions and dance moves in their program. V&T have a huge rest section in the middle where she kneels to the ice and prays while he skates around her. S&K have unexpected entries into their elements. They vary the placement of elements, like the 3sal at the end of the program. I spoke to someone who was there and she felt S&K had good speed and ice coverage. Interpretation may be subjective, but how can anyone deny that "The Addams Family" is a very specific program that is clearly choreographed for that music and timed to that music? V&T's choreo is more or less last year's layout set to different music. They hit those huge lifts on the "Jesus Christ Superstar" part but there's not a lot of nuance that goes into interpreting the music.

    Now, I can even accept that under the usual circumstances, V&T's skating skills and execution of their program is so strong that it can overcome that and they end up with higher PCS marks in all categories, because the judges mark on overall impression. V&T's speed, edges, the height and security of their elements, and their unison and extension, on a good skate, really do make them clear winners in any competition where S&S aren't skating. However, this was a very rough performance, and they had huge splat on 3 of 4 jumping passes and didn't have the usual attack. I'd rank K&S ahead on PE here too. So now V&T have only the SS mark, and their SS wasn't as strong as usual, and it should not be enough for them to win the PCS.I'm not advocating a return to 6.0 but one thing I did like is the accountability of having to give an ordinal, and to make a decision about who was better (and it wasn't anonymous until 2003). Some skaters did win with falls but in this particular competition I doubt many would have the guts to put V&T first, even if that upset their plans. Now you can say that V&T lost the LP here and so the result would have been the same under 6.0 (as S&K were 4th in the SP) but it's very troubling that V&T won the PCS by over 5 points, averaging 9s in most categories and an 8.93 in PE. The judges are using it as an overall impression mark, but the overall impression was bad! They only lost that segment because of the TES.

    It's like the judges are afraid to mark the PCS outside the corridor, so they just hope that TES takes care of it and if it doesn't, oh well. It seems like they don't want to get written up so they just go with the normal flow.
    Quote Originally Posted by gkelly
    I timed the disruptions in V&T's freeskate. Trankov's step out of the 3S broke the flow of the program for less than 2 seconds, closer to 1. Volosozhar bounced up immediately from the fall on the throw 3S, so that was also less than 2 seconds. Trankov did take more than a second to get up from falling on the 3T and then he skated a couple of strokes out of character to catch up with the choreography, so the total break in unison for that mistake was 4-5 seconds.

    I.e., the total disruption for three errors added up to less than 9 seconds out of a program that lasted for 276 seconds, or a little over 3% of the total time.

    I didn't really see other disruptions.

    [...]

    It's not that these things are preset, but rather that the judges look at all aspects of the whole program and reflect everything the skaters do according to all the PCS criteria.

    The Program Components Overview defines 7 as good, 8 as very good, 9 as superior, and 10 as outstanding.

    It also shows scores in the 7s and 8s as representing performances that meet the criteria for that component ~75% of the time, with the implication that programs that scores in the 9-10 range would meet the criteria for a larger percentage of the program time.

    Different observers (judges, fans, etc.) may have somewhat different mental definitions of what counts as good, very good, superior, or outstanding. And there isn't any written guideline for how to balance those levels of quality with the percent of time they're achieved.

    So, theoretically, what would we think is an appropriate score for a program that is generally Superior to Outstanding for 90% of its duration, maybe just Good for 7%, and Poor to Very Poor for 3%, vs. a program that is maybe Good to Very Good for 95% and Average to Above Average for 5%?

    [...]

    If I were judging these performances, based on what I saw in the youtube videos, I'd score V&T higher on PCS than S/K as well.

    To me, this wasn't a very very weak or messy performance; it was a very strong performance for well over 90% of its duration with significant -- but temporally brief -- errors/failures on three important elements.

    It might be interesting to go through one or both of these programs and compare notes on the negative, very negative, positive, and very positive moments and program-wide qualities we each take note of and discuss how we would each weight them to come up with one number for each of the program components. I'm sure we won't all see the same things and won't all agree on how to translate our perceptions into numbers or how to balance the positives and negatives. But it would be instructive to compare notes and find out what others saw that we missed.

    [...]

    It's easy to see falls, and it's easy to see the best moments of a performance we're rooting for, the weakest moments of a performance whose official scores we resent. But what else can we see and discuss?

    And I don't doubt that ISU pair judges will see pretty much everything we could come up with as a group as well as some other details that none of us noticed.

    Should we start a separate thread for that kind of nitty gritty and keep this one more broadly theoretical -- such as whether a rule change be appropriate that penalizes a second fall more than the first?
    Quote Originally Posted by snoopy
    This logic seems a little dubious to me because there can be varying degrees of disruption. “x seconds out of y” ignores the level of disruption – was it 9 seconds of being behind the music or 9 seconds of being on your butt? I don’t have a dog in this particular fight but 3 falls indicates, to me, bigger issues with the PCS that the “x seconds out of y” argument suggests.
    Quote Originally Posted by gkelly
    There were two falls and one step out.

    The falls themselves were quick. Volosozhar pretty much bounced right back up from her fall. I'd say there was no more than 3 seconds total in which anyone's butt was on the ice.

    Now, the amount of time that an observer spends thinking "fall, bad, that's the second mistake, this is messy, they can't win now" and other negative thoughts, and failing to mentally catch up with the recovered performance, may significantly exceed the actual duration of the disruption. Should the performance be penalized for the amount of time the skating was disrupted or the amount of time that the observer spends thinking about what happened earlier in the program and ignoring what's happening right now?

    Also, what about smaller negatives that are less disruptive but that might add up to a much larger amount of time and more
    significant sense of negativity to someone who is bothered by those details -- or not even noticed at all by someone who's mostly focused on large details and not the small ones.
    E.g., out of unison, far apart, on two feet, losing speed, scratching on toepicks, breaking at waist, reaching for hand hold, etc.

    And those might also be only a few seconds here or there for a high-level team, although they can add up. For a lower-level team, such weaknesses might characterize a significant percentage of program even in the absence of outright mistakes. So should a team with weaknesses throughout, fewer strengths, but no blatant errors score higher or lower in PCS than a team with many strengths throughout and a few noticeable errors?

    With 6.0, each judge had to decide for him-/herself how to compare each pair of teams. And often programs with a lot of strengths and a few obvious mistakes would come out ahead of programs with fewer or less-strong strengths and weaknesses or subtle errors.

    With IJS, judges just have to decide whether the overall Performance/Execution, Interpretation, were Average or Good or Superior for the program as a whole, adjusted down to reflect errors as applicable.

    The outcome is often the same, though -- a skater or team with more strengths and more mistakes often wins over a skater or team with fewer of both.
    Quote Originally Posted by RFOS
    Just curious, those of you saying V&T's program "isn't very good" or "fell completely flat," what kind of marks out of 10 would you give in the PCS to reflect those opinions? Ideally, please refer to the guidelines and address multiple specific aspects of the program component marks http://usfigureskating.org/content/J...compexplan.pdf). I feel like those opinions have to be rooted in disappointment in the performance because they are clearly capable of better, and perhaps anger or extreme emotional dissatisfaction with the results, but I'm guessing there were many, many positive qualities they still exhibited in spite of making several brief errors (I haven't had a chance to watch any of the pairs performances yet).

    Cherub721, I do appreciate your specific thoughts on the programs.

    One thing to remember is that even under 6.0, the marks weren't supposed to be directly dependent on each other, but often there was a much higher correlation than perhaps should have been the case, especially in the free skate where you'd often see the vast majority of skaters with marks within a couple of tenths of each other. (In the short program judges generally had a high correlation between the required elements BASE mark and the presentation mark, with specific mandatory deductions taken from that base required elements mark, so if you added the deducted points back to the required elements mark you'd have a very high correlation). IJS by its design allows judges to more easily judge the presentation/program components from the technical mark. I think this is a positive of IJS in general, but of course it doesn't mean I agree with the marks given in every specific case.

    (One reason the 6.0 marks might have been more highly correlated than they should have been is that it is easier to determine ordinals when the two marks aren't vastly different for most of the skaters).
    Quote Originally Posted by bek
    [...] To me it doesn't matter how long a fall is, a fall leaves a strong level of impression on the audience and on the skaters (look at V/T's face on the medal ceremony do you think they felt that the performance was worthy of a 9? No.

    I can forgive one fall, but two and a big step out. No. It makes it look like the skaters cannot control themselves or their skates. Its a glaring error IMO.
    Quote Originally Posted by RFOS
    I think there's something to be said for your point, because falls and major errors do mar the overall impression to the extent that the skater and audience are going to be less satisfied. I would say that a 3% reduction in P/E mark would definitely not be enough (that would probably only be 0.25) even if the program was presented as well as it would've been without the falls aside from that 3% of the time. However, arguing that the mark should be way, WAY lower (like 50% just as an example) just because of that 3% of the time takes it to the opposite extreme. I've made the argument in another thread that that 3% probably does weigh more heavily in one's mind than the other 97% in the overal impression and therefore might be weighted more than 3%, but it shouldn't be weighted SO heavily as to cancel out the other 97% of time either. Would you be satisfied if they had gotten an 8 instead of almost a 9 for performance? A 7.5? How low do you think it should have been?
    As an experiment, I tried using a GOE/PCS conversion I came up with for my transitions system to come up with a "performance/execution of elements" mark.

    For my transition system, I originally had
    "Outstanding" = 10.00 = +3
    "Excellent" = 8.00 = +2
    "Good" = 6.00 = +1
    "Fair" = 4.00 = 0
    "Poor" = 2.00 = -1 or -2
    "Failed" = 0.00 = -3

    However, for both purposes, I just added "Weak" to be 3.00 or -1 and changed "Poor" to 2.00 or -2.

    One might argue that it should by symmetrical with the 0 GOE equaling a 5 for PCS but in practice I'd expect a skater with 5.00-level PCS to be able to get a mix of 0s and +1s. If everything was merely just completed to a 0 standard that to me is a little lower than 5 (Juveniles could certainly commonly do that), so 4 seemed appropriate to me. I think of below 2.00 for PCS as being painful, so -2 being not-quite-painful but still pretty seriously flawed sort of made sense to me too. A -3 error would be a blip where the performance would almost certainly be significantly marred and they probably wouldn't meet any P&E criteria while actually falling. (An exception where a -3 GOE wouldn't necessarily mar the program would be doing fewer revolutions than required on a jump in the SP. While the skater could still executed the lower jump well, the "ding" to the P&E marked could be due to the emotional disappointment incurred by the judge having to punch in a -3 and because it was a "failure" in a sense.)

    At their best, V&T are capable of +2s or +3 on every element, which could correspond to around a 9 if there was an even mix.

    I used that table to convert each of the unfactored average GOEs for each element into PCS with this formula:
    X = GOE
    Y = PCS equivalent
    3 < X < -2, Y = 2*X + 6
    -2 < X < 0, Y = X + 4
    0 < X < 3, Y = 2*X + 4

    I counted all elements equally. I suppose I would consider an element that received no value (and thus officially no GOE since they would just show up as dashes) due to a complete failure as a 0 also, whereas something that was asterisked due to being over the maximum could still be well-executed and I'd use the GOE I would've given it.

    Using that method the "performance & execution of elements" score for V&T was about 6.571. This would include the performance & execution of the elements themselves and perhaps 1.5-2 seconds before and after each element, enough time to quickly recover from a fall. (I wonder what percentage of the average program this would actually work out to be). Using a similar rationale to my transition system, this could be averaged with a "performance & execution between the elements" score. This would be affected by falls between elements or falls on elements that the skater took longer than 1.5-2 seconds to recover from. For argument's sake, let's say this was a 9.00 for V&T. That would average to a P&E mark of 7.786, rounding to 7.75.

    Because V&T had errors on 3 out of 12 elements (25%), and the "performance and execution of elements" is 50% in this experiment, essentially 12.5% of that mark was affected due to the errors (perhaps slightly more if they took time to get back into the program after one of the elements). This seems like a potential middle ground where it affects the mark more than, say, 3%, but certainly doesn't cut it in half or anything extremely drastic like that.

    It does make sense that the *execution* of elements would be an important part of the performance & *execution* mark, not just such that "not clean" programs are given lower scores than "clean" programs, but because the qualities that would be needed to get +2 and +3s or that would be lacking if a skater got -2s and -3s would also in most cases somehow relate to the P&E component criteria, even if the P&E component description doesn't explicitly mention anything about the execution of technical elements. For example, a +2 or +3 jump would almost certainly have to have a really good air position and presentation of the landing (going into clarity of movement and projection and adding a degree of emotional excitement in the viewer), similar for a +2 or +3 spin, lift, step sequence, etc. Additionally, the elements take up quite a bit of time in the program and that time has a major impact on the overall impression conveyed to the viewer so that time should be reflected in the mark also.

    I suppose that rationale could be applied to some degree to all components except for transitions (since that component quite specifically deals with only the connections BETWEEN elements), but I think the other components would be much less affected by the elements themselves. For example, when the skater is actually in the air for a jump they aren't going to be demonstrating anything in the criteria for skating skills so that time would have to be discounted. Falling on a jump (or other element, or between elements) would to at least an extent show a lack of sureness and should probably be reflected a little bit in the skating skills mark. I would argue that spins really wouldn't have much affect on the skating skills mark generally, but a footwork or choreographic sequence definitely is a place where a skater would demonstrate basic skating skills (and skills in all of the components except for transitions).

  2. #2

    Join Date
    Sep 2003
    Age
    52
    Posts
    10,239
    vCash
    500
    Rep Power
    10899
    Interesting.

    Would the goal be to generate rules or guidelines for how judges should reflect element quality and/or visible errors in the PCS?

    Should the guidelines be tailored differently specifically for each component according to its criteria?

    Is the idea that the 1.0 fall deduction is not enough and doesn't apply to non-fall disruptions, so it's better to build a more systematic way of lowering the component scores rather than a systematic way of further lowering the total segment score?

  3. #3

    Join Date
    Nov 2005
    Posts
    16,929
    vCash
    1600
    Rep Power
    4720
    To me, I know I'm one of the loudest on this, but at this point I'm just waiting for a new scandal to come through. However, I suspect that top skaters won't be allowed to fall multiple times and get huge PCS. Showing the hypocrisy that its allowed now. But if Trankovinflation chanflation, etc gets to be introduced to the general public. it might be interesting to watch, people denigrate the sport even further, because they will. Performance/Execution=9s.

  4. #4

    Join Date
    Dec 2002
    Location
    Infected with the joy of skating!!
    Posts
    10,391
    vCash
    500
    Rep Power
    7460
    The problem seems to be caused by the inability to give a skater an 8 in a category they deserve it in, and a 5 for P/E when they deserve a 5.
    Keeper of Nathalie Pechelat's bitchface.

  5. #5

    Join Date
    Mar 2010
    Posts
    3,466
    vCash
    500
    Rep Power
    1553
    I think some 'fans' assume if skaters fall they should turn the music off and leave the ice immediately and average skaters should win because they remained upright even if they weren't the best skaters.

    That certainly isn't the reason the COP was created. V/T skated wonderfully - even with the few seconds of imperfection the rest was far better than anything else anything the other pairs came up with. You may 'like' S/K better but personally feelings aren't on the score sheet and unless they combine the sp and fs skaters have to be equally good in each. If not the score difference from the sp is enough to win if a pair doesn’t' have a good enough fs.

  6. #6

    Join Date
    Nov 2005
    Posts
    16,929
    vCash
    1600
    Rep Power
    4720
    Quote Originally Posted by julieann View Post
    I think some 'fans' assume if skaters fall they should turn the music off and leave the ice immediately and average skaters should win because they remained upright even if they weren't the best skaters.

    That certainly isn't the reason the COP was created. V/T skated wonderfully - even with the few seconds of imperfection the rest was far better than anything else anything the other pairs came up with. You may 'like' S/K better but personally feelings aren't on the score sheet and unless they combine the sp and fs skaters have to be equally good in each. If not the score difference from the sp is enough to win if a pair doesn’t' have a good enough fs.
    Once again, I never said I normally like S/K better. I don't necessarily have a problem with V/T's overall win. However, I think 9s in P/E for that kind of program is to much. And you know what sometimes in other sports the average player, the average team etc wins a game or two? Why would that be so horrific if it happened in skating?

    I don't have a problem with one fall, but we aren't talking about one fall.

  7. #7

    Join Date
    Nov 2002
    Location
    Germany
    Posts
    136
    vCash
    500
    Rep Power
    0
    I think I suggested this already during the aftermath of Chan's last world's win, but to my mind the simplest way of having execution scores more clearly reflect what happens on the ice is to separate "performance" from "execution". I would lump performance together with the interpretation component in order to have one component dedicated entirely to execution. And to make things easy on the judges, I would require that the maximum score a skater can earn reflect the number of falls in the skate, i.e., 1 fall = maximum score of 9, 2 falls= maximum 8, etc. Other visible errors such as stumbles, step outs, hands down, etc., should be penalized a minimum of .25, depending on the severity at the discretion of the judges. Under this system, V&T at Euros, as an example, could get a maximum execution score of 7.5, which to me seems about right. I think my system is pretty similar to RFOS's in intent, just with much more basic math.

  8. #8

    Join Date
    Jun 2002
    Posts
    776
    vCash
    500
    Rep Power
    697
    Quote Originally Posted by julieann View Post
    V/T skated wonderfully - even with the few seconds of imperfection the rest was far better than anything else anything the other pairs came up with.
    You can't be serious. I think even Max and Tanja would be embarrassed to read anyone describing what they did on Sunday as "wonderful."

    Beyond V/T's mistakes (and Trankov sulking and taking 2 seconds to snarl and wipe his hands after falling on his 3toe, then breaking character and looking visibly disappointed after Tanja fell on the throw) the JCS program is empty.

    S/K have really excellent basic skating too - maybe an 8.75 or 9.00 to a V/T 9.25 to 9.50. After that it was all S/K - they had far more transitions and choreography in their Adams' Family program, they vary their speed and movements to match the music, not to mention they never broke character and presented up and out to the audience with commitment and expression from beginning to end, particularly Ksenia who rivaled an ice dancer in her commitment to expressing the character. They demonstrated excellent unison with leg and arm lines that really matched even throughout some nonstop choreographic nuances and have good posture and lines overall (to nitpick, I do wish Ksenia would point her toe more.)

    V/T's JCS is a lot of simple (albeit excellent quality) angsty pretty skating. Lots of plain crossovers into the throws and lifts, a long smoke break after the SBS spins. There are no complex in-betweens, no sections where their unison is challenged, very little variation in speed or character and really nothing to challenge them in terms of expressing the music or character other than being angsty on one level throughout. They can do much better and more challenging transition, choreography and character work (as their SP proves) but they don't bother. The marks should reflect what they do on the ice not their potential.

    This has nothing to do with "liking" one team over the other. I like V/T better but I see all the shortcomings of that free skate and when they don't deliver it to perfection they should be vulnerable to the other 3-4 top teams in the world like S/S, Pang/Tong, and yes, S/K.

    Quote Originally Posted by julieann View Post
    I think some 'fans' assume if skaters fall they should turn the music off and leave the ice immediately and average skaters should win because they remained upright even if they weren't the best skaters.
    Please. With 3 large mistakes should V&T still beat, for example, the Israeli couple? Of course they should! And they still should even beat James/Cipres and Berton/Hotarek who are very good but still not on the same quality level. But S/K have gotten very close to V/T basic quality level and delivered a well designed and very challenging program close to perfection (the only real error being a pretty well covered up lift exit mistake).

  9. #9
    Registered User
    Join Date
    Oct 2004
    Age
    31
    Posts
    1,441
    vCash
    500
    Rep Power
    0
    Quote Originally Posted by Coco View Post
    The problem seems to be caused by the inability to give a skater an 8 in a category they deserve it in, and a 5 for P/E when they deserve a 5.
    I agree and I brought it up a while ago in another thread and someone responded (it might have been Eislauffan but I'm not sure) that judges are encouraged to keep the PCS score within a certain range. It would explain why they never waver much but it so doesn't make sense. Having excellent skating skills has nothing to do with having excellent interpretation, so if the interpretation sucks, judges should be allowed to give a 4 even though they might have given a 9 for skating skills because the skating skills deserved it. A fall also has nothing to do with overall skating skills or interpretation but needs to be reflected in the P/E mark.
    It's just so out of this world to keep the PCS scores within a certain range!

  10. #10

    Join Date
    Sep 2003
    Age
    52
    Posts
    10,239
    vCash
    500
    Rep Power
    10899
    Quote Originally Posted by ballettmaus View Post
    I agree and I brought it up a while ago in another thread and someone responded (it might have been Eislauffan but I'm not sure) that judges are encouraged to keep the PCS score within a certain range.
    Officially, de jure, that's not true. The rules for determining whether judges are out of line on PCS give them more leeway to disagree with the panel if they use a wider range between components for the same skater than if they keep them close together.

    De facto, it seems that many judges are afraid to trust those calculations and may believe that they're more likely to stay in the corridor of the rest of the panel if they keep all their marks close to the Skating Skills mark for that skater. And then it becomes a self-fulfilling prophecy -- if all the judges aim to match each other by using tight ranges, then the ranges remain tight and it becomes riskier for judges to spread the marks as intended.

    My understanding is that a lot of the more recent judge training tries to encourage judges to spread the marks. But more conservative judges may still feel that they need to aim for staying in the middle of the range of the panel. Similarly to the way that under 6.0 judges tended to keep the presentation marks within a couple tenths up or down from the technical merit mark (or SP pre-deduction required elements base mark) even if the skater was significantly better at either technical merit or presentation.

  11. #11

    Join Date
    Mar 2010
    Posts
    3,466
    vCash
    500
    Rep Power
    1553
    Quote Originally Posted by BreakfastClub View Post
    …the JCS program is empty…V/T's JCS is a lot of simple (albeit excellent quality) angsty pretty skating. Lots of plain crossovers into the throws and lifts, a long smoke break after the SBS spins. There are no complex in-betweens, no sections where their unison is challenged, very little variation in speed or character and really nothing to challenge them in terms of expressing the music or character other than being angsty on one level throughout. They can do much better and more challenging transition, choreography and character work (as their SP proves) but they don't bother. The marks should reflect what they do on the ice not their potential.
    Judges can only score what they see, not what the potential is, you said it yourself. They can also only score what is on the score sheet and they hit all the boxes, they can’t be faulted for knowing the sheet.

    This has nothing to do with "liking" one team over the other. I like V/T better but I see all the shortcomings of that free skate and when they don't deliver it to perfection they should be vulnerable to the other 3-4 top teams in the world like S/S, Pang/Tong, and yes, S/K.
    Again, what shortcomings are you seeing? Judge what you see not what you don’t see. Otherwise you are doing the skaters a tremendous disservice. There is no reason that a few mistakes that took a few seconds to “perform” mar the rest of the “program”. No different than if crappy skaters do a really good elements and don’t get rewarded with +2 and +3.

    With 3 large mistakes should V&T still beat, for example, the Israeli couple? Of course they should! And they still should even beat James/Cipres and Berton/Hotarek who are very good but still not on the same quality level. But S/K have gotten very close to V/T basic quality level and delivered a well designed and very challenging program close to perfection (the only real error being a pretty well covered up lift exit mistake).
    So you excuse S/K "pretty well covered up" error but not any of V/T's? S/K still have problems with their skating and only just started doing the triple twists and it's no where near V/T's height (or B/L) so stop putting them on a pedestal they don’t deserve, not yet. Klimov is okay to watch on a good day and if it wasn’t for her I may not watch them at all. He is the Yuri of their team. She is wonderful and has great landings on the throws, their jumps are perfect but lifts labored, spins are ok and the footwork is alright. By 2018 they can be OGM but they need to be a coaches #1 team and right now they aren’t.

    Maybe they could have won if they had got full levels for their elements, but they didn’t, that is a big aspect in skating and they didn’t fulfill that criteria so no matter how well you ‘thought’ they skated or how bad you thought V/T should have been penalized the levels were in the hands of the skaters and they dropped the ball. They didn’t perform as well as they could have in the sp, they didn’t get full levels, their skating is not on par with V/T and got silver.

    To look at the competition as a whole as think S/K should have won is laughable.

  12. #12

    Join Date
    Jun 2002
    Posts
    776
    vCash
    500
    Rep Power
    697
    Quote Originally Posted by julieann View Post
    Again, what shortcomings are you seeing? Judge what you see not what you don’t see ... (of S/K) stop putting them on a pedestal they don’t deserve... Otherwise you are doing the skaters a tremendous disservice … To look at the competition as a whole as think S/K should have won is laughable.
    Excuse me? Please re-read my post. I laid out pretty clearly where V/T's JCS program is empty in terms of transitions, choreography and interpretation and in no way challenges them to skate to their incredible potential. And I didn't put S/K on a pedestal or say they should have won the entire thing. I laid out where they have a much more challenging free program designed exactly with an eye towards COP's PCS standards and nailed it with one very minor error. V/T are better skaters but the JCS program is not.

    Quote Originally Posted by julieann View Post
    So you excuse S/K "pretty well covered up" error but not any of V/T's?
    Correct, there's nothing to excuse. S/K's error was minor and didn't break the program at all. All three of V/T's errors were disruptive and Trankov made it worse with his sulking after a couple of them.

    LOVE V/T but they need to skate clean with that conservatively constructed free skate. Otherwise they need a better one that challenges them to skate up to their immense talent just like their magnificent SP does, not this cautious one that opens the door to other teams (albeit only very top challengers) as soon as they start screwing up in a significant way.

    I've said my peace and I'm done.

  13. #13

    Join Date
    Nov 2005
    Posts
    16,929
    vCash
    1600
    Rep Power
    4720
    I could barely notice the error on S/K's lift to compare that to falls is laughable. As for levels that was reflected in base marks. The issue here is PCS for V/T.

  14. #14
    Registered User
    Join Date
    Oct 2002
    Location
    LEAVE EDMUNDS ALONE!!1!
    Posts
    20,089
    vCash
    500
    Rep Power
    0
    I think that the issue here is people struggling with ambiguity and integrating negative and positive aspects together.

    The line of thinking that I keep seeing time and time again is: A skater makes a mistake > bad performance > so low(er) marks

    But under IJS, you are not supposed to give away one mark for 'general impression.' That's what the judges essentially are doing, as a result of the above described thought process, but what they are supposed to do is reward the execution of individual elements and how well the skater/team fulfills the PCS criteria.

    A skater can make mistakes and then get rattled or disoriented by it and either give up or become very tense and the performance aspects suffers as a result. In this case, you should obviously reflect that in the PE and other relevant scores.

    But a skater can continue to keep on fighting and presenting their program to the best of their ability despite the mistakes. In that case, I see absolutely no reason to lower their PE score (if a skater falls but gets up straight away, it's even hard to argue that the flow of the program was disrupted).

    But it seems that a noticeable mistake anchors people's judgement and becomes a lens they see everything else through. Come on, you can do better than that.

  15. #15
    Blergh
    Join Date
    Jul 2003
    Location
    I Want to Go to There
    Posts
    9,309
    vCash
    500
    Rep Power
    17257
    Ziggy, you seem to be able to view skating as a collection of individual and separate elements and enjoy viewing skating that way. However, I think that I and other viewers like me think the whole is very important. Although I do agree that some falls are much more disruptive than others, I do think there needs to be something the judges can do to deduct for general impression when there's a performance with multiple falls (at least more than they are doing now). I don't care how fast a skater gets up because 2 or 3 falls is disruptive (even if the way the skater handles such mistakes can be less disruptive than it has to be). I know IJS is supposed to be rewarding individual elements and letting the numbers fall where they may, but I also think the overall product has to be rewarded as well.

    I don't think the whole picture should be dismissed just because skating is now looking at all the individual elements with a critical lens. I really do not have the answer as to how it can be done (and I also don't know how to correctly balance the demand for skaters to attempt high risk elements/choreography but not encouraging them to fill it up with things they really cannot handle (and instead encourage them to execute elements within their control…even the riskier ones)).
    "Corporation, n. An ingenious device for obtaining individual profit without individual responsibility." - Ambrose Bierce

  16. #16

    Join Date
    Mar 2002
    Location
    with the traditionless
    Posts
    5,439
    vCash
    500
    Rep Power
    4854
    I agree with comments made by VIETgrlTerifa and bek.

    1) skating is IMO inherently a performance sport and the general impression, or execution of the performance should matter in scoring.
    2) I am okay with an average skater beating a superior skater if the superior skater didn't compete well that day. Otherwise why have competitions? Just have a bunch of test skates each year and hand out the medals accordingly.
    3) I don't think one fall should tank a skater but perhaps what I would do is make the penalty for falls scale up. Meaning, 1 point deduction for 1st fall, 2 points deduction for 2nd fall, 3 points deduction for third fall, etc.
    Figure skating is hard.

  17. #17

    Join Date
    Jul 2009
    Posts
    4,373
    vCash
    500
    Rep Power
    0
    Quote Originally Posted by julieann View Post
    V/T skated wonderfully - even with the few seconds of imperfection the rest was far better than anything else anything the other pairs came up with. You may 'like' S/K better but personally feelings aren't on the score sheet and unless they combine the sp and fs skaters have to be equally good in each. If not the score difference from the sp is enough to win if a pair doesn’t' have a good enough fs.
    Are you sure you watched V/T's Euros skate and not one of the GP ones? They did not skate "wonderfully" and the program was nothing special to begin with. S/K were much better on the day and the marks should have reflected it to a greater extent.

    I can see why V/T get high marks for their SP, but the JCS LP doesn't warrant high marks for anything other than skating skills and P&E, and looking at how they performed it on Sunday - not just the falls but the actual performance, expression and projection - their P&E mark should have taken a hit. But of course, all marks follow from skating skills, because as we all know skating skills magically imbue skaters with great interpretation skills and good choreography (the rolleyes is for the ISU, not for you).

    Re P&E in general, it should be affected by falls only if they affect the rest of the performance and the criteria listed for this component. Falls should be penalized with deductions, not necessarily with a PCS hit. I would, however, welcome some kind of general impression, more than the sum of its parts component or score.

  18. #18

    Join Date
    Jul 2007
    Posts
    2,023
    vCash
    5550
    Rep Power
    7820
    Quote Originally Posted by Ziggy View Post
    I think that the issue here is people struggling with ambiguity and integrating negative and positive aspects together.

    The line of thinking that I keep seeing time and time again is: A skater makes a mistake > bad performance > so low(er) marks

    But under IJS, you are not supposed to give away one mark for 'general impression.' That's what the judges essentially are doing, as a result of the above described thought process, but what they are supposed to do is reward the execution of individual elements and how well the skater/team fulfills the PCS criteria.

    A skater can make mistakes and then get rattled or disoriented by it and either give up or become very tense and the performance aspects suffers as a result. In this case, you should obviously reflect that in the PE and other relevant scores.

    But a skater can continue to keep on fighting and presenting their program to the best of their ability despite the mistakes. In that case, I see absolutely no reason to lower their PE score (if a skater falls but gets up straight away, it's even hard to argue that the flow of the program was disrupted).

    But it seems that a noticeable mistake anchors people's judgement and becomes a lens they see everything else through. Come on, you can do better than that.
    I do think errors should affect the P&E mark *to some extent* because at least during the brief time skaters are making the errors the criteria aren't going to be fulfilled. However, the above was just an experiment and after finally actually watching V&T's performance I wouldn't ding them as much as that exercise suggested. I personally feel that it was presented well other than the disrupted 9 seconds or so (I did find S&K's choreography more interesting in its details and would give them higher on that mark but didn't feel they "sold it" as much as they could have), and the quality of the amazing elements they did do actually stood out more than the mistakes. The amazing air and landing positions on the twist and throw loop, amazing unison on the side-by-side spins, etc., enhanced their fulfillment of the P&E criteria more than the errors disrupted it IMO.

    You're definitely right that people, especially judges, should be careful not to let an error in one element affect their overall judgment of other aspects of the program that weren't affected by the error. Unfortunately, that's something that casual fans definitely aren't good at and if they see programs with many obvious errors beat a program with few or no obviously visible errors they get the impression that it's fixed. Many experts feel strongly too and those results aren't helping win fans for the sport. Part of it may be carryover from the 6.0 system where the presentation mark usually did go down noticeably if the technical mark went down significantly, and there was a higher correlation between the marks than there is in IJS and higher than there theoretically should have been if they had used the marks independently.

  19. #19

    Join Date
    Sep 2003
    Age
    52
    Posts
    10,239
    vCash
    500
    Rep Power
    10899
    Quote Originally Posted by VIETgrlTerifa View Post
    Ziggy, you seem to be able to view skating as a collection of individual and separate elements and enjoy viewing skating that way. However, I think that I and other viewers like me think the whole is very important.
    The question is whether the sport wants an overall sense of the whole to be more important than the sum of the parts as much as the fans want that.

    I think that to some extent they do, but the current rules/values don't require it. So it probably would be a good idea to add some additional rules that will allow and encourage scores to go down more when errors affect the overall impression.

    Even so, I don't think anyone wants an otherwise strong performance with a few obvious errors to lose to an otherwise easier or weaker performance with fewer, smaller, or even no actual errors. (Fewer errors doesn't automatically equal stronger.)

    As long as judges tend to see more differences in overall quality and mark accordingly, and others may not see as many quality differences but do see obvious errors as the most salient aspects of a performance, there will be discrepancies between who the judges thought skated best overall and who the others thought was best overall.

    So we're never going to get rid of the problem entirely. Even if the rules build in more penalties for multiple obvious errors and judges are trained to apply those penalties strictly, so that the best skaters start losing more often once they get past two obvious mistakes, and all the knowledgeable fans and coaches and other skaters start agreeing with the overall results more often, there would still occasionally be times when all the good skaters make mistakes, only a mediocre skater gives a superficially "clean" performance, and casual viewers would still be confused over why that clean performance didn't win.

    However, stricter rules would make winning with multiple errors less common, so some rules in that direction would probably be a good thing.

    Although I do agree that some falls are much more disruptive than others, I do think there needs to be something the judges can do to deduct for general impression when there's a performance with multiple falls (at least more than they are doing now).
    Quote Originally Posted by snoopy View Post
    1) skating is IMO inherently a performance sport and the general impression, or execution of the performance should matter in scoring.
    2) I am okay with an average skater beating a superior skater if the superior skater didn't compete well that day. Otherwise why have competitions? Just have a bunch of test skates each year and hand out the medals accordingly.
    3) I don't think one fall should tank a skater but perhaps what I would do is make the penalty for falls scale up. Meaning, 1 point deduction for 1st fall, 2 points deduction for 2nd fall, 3 points deduction for third fall, etc.
    So some penalties built in as automatic, and additional penalties left to the discretion of judges depending on the severity of the disruptions?

    For the automatic penalties, we already have the fall deduction. 1 point is not really significant at the top levels but very significant at lower levels -- including seniors who might be fighting for a place in the long program. So when this topic came up a year or two ago mostly focused on the men's event, after thinking it through I suggested that it might work to 1) change the fall deduction from a flat 1.0 points for all falls in all programs at all levels to, instead, a percentage of the total segment score (which would tend to make them more costly in freeskates as in short programs) and also 2) increase the penalties for subsequent falls. So the deduction for the first fall could be 1% of the total score, for the second fall 2%, etc.

    Of course not all disruptive errors are falls and not all falls are especially disruptive. And not all disruptive errors occur on elements scored under TES. But to capture those disruptions on a discretionary basis, guidelines could encourage judges to reduce PCS in response to falls and other disruptive errors. It wouldn't be automatic as with the fall deduction, but if the guidelines are written into the rules, judges would take them more often in addition to the increased automatic fall deductions.

    Maybe written guidelines suggesting that for each fall or other break in the program, judges should reduce the Performance/Execution mark by at least 0.25 and up to 1.0 per occurrence depending on severity, and also take off from the other component marks as applicable.

  20. #20

    Join Date
    Jan 2012
    Location
    Oz
    Posts
    2,551
    vCash
    400
    Rep Power
    6051
    Great debate. Learning a lot. Keep it up.
    Prosperity makes friends, adversity tries them. – Publilius Syrus

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •