1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Hypothetical question on TR and SS

Discussion in 'The Trash Can' started by midori, Nov 21, 2010.

  1. midori

    midori Well-Known Member

    I am getting confused how TR and SS should be scored, so please help me with scoring some extreme hypothetical cases.

    Assuming the following list shows how a same skater performed in 6 different competitions. Assume her major skills stayed same, avarage difficulty of transion moves stayed same. Judges were consistent and fair. Even other competitors and their performances stayed same.

    Only differences between the competitions are her elements and amount of the transitions, hence the variety of transition moves, as she does not repeat the same move during a program.

    She did not stop for taking a breath during any programs.

    - Competition A:
    All single jumps(intentional), level 1 elements(intentional). Executed all 12 elements in a well balanced order with some transition between each element. All elements got GOE 0.
    Total time of the transitions were 2min 0sec.
    Assume her SS and TR were both 4.0 in this competition A.

    - Competition B:
    The same elements as Comp A but very frontloaded. Finished all jumps without any other elements nor transitions between.
    Total time of the transitions were again 2min 0sec, all after the jumps were finished.

    - Competition C:
    Only valid element was 1A at the very end of the program. The GOE was 0. She did not even attempt any other element than the 1A but kept doing transition moves.
    Total time of the transitions were 4min 0sec.

    - Competition D:
    All triples and level 4 and the GOEs were 0. The number of elements and the balance were the same as Comp. A.
    She gets faster during entrance of elements, but during the transitions her speed stays same as Comp. A.
    Total time of the transitions were only 1min 0sec because difficult elements took more time.

    - Competition E:
    The same elements as Comp A but none of them was intentional and fell on every element. Singled all jumps planned triples, and received all level 1 for elements all planned level4, GOE -3. She recovered from each fall very quickly and did not miss any transition moves.
    Total time of the transitions were 2min 0sec.

    - Competition F:
    Almost the same as Comp A but executed the elements very well with difficult entrance and all, then received all GOE +3.
    Total time of the transitions were 2min 0sec.

    How should her scores for TR and SS be in each competition?

    Apparently the skater is major headcase for fans and coaches, but I believe that does not affect the scores ;)

    ETA: add bold to some assumptions.
    Last edited: Nov 21, 2010
  2. Aussie Willy

    Aussie Willy Hates both vegemite and peanut butter

    I am sorry but I don't understand what you are asking. How can anyone comment without seeing the skater and then be asked to give a score? Because transitions do not just depend on quantity but also the quality and difficulty.

    To just say a program is full of transitions does not paint the picture.
  3. midori

    midori Well-Known Member

    Please assume the overall difficulty and quality do not differ. Only the quantity differs between the above programs.

    e.g. In Comp A, B, E, F, her transition moves consist exactly the same 2 min total, and she executed in the same quality, as if copy+pasted.
    In Comp D, she did only half of them because she took more time for elements but the average difficulty and quality of transition was the same as Comp A,B,E, and F. Just with less quantity and hence less variety.

    TR4.0 at Comp. A is given to suggest how difficult and how good it is. If the number 4.0 is not right then you could use any number X for the reference score at competition A. Or just "much higher than Comp A" or "same as Comp A" would be good.

    The question is all hypothetical because real programs have too many noises to evaluate. Real skaters cannot repeat exactly same moves in two competitions :)

    Actually this question is here inspired by your post, Aussie. I remember you once (or more) explained that if a program was all filled with transitions it should get 10 and if half of if it gets 5, or something like that. (I think it was you, wasn't it?)

    If the assumptions above are not enough, what kind of assumptions do you think we need?

  4. gkelly

    gkelly Well-Known Member

    I'll try to get specific later when I have more time.

    One of the criteria for Transitions is Intricacy, which involves how closely the elements are connected to each other or to transitional moves (lack of telegraphing). Another criterion is Difficulty, which includes not only the difficulty of the transition move, but also the difficulty of how it's connected to an element (if any), which includes the difficulty of the element.

    Example, back outside counter in isolation, back outside counter into single axel or double axel or triple axel.

    The turn itself adds to the difficulty level of the connecting steps. Doing it directly into an element also adds to the intricacy. Doing a difficult turn directly into a very difficult element such as triple axel adds a whole lot more difficulty than doing the same turn in isolation.
  5. Aussie Willy

    Aussie Willy Hates both vegemite and peanut butter

    :lol: I didn't realise I was so inspirational. I will take that as a compliment. Thanks.
  6. Doubletoe

    Doubletoe Well-Known Member

    Transitions are scored not just in terms of difficulty and quantity, but also in quality of execution, so simple transitions executed smoothly and with speed can give a skater a higher transitions score than more difficult transitions executed with difficulty. This includes transitions in and out of jumps and spins. Let's say the same skater skated the same exact program with the same choreographed transitions in-between elements but in the first competition she fell on every jump and in the second competition she landed every jump solidly. In the program with the landed jumps, the transitions score should be higher because the transitions out of the landed jumps would be seamless and controlled. It is impossible to have a seamless, controlled transition out of a jump landing when you have fallen on the jump.
    Last edited: Nov 23, 2010
  7. dinakt

    dinakt Well-Known Member

    I would be very grateful to a person who attempts an objective analysis of TR of particular skaters, for ex. Chan, Kozuka, Abbott and Takahashi; and the underlying similarities and differences.
  8. AliasJohnDoe

    AliasJohnDoe Headcase Addict

    I actually like these discussions. I've read the definitions of each component of PCS, but these discussions help me understand it that much better. Thanks.

    I'm assuming Sawyer's cantilever is an example of a Transition with good quality and difficulty then.?
    Last edited: Nov 23, 2010
  9. Aussie Willy

    Aussie Willy Hates both vegemite and peanut butter

    I love Cantlievers and I would definately classify them as a difficult transition. Not many skaters can do them and do them well.
  10. gkelly

    gkelly Well-Known Member

    OK, I'll go through and give some thoughts on the transitions I see. Short programs, because they're shorter. :)

    Takahashi 2010 NHK first:

    Various poses with upper body moves
    Back crossover into BI and FO (i.e., counterrotated) mohawks, cross to RBI counter

    Picks up speed with various simple turns (threes and mohawks, including LBO three with nice flow on the exit) and three crossovers in between, mohawk into flip combination

    Forward crossovers to deep LFI choctaw; second edge not held -- immediate crossover to RBO extended edge to RBO counter

    Mohawks and crossovers with some upper body movement into simple BO approach to triple axel (bad landing here)

    Cross in front strokes to get back to speed

    Can't tell how he turned back to front, but it was clockwise

    FI three, step into flying spin

    Little backward edge wiggles

    Clockwise FO twizzle or traveling threes, little falling leaf

    Crossovers, threes, mohawks, w/ some little shuffling steps and free leg kicks, quick rocker-choctaw (? so quick the edges aren't important -- I just call that move "step outs" when performed quickly in footwork);

    shallow LFO rocker into lutz (?) -- this is risky because it doesn't look much different from skaters who unintentially "lip" by turning what's supposed to be a three turn into a flip into a rocker instead -- especially since he'd already done a triple flip in this program so if the tech panel saw it that way and called it "3F e" instead of "3Lz" he'd get no credit for the element -- but maintaining the outside edge through the turn certainly is a difficult entrance into the lutz. (He did get an e call on the flip combination, so there really wasn't much difference between the takeoffs of those two jumps, despite the different preceding steps.)

    Turns directly into sitspin

    One or two steps on two feet to set up the straight-line steps

    And then inside three again straight into combo spin

    He keeps up the flow, keeps moving through the program, with nice soft knees., fairly deep edges (especially on that sequence that I boldfaced)

    Hard to tell absolute speed on video, but it looks like he gets his speed pretty effortlessly thanks to those soft knees. The edges impress me more for softness and flow than for precision. Pretty good balance of multidirectional skating.

    For transitions, the quality looks good to me, some difficult and moderately difficult turns. Difficult lutz entry. Combo was the opposite of telegraphed. Last four elements were chained pretty seamlessly one after the other, so pretty good intricacy -- only the 3A entrance really seemed lacking in transitional moves in or out. No real highlight elements such as half jumps or extended glides in difficult positions, so there could be more variety.

    More later.
    Last edited: Nov 23, 2010
  11. nashvilledancer

    nashvilledancer Active Member

    Please gkelly, analyze/compare Jeremy's transitions.
    gkelly and (deleted member) like this.
  12. midori

    midori Well-Known Member

    Thank you Doubletoe! So, TR score for Comp E would be lower than Comp A, perhaps, right?

    How about SS? Some says that falls in elements should not matter in PCS becaues PCS should be scored separately from elements. Then if everything else is the same but she fell on every jump in a competition and landed solidly in another, should SS be different?

    Thanks a lot gkelly!

    If I interpret it correctly, you are saying even if the difficulty of each move is the same, connecting one another makes it more difficult, especially when connecting into a difficult element, right?

    So, if she could have put the exactly same moves into Comp D (all triples) as Comp A (all singles), she would get higher TR in Comp D than Comp A, even though only differences are elements.

    How much would the difference be? In other words, if she does a simple arm movement before a triple jump, how complex moves into a single jump would be considered equal value on TR point of view to the arm move before the triple?

    Also thank you very much for the analyzing the program! I will watch it later when I can concentrate :)
  13. gkelly

    gkelly Well-Known Member

    The guidelines aren't that specific. There's no scale of values for program components. Judges just have to take as many different aspects of the criteria into account as they can to come up with a number on a scale of 0 to 10.

    For Transitions, I guess 0 would mean nothing but the simplest possible stroking with very weak skating quality and very telegraphed elements. Even the worst juveniles or even preliminary skaters will do a bit more than that.

    I guess 10 would be very difficult and varied moves throughout the program, including directly into and out of elements, or elements leading directly into other elements, all performed with very high quality. No simple stroking and no pauses to set up elements or to recover afterward. And nobody quite achieves that ideal either.

    But each judge has to decide for him or herself where to fit a given performance somewhere in the middle of that scale of 1-10. High middle if we're talking about elite skaters.

    Would there be any way to make the guidelines more specific so judges would be more consistent among themselves on what numbers to use for certain kinds of transition content?
  14. Lainerb

    Lainerb New Member

    So, when people complain about Miki's lack of transitons they probably aren't taking into account the short entry that she takes into her 3salchow from the 3lutz and the short entry into the 2x+2l+2l from her 3toe.

  15. jjane45

    jjane45 Active Member

    Love this thread!
  16. dinakt

    dinakt Well-Known Member

    O wow, gkelly, THANK YOU! I am looking forward to poring over what you wrote with the corresponding video. If you have time, please do more!!!
  17. Jun Y

    Jun Y Well-Known Member

    Thank you, Gene Kelly for the detailed rundown.

    I have some philosophical quibbles about the current design of the 5 components in the second score. I do not disagree that the quality of executing transitions should be taken into consideration when scoring TR. However, the quality of transitions is very closely associated with skating skills. It goes directly to the reason why all 5 components are so often in lock steps. I used to think it is because judges do not have the time and mental energy to assess each component separately. However, it is also true that skating skills indeed affect the other components as they are defined.

    Meanwhile, the TR component score is also directly and substantially correlated with the GOE of most technical elements and sometimes correlated with determining the level of a technical element via difficult entry into the element.

    In psychometric instruments (composite scales used to quantitatively assess things like depressive mood, anxious mood, fatigue, quality of life), questions that measure the same symptom should not be confused with questions that measure different symptoms. An instrument is only valid if it measures what it intends to measure. If a question is supposed to measure depressive mood, but it really measures fatigue or overlaps with another question in the same instrument that also measures fatigue, you are giving disproportional weight to fatigue than to depression. Thus the total composite outcome score will end up not giving you the information you are looking for, because the depressive symptoms were unduly discounted in the total score.

    If the TR component really measures 50% the difficulty/complexity of transitions choreographed into the program and 50% the skater's innate* skating skills, which is already given 100% weight in the SS component, the scale is basically rewarding the innate skating skills 3 times as much as the actual amount and difficulty of the transitions in a program. It seems that the other 4 components have very significant overlaps with the SS component. If so, why have 5 components? This structure gives an illusion of 5 discrete aspects of skating that are being judged, when that is not true. Clearly skaters are judged primarily on their SS, but to what extent nobody can really tell because of all the overlaps and correlations.

    *By "innate," I mean a skater's mastery of the blades, to put it generally, that is relatively stable in any given season that is unlikely to be affected by jumps and spins.


    That IJS measures and sums up discrete aspects of a program performed by a skater at a given occasion is largely an illusion because all the separate scores, including the score for each technical element executed (base value +/- GOE) and the score of each component, are highly correlated and therefore muddled. On the other hand, a jump and a spin and a step sequence are discrete elements, but a step sequence and the SS component score are not. It is not possible to clearly understand how much each discrete aspect of skating actually counts in the scheme.

    I would think that skating skills, interpretation of music, and projection and emotional/intellectual involvement with the audience are fairly distinct from each other. But that's another subject altogether.

    This "muddling" phenomenon is much bigger in the judging of ice dance than in the other disciplines. It often seems arbitrary that some attributes/characteristics are written into the criteria for level determination and other attributes are written into GOE criteria. And in the end everything seems to be determined by skating skills, with a little space left for posture and extension and another sliver left for "reasonably tasteful music."


    If the rules cannot clearly establish what attributes/outcomes are measured in the judging system and why, and how much each actually counts, it is not only the casual audience or the fans who are confused. Clarity is even more important for coaches, choreographers, and skaters.

    That is not to advocate for going back to the 6.0 system, which was hardly a model of clarity either. However, numbers alone do not guarantee accuracy, precision, and validity.


    For TR analysis of short programs, gkelly, may I suggest Adrian S... OK, sorry, not him --- it's just my uberdom talking. :p How about Kozuka's? Or Joubert's? Abbott's SP would be interesting as well, as an illustration of "How much should upper body movements count as transitions?"
    Last edited: Nov 24, 2010
  18. gkelly

    gkelly Well-Known Member

    Abbott 2010 Cup of Russia

    Lots of posing and arm movements, with simple skating including a clockwise FO three

    LBO counter to LFO three into flip combo

    Stretch with little forward edge changes to a quick 360-degree turn -- I think it's a RFI mohawk and a toe-assisted LBI three

    Mohawks, crossovers, arm movements

    LFO choctaw continuing into RBI double three, edge change, step forward into triple axel

    Controlled RBO three on the landing to RFI rocker to two-foot posing

    Inside three to sitspin

    Crossovers with arms, clockwise three and mohawk, cross in front, into lutz

    Exit landing to brief inside spread eagle, two-foot pose glide, twizzles both directions, two little scissory hops, inside spread eagle, brief two-foot pose glide, half loop, quick two-foot turn into flying spin

    Simple glide to end of ice to start straight-line steps

    Inside three to combo spin, exit straight to kneeling pose

    Edges seem well-controlled, probably a little more precise but not as effortless as Takahashi's, not the same lilt in the knees

    A bit more variety in the kinds of turns and other connecting moves (hops, spread eagles); no really big highlights -- bold for the combinations of turns and edges that strike me as most difficult

    Fine use of turns and stroking in both directions

    I'm especially impressed by the approach to the axel (same as last year) because of the seamlessness of changes of direction and the amount of ice covered through those few moves, leading into a difficult jump . . . and then more turns continuing seamlessly out of that difficult jump

    Entries to two of the spins and the step sequence were pretty basic, neither directly connecting the steps to the spins before and after nor telegraphing them

    I don't know how the arm movements would figure into the Transitions score. They seemed less obtrusive to me in this CoR performance than at NHK. My impression is they add more to the choreography than to the difficulty.
  19. Fallcolor

    Fallcolor Member

    But if you look at his short at Skate America…this is exactly where my confusion was :confused:... wasn’t the entrance just from a three turn—left forward inside edge right into the LBO “lutz” edge? There was a change of edge from going in and out... would that still be considered a rocker? And considering the lutz on its own *could* have been marked as a flat/slight inside edge, I was pleasantly surprised he got credit for a flip as well. http://www.youtube.com/watch?v=qKWvLyJ2u9s The NHK SP camera angle was slightly ambiguous I thought.
    Last edited: Nov 24, 2010
  20. Artifice

    Artifice Guest

    Jun Y, I totally agree with your brilliant analysis !
    You are one of the few, if not the only one I've heard or red that understands clearly what components are about.

    I was talking with a friend who wants to become a coach and she sometimes judges some competitions. I was explaining her how difficult it is to get a precise idea of what SS are and how accurate are transitions evaluations.
    She messed up, didn't understand the point and mixed everything in the bad way. Other coaches also have difficulties to gest the real point. As for judges I bet it's the same.
    The problem is that none of those want to agree with what you say because they don't want to recognize in front of their students that there is something not really rationnal nor really precise in their judgement.
  21. gkelly

    gkelly Well-Known Member

    Yeah, you don't really mean inborn, because they're not born with skating skills :) . . . but whatever skills they've developed by the beginning of each season are not likely to show huge variances from one performance to another, regardless of mistakes on the risky elements.

    I think the current rules are definitely a step forward from the 6.0 guidelines, but there's still a lot of room to make them even clearer.

    Especially the Transitions guidelines, which don't have any written expansions. All the rulebook says is to judge them on a scale of 1-10 on the basis of Difficulty, Quality, Variety, and Intricacy, without even officially explaining what each of those bullet points means.

    So what would be good solutions to address the problems that

    1) Transitions scores (and also Performance/Execution, Choreography, and Interpretation) appear to be too highly correlated with Skating Skills and not judged independently, and

    2) The official criteria for judging Transitions are very vaguely spelled out?

    One possibility would be to get rid of the separate score for Transitions, roll up most of it into Skating Skills and give a larger factor to that Skating Skills score, and also consider some of the variety and intricacy aspects of the transitions under Choreography.

    Another possibility would be to spell out the Transitions criteria more specifically so that judges would all be on the same page about how to judge them and skaters/coaches/fans would be able to read that page and know what to expect.

    I guess a subset of that would be to make a scale of values for (certain kinds of) transition moves.

    How could the Transitions criteria be written more clearly and objectively? Any suggestions?
  22. Jun Y

    Jun Y Well-Known Member

    I would recommend that the judging system be revised with the help of some psychometricians while involving many experienced coaches or athletes with a deep understanding of what figure skating should be about. Then determine how these ideals of figure skating should be measured with valid instrument.

    I would also recommend REDUCING rather than increasing the complexity of the scoring/judging system. Instead of adding to the judges' cognitive burden (especially those poor people who have to sit through 24 skaters in one Worlds competition, ugh), the system should SIMPLIFY. Have fewer, but better-defined criteria. Make the guidelines easier to remember, easier to apply, and clearer.

    Because figure skating is very complicated, paradoxically having more items on the list only makes a judge's judgment less accurate. For example, a judge is expected to assess variety, difficulty, intricacy, quantity, quality of all the transitional moves, all accurately and all in real time. Just for 1/10 of the total score. And judge 4 other components and GOEs at the same time. Are you kidding me?!

    The skating community has to really search their soul and decide what figure skating is all about. Set aside the endless circular bickering about whether Plushenko or Lysacek or Takahashi should have been the Oly champ. Stop trying to devise a judging system that, if applied backward, it would have given the OGM to a particular skater.

    I would suggest that the skating community think abstractly about what figure skating is all about, pick out the 3 most important and reasonably separate aspects (although total separation may not be possible). Create and validate a composite instrument that is able to measure these 3 components. Ensure the judging criteria for each component is clearly and simply defined with no more than 3 sentences. Train the judges well. Set up a system for monitoring and accountability to minimize cheating or gross negligence. Then set them loose. Allow judges to come up with a relatively holistic SS or IN/Performance score based on their below-consciousness summary impression.

    I personally feel the TES system is clear and sufficiently discrete among elements as to need no significant revision.

    However, I think the above-described revision for "the second score" will probably not be accepted. Muddledness, or the lack of clarity, has always been a part of figure skating as a judged sport. This historical lack of clarity may have had some interesting unintended consequences. But that's another topic altogether.
  23. gkelly

    gkelly Well-Known Member

    Do you volunteer to give it a try? In collaboration with other fans who have strong knowledge of psychometric criteria and/or of what figure skating values?

    I'm not sure that's true. If you reduce the number of criteria, you may get more accuracy and agreement in judging those criteria that are left, but you'll still have many other skating qualities that have historically been valued (but not quantified, especially under 6.0) that are not addressed at all in the official criteria.

    So either all the judges will just ignore those criteria that didn't make it into the official rules and things that used to be important will get lost, or they will let their gut feelings about those qualities and memories of when they used to matter color the way they judge the remaining criteria.

    OK. I'll give it a shot.

    I'll call them

    *Blade Skills
    Everything that's currently in the Skating Skills criteria, along with difficulty and quality of the connecting moves.

    Summarized as form (carriage and body line), relation to the audience, "attack" and overall impression of mastery

    Design of the program in space and time; relation of the movement to the music in style and phrasing; originality and variety

    But mainly that's just rearranging the existing criteria and deemphasizing some of the vaguer ones.

    What would a validated "composite instrument" that is able to measure these 3 components look like? Can you offer some specific suggestions?

    Under that division, I would probably want the Blade Skills component to be worth the most, maybe equal to the other two combined, for a technical sport competition.

    But I bet a lot of fans who enjoy skating mostly for the artistry, especially those who watch only on TV where they can't see and hear the differences in blade skills and don't have the knowledge to tell a difficult skill from an easy one, would prefer for the Performance and Choreography components to be worth more.

    And you're still going to have judges disagreeing with each other and fans disagreeing with some or all of the judges. And there's still no way to read judges' minds as to whether their scores are arrived at through honest holistic impression or through cheating or incompetence.

    So you'll still have fans complaining about results they don't like and accusing judges of dishonesty. The only thing you'll lose is complaints about the scoring pretending to be more precise than it really is.
  24. Coco

    Coco Well-Known Member

    GKelly!! thanks!!
  25. Fallcolor

    Fallcolor Member

    :respec: to the idea of clearing up the guidelines for transitions. Although this would never happen, if a skater manages to skate a clean 7-8 triple program, possess excellent skating skills, good use of one-foot skating, but has no transitions or any difficulty in linking movements-> somehow I'm not sure if the judges would mark the TR score alone, accordingly. Although these kinds of extremes are rare, usually find the first 3 categories- most correlated in terms of scores. Of course having TR are related to SS, the converse may not be true. Same can be said for CH and SS.
  26. Jun Y

    Jun Y Well-Known Member

    I know, Gene Kelly. It's not easy. :D Greater clarify does not always give everything everyone wants. A vague system allows more room for ... various ideas and personal projections. It's kind of a philosophical question. That's why I sort of brought up the historical lack of clarity and its unintentional consequences, but could not go on.

    I completely agree with you about the "artistry" issue, which, if we are honest with ourselves, is practically (although not theoretically) unjudgeable. Even the current IN component is practically unjudgeable. I have become convinced that some people (skaters as well as judges, not to mention viewers) who can do or assess other aspects of skating well enough have a very hard time picking up details in music, from rhythm to phrasing. This may be an inborn trait, as Oliver Sacks pointed out. Is it fair to include musical interpretation as a component? Is it still figure skating if it's not included? Ah, the mind boggles ...
  27. gkelly

    gkelly Well-Known Member

    What do you think the TR score alone should be, if marked accordingly?

    In what way do they have good use of one-foot skating but no transitions? In between the crossovers they do a lot of neutral one-foot glides on various edges, and simple threes to turn forward to backward, but nothing else?

    Now I'm wondering what much more objective, spelled-out guidelines for Transitions might look like.

    Everyone starts out at 0, or at a baseline equal to one-half or one-quarter of the Skating Skills mark.

    Add 0.25 for every element that's entered from other than the most common preceding edge or turn.

    Add 0.25 for every connection between elements with no more than 1 additional edge that is not part of either element.

    Add an additional 0.25 for every element entrance or exit that is especially difficult.

    Add an additional 0.25 or 0.5 to the above for exceptional quality and seamlessness.

    Subtract 0.25 for every second of telegraphing.

    Add points for highlight moves between elements that do not receive base values (nonlisted jumps; spirals, spread eagles, Ina Bauers, shoot-the-duck/hydroblades; turns and steps other than mohawks and forward threes outside the step sequence) -- 0.25 just for their existence if they're recognizable, with additional 0.25 increments for the difficulty, quality, and intricacy of each up to a maximum of 1.0 for any highlight move performed in isolation or 1.5 for any highlight move directly connected to an element

    Add 0.25 for variety if the skater uses three different categories of highlight moves (nonlisted jumps, glides, steps), and an additional 0.25 for three different kinds of elements within any of those categories

    Are guidelines like that more what people would like to see?
  28. Fallcolor

    Fallcolor Member

    Sorry, I should have been a bit more clear-while moves in the field are marked as transitions, if the skater does a short sequence of one-foot glides, holding edges, but then proceeding with a series of crossovers, and say, the simplest entrances going into the jump. Or skaters who are able to do one foot turns/skating that relates to the choreography, but stalk the entrances going into elements (or, with considerable time to set up)?
  29. Jun Y

    Jun Y Well-Known Member

    I have given your challenges a little more thought, gkelly, and here are some very rough ideas:

    For the "second score", I came up with just two components/domains: 1) Skating Skills and 2) Choreography/Expression. To put them simply, SS = blade skills that are not accounted for in TES (i.e., everything in between that are not crossovers), which would include transitions; C/E = the "artistry" part, which would include interpretation of the music and creativity/vision/originality, the intangible stuff.

    Each of these two components are basically made up of two considerations: How difficult or how good is the "in-between" stuff built into the program by the choreographer? How well does the skater execute the choreography --- technically and artistically?


    In medicine (including psychiatry), psychometric instruments widely used to measure subjectively reported outcomes (e.g., physical function, quality of life, mood) by patients, caregivers, or clinicians, usually have several basic formats:

    1) a numeric scale (known as a Likert scale) using only integers: -3 to +3, or -5 to +5, or 0-5, or 0-10. (For example, -3=unable to function, while +3=no difficulty with movement.) Of course, FS judging would require more digits.

    2) a visual analog scale: The reporting/assessing person marks a point on a continuous straight line that is 10 cm in length, with 0 being the worst and 10 being health/normal. The placement of the mark is measured as x.x cm, as a reflection of how well or poorly a person feels or does.

    3) Dichotomous scale: yes or no. Yes = 1 and No = 0. Or Yes = 3, Maybe = 1, No =0.

    The scores of all questions can be simply added up, or weighted with different percentages and then summed up.

    The design of the questionnaire should be simple: easy to understand and easy to use by the person doing the reporting or assessment (patient, caregiver, or clinician). However, the validation is critical: Does the instrument measure what it intends to measure in the target population or circumstance? How sensitive is it? How specific is it? How consistent is it? Can the instrument maintain its ability to measure what it should measure across the spectrum, e.g., from juvenile to elite senior skaters? Can one judge use it consistently for different kinds of programs and skaters (intra-person validity)? Can different judges look at the same skater's same program and come up with very similar scores (inter-person validity)? These are the questions generally used to test for the sensitivity, specificity, and validity of such instruments.
  30. Artifice

    Artifice Guest

    You guys put very interesting points in this thread !

    As for my two cents, first of all I believe there are too many components in the current system, too many because it seems that judges have difficulty to separate them in their judgement, and also several that are so closely related that I don't see the point of keeping them separate.

    To explain it more precisely, here are the current components :

    Skating skills

    We can assume that these components include all the elements that actually make figure skating and that there is no other aspect to include in figure skating. Some elements that are not clearly visible through the components title are actually part of these components. Ex : capacity to move with musicality, carriage...

    So, some components that are judged separately are actually very related to each other and concern the same global skills :

    Transitions are very much a part of skating skills. The variation of connecting moves as well as their difficulty and quality of execution are directly linked with the basic skating skills. Indeed judges mark them very equally. And it's very difficult to make a clear difference between fluidity, speed and quality of connecting moves. Without skating skills it's impossible to have speed and fluidity. And there is always questions about how to judge speed vs difficulty. Actually in figure skating schools both aspects are taught simultaneously and no choice is recommanded between skating fast and doing hard moves. It's always "do both" ! So I don't see any reason to mark them separately in competition.
    To me SS and transitions should therefore be one component.

    As for the other criterias : performance, choreography and interpretation, if we keep them separate like today it would increase their relative value comparatively with the SS/TR component (since there would be only one mark given to the SS and TR criteria instead of two). And I dont think it would be good to decrease the SS/TR mark because this should actually be the main criteria, which is actually the case today.

    So, either I would regroup some components among the PE, CH and IN criterias, or I would factorize the SS/TR mark so that it reaches the right proportion within components.

    I actually think it's relevant to keep the PE, IN and CH components separate since I see pretty clearly the difference between them. For instance, Chan is a good exemple IMO : choreographically wise he shows interesting moves and variations of steps and upper body positions. His carriage and overall presentation are good. However I don't think his interpretation is good, he lacks charisma and is more like a "good student" who does what he has to do than a real actor or dancer who would put his soul into his program.

    To me the fairest and mathematically perfect choice would be to have the following components :

    Skating skills (SS + TR), factorized (ex : x1,3)

    But are judges always capable of making the difference ? Not always actually.
    I don't think they all have the knowledge and competencies to differentiate clearly all those aspects, especially in the context of a competition, while watching 20 or more skaters live and have to mark them quite fast. The experience prove that they indeed tend to give similar marks for all components.

    So, since I doubt of the possibility for judges to judge those criterias with clarity and knowledge of what they actually judge, I believe it would be relevant to simplify things and make components look like the following :

    Skating skills (SS + TR) -> everything relating to blades work, edges, speed...
    Performance / Interpretation -> everything relating to presentation, acting, musicality and overall personal implication.
    Choreography -> external input (choreographer) for making everything look like a whole, originality, accuracy of moves with the music choice.

    This way there would probably need no factorization for any of the component and it would simplify the cognitive aspect of judgement for judges without removing any aspect of the overall skating.
    As a result Chan would have the following evaluation under the current components :
    SS : excellent (what he gets today)
    TR : excellent (what he gets today)
    PE : very good (he gets too much today)
    CH : good (he gets a little too much today !)
    IN : average (he gets way too much today !)

    Under my new components it would give :
    SS/TR : excellent (which is fair for Chan)
    PE / IN : between average and good (which overall is fair : he looks good on the ice but not much)
    CH : good (everything is well put together, but still not making a big artistic impression)
    Last edited by a moderator: Nov 25, 2010