So true!
What is the difference between Choreography and Interpretation?
- Thread starter Mathman
- Start date
So true!
- Joined
- Jun 21, 2003
I don't know what I feel about this. I think it's different at the elite level than for kids who are trying to pass their pre-juvenile tests.
At the elite level, I think the ISU is very aware -- or at least it should be -- that big time sports is first and foremost entertainment. Send the people home happy and wanting more.
In basketball, the referees could blow a whistle on every play. But that would disrupt the entertainment value of the game. Basketball players make millions and owners of basketball teams, billions, by putting out a product that folks will spend their hard-earned on.
When Slutskaya brings the house down in her home town at the World Championships, that's money in the bank. The scores come up -- a new all-time world record!!!! Yay!!!! Ten thousand Muscovites are dancing in the streets.
In principle, the Grand Prix series is supposed to work like that, IMHO. You have a "regular season" culminating in a title match. Then you go on to the "post season" with Four Continents, Europeans and -- the Super Bowl of figure skating -- Worlds.
Personally, I think this is a good format. The only problem is, the overall interest in the sport is not sufficient to sustain it.
IMHO the balance struck by the current version of the ISU judging system is not bad. A triple flip is 5.5 points, cut and dried. If you wow the audience by emoting to the music, that's more subjective, but the judges have ample opportunity to reward you for it.
Sport and art, LOL.
- Joined
- Jul 11, 2003
I have never read anything that the ISU is aware of or should be aware of that sports is first and foremost entertainment. That to me, seems to be one minority thought. Most sportsminded people have a personal interest in sports and another interest in threatrical performances. I presume you combine the two into one general sweep, smile, and get back to your more serious statistics.
No problem, to each his own.
In statistics, one could emphasize either side of the bell curve. But, wouldn't that disrupt the entertainment value of the meaning of statistics. rarely statisticians make millions and their publishers put out products that shock the general public. But it's entertainment.
Kwan made a fortune on her fame and no serious hip problem was going to prevent her from making that fortune - win or lose a competition. No matter her fans loved her and the family made good use of it. Entertainment? NO. They loved her as a sportswoman.l
Exactly, the sport is in a downward spiral (pun intented). Judges do have an ample opportunity to reward you for wowing the public or to disaward you. That's what subjectivity is all about.Personally, I think this is a good format. The only problem is, the overall interest in the sport is not sufficient to sustain it.
Tosca preferred Art and Love, so maybe the whole thing is entertainment and we should go to Basketball and watch a real game of sports.
Joe
A few more thoughts on post 77 after a restful 2 hour drive home tonight.
The point I am flogging is that without adequate markers people typically do no better than 10-15% in perceiving differences in things, and in making absolute measurements. The counter example about estimating the height of people to an inch (1.5%), however, is not applicable because that is a measurement that uses more than the unaided eye. You look the other person over, in your mind you compare them to your own height which you know, maybe you compare them to a person they are standing next to, or the door frame, or the furniture, or you use knowledge of the typical height of a person, or the height of people of a given age, or sex. You have all manner of assistance/guides you draw on when estimating the height, that you use at the moment you make the estimate. But if you take all those things away you would not do so well.
For example, have a friend pick ten strangers and dress them up all the same. Put a bag over their heads so you don't know their age or sex. Put them in front of a large white background with no markings on it, one at a time, in the middle of a field 100 feet away like they were at center ice, in random order of height. All you get to guide you is one view of a 6 foot tall person at the beginning of the experiment. Try and correctly determine the height of each person to within one inch. Ain't gonna happen.
Of course, even in this example you still know they are people so you still have some information to help. So better yet, have your friend stick poles of different but similar length in the ground, one at at time. You will probably do even worse. (Another interesting part of this experiment would be to see by how much your judgement drifts during the experiment, and how quickly.)
The 10-15% rule doesn't mean that is the best a person can do, ever. It means in the absence of adequate markers that is the best a person can do. If you have sufficient markers (benchmarks, a ruler, supplemental information, etc.) a person can do better.
Currently, under IJS, the PCs have few markers specified or that the judges can rely on as they observe the program, so accuracy and precision of the judges' marks are limited by the 10-15% rule. The spread in marks among the judges in the PCs in competition consistently bears this out. For the GoEs where the judges do have better guidelines and better markers the spread in the GoEs is usually fairly low. (I went back to one of those competitions I mentioned in a previous post, and the spread in the PCs was typically 1.5 to 2.0 points out of a panel average of 4 to 6 points -- a spread of 25-50% relative to the panel average for Juvenile through Junior events.)
So the question (for me) is, what can be done to the PCs and the process of scoring them to do better than the 10-15%? How can the criteria be improved to achieve this? How can markers and better guidelines be added? It is not fair to the skaters to say we will select the places to within 0.01 of a point (1/200 of 1% of the total score in the men's event) when the judges currently are floundering with scores that are spread out by 10-15% for an individual judge and uncertain by 1-2 points for even a 10 judge panel as a whole (and some USFSA clubs are using 3 judge panels with IJS -- OMG!).
The point I am flogging is that without adequate markers people typically do no better than 10-15% in perceiving differences in things, and in making absolute measurements. The counter example about estimating the height of people to an inch (1.5%), however, is not applicable because that is a measurement that uses more than the unaided eye. You look the other person over, in your mind you compare them to your own height which you know, maybe you compare them to a person they are standing next to, or the door frame, or the furniture, or you use knowledge of the typical height of a person, or the height of people of a given age, or sex. You have all manner of assistance/guides you draw on when estimating the height, that you use at the moment you make the estimate. But if you take all those things away you would not do so well.
For example, have a friend pick ten strangers and dress them up all the same. Put a bag over their heads so you don't know their age or sex. Put them in front of a large white background with no markings on it, one at a time, in the middle of a field 100 feet away like they were at center ice, in random order of height. All you get to guide you is one view of a 6 foot tall person at the beginning of the experiment. Try and correctly determine the height of each person to within one inch. Ain't gonna happen.
Of course, even in this example you still know they are people so you still have some information to help. So better yet, have your friend stick poles of different but similar length in the ground, one at at time. You will probably do even worse. (Another interesting part of this experiment would be to see by how much your judgement drifts during the experiment, and how quickly.)
The 10-15% rule doesn't mean that is the best a person can do, ever. It means in the absence of adequate markers that is the best a person can do. If you have sufficient markers (benchmarks, a ruler, supplemental information, etc.) a person can do better.
Currently, under IJS, the PCs have few markers specified or that the judges can rely on as they observe the program, so accuracy and precision of the judges' marks are limited by the 10-15% rule. The spread in marks among the judges in the PCs in competition consistently bears this out. For the GoEs where the judges do have better guidelines and better markers the spread in the GoEs is usually fairly low. (I went back to one of those competitions I mentioned in a previous post, and the spread in the PCs was typically 1.5 to 2.0 points out of a panel average of 4 to 6 points -- a spread of 25-50% relative to the panel average for Juvenile through Junior events.)
So the question (for me) is, what can be done to the PCs and the process of scoring them to do better than the 10-15%? How can the criteria be improved to achieve this? How can markers and better guidelines be added? It is not fair to the skaters to say we will select the places to within 0.01 of a point (1/200 of 1% of the total score in the men's event) when the judges currently are floundering with scores that are spread out by 10-15% for an individual judge and uncertain by 1-2 points for even a 10 judge panel as a whole (and some USFSA clubs are using 3 judge panels with IJS -- OMG!).
Last edited:
I think this question coupled with what really is the difference between ch, in pe (as we were discussing earilier) since they can be defined separately but interact/depend upond/etc each other get at the heart of it.
I think gkelly's macro-micro distinction is really useful for thinking about how marks are relative to the field before you in two kinds of ways (for simplicity, bad and good skater, then within the bad and within good). I think grossano's discussion explains a lot about how difficult it is to make the gradations - again for simplicity - among the bad and among the good; by making the gradations I mean deciding among and ranking within a group.
Both the old and new system did/do this ranking and judges had a whole lot explict instruction/knowledge that they surely could articulate to explain why a particular ranking (and surely a whole lot of tacit knowledge too that would necessarily be harder to articulate but would include a range of assumptions about ranking and the classifictory scheme used to place people etc)...the problem with respect to the new seems, at this point to me, to be exactly what MM asked in the quote above. Like grossano said - we don't have a ruler - or something that says choreography like this is worth xx points, transitions like that are worth yy points. So there is a tension between the veneer of absolute points and the reality of relative points.
IMO from what I have read so far, and just thinking about skating - I imagine it would be darn near impossible to make these points (ch, in, pe) absolute because while skaters might skate at a particular level, they are so varied in their actual programs - UNLESS you have one program that everyone performs (which to start with might make ch a very, very stupid mark or redundant at best).
So, here's my new question: at the national level (say within USFS or Skate Canada) are their discussions among judges (formal ones like workshops or the like) where the question of why a.25 difference here, why a .50 difference there is debated and fleshed out in the attempt to create a shared framework of understanding for the meaning of particular point groupings? does this happen at the ISU level?
I just wanted to add - as I continue to mull this over - that my last questions were derived from thinking about:
and
I also want to add - again, that I don't see these questions as critizing judges; I am in awe of their expertise. Just trying to figure out the architecture/broad strokes of how they use this system/what the categories and numbers mean to them. I also want to be clear that I don't have some expectation that if the categories and meaning of numerical increments are 'perfectly' defined (as if that could happen!), that all of the judges on every panel will always agree OR be very very very close in agreement. I do like that judged sports rely, well, on judges - their knowledge/expertise/gut reactions etc are part of the competition. So, just like our competitors overlap in skills etc, so too will judges, but just like the competitors who differ (at times very subtley) so too will judges. I guess what I'm saying is I'm NOT asking for computer/robot like humans (and just wanted to make that clear).
and
I also want to add - again, that I don't see these questions as critizing judges; I am in awe of their expertise. Just trying to figure out the architecture/broad strokes of how they use this system/what the categories and numbers mean to them. I also want to be clear that I don't have some expectation that if the categories and meaning of numerical increments are 'perfectly' defined (as if that could happen!), that all of the judges on every panel will always agree OR be very very very close in agreement. I do like that judged sports rely, well, on judges - their knowledge/expertise/gut reactions etc are part of the competition. So, just like our competitors overlap in skills etc, so too will judges, but just like the competitors who differ (at times very subtley) so too will judges. I guess what I'm saying is I'm NOT asking for computer/robot like humans (and just wanted to make that clear).
I think it would be a good idea to have a table for component scores, so that judges have something concrete to help them out. For example, for transitions, you could have:
0 - no transitions between elements (so one jump, stop, turn around, do a spin, stop, do another spin, etc.)
1- exclusive unidirectional crosscuts between elements
2- Crosscuts in all directions (forward, backward, clockwise, counterclockwise)
3- Some transitional moves here and there (like 3 or 4 in the whole program, for an LP)
4...
5...
6...
7...
8...
9...
10 - Non-existent crosscuts, which makes this nearly impossible, just as perfection is not of this world.
You get the idea. This way, judges could have an idea of what range of marks would be. Increments would be for the fact that, say, a 2.50 would be for a skater who majoritarily used crosscuts, but inserted maybe 2 transitional elements.
Kypma
0 - no transitions between elements (so one jump, stop, turn around, do a spin, stop, do another spin, etc.)
1- exclusive unidirectional crosscuts between elements
2- Crosscuts in all directions (forward, backward, clockwise, counterclockwise)
3- Some transitional moves here and there (like 3 or 4 in the whole program, for an LP)
4...
5...
6...
7...
8...
9...
10 - Non-existent crosscuts, which makes this nearly impossible, just as perfection is not of this world.
You get the idea. This way, judges could have an idea of what range of marks would be. Increments would be for the fact that, say, a 2.50 would be for a skater who majoritarily used crosscuts, but inserted maybe 2 transitional elements.
Kypma
No!
Excellent summary emma (#86) of the complexity of what is going on!
As for the question of whether the judges discuss the diversity of the opinion in the marks, informally in a non-qual competition and in a more formal way in qualifying competitions. But for the PCs the discussion is mostly limited to why one judge might have scored a skater in the 3s and another in the 4s, for example. Nobody quibbles much over differences of opinion at the sub-point level -- and certainly not at the 0.25 point level. If we all agreed at the 0.25 to 0.50 point level for all the skaters we would all be thrilled!
Last edited:
- Joined
- Jun 21, 2003
Personally, I am very doubtful that this question will ever be satisfactorily addressed.
To me, GKelly's excellent and thought-provoking post (#77) makes it quite clear how the judges actually judge program components. Within a general range corresponding to the level of the competition, they give scores at the high end of the range to the best performances and lower scores to the perfomances that were relatively weaker in the area (choreography, etc.) under view.
In other words, ordinal judging.
I do not see how it could be otherwise. Judging and measuring are two different things. In the IJS we measure the tech and judge the program components. The first mark, and the second.
As Emma put it above (cutting right to the chase, LOL)...
:yes:emma said:
- Joined
- Jun 21, 2003
Just a technical question here. Do moves in the field, like Ina Bauers, split jumps, spread eagles, hydoplaning, etc. count in the Transitions category?
What about a Charlotte? Is that a spiral to be judged under TES, with added GOEs, or is it a Transition?
What about standing on your toe pick and doing a couple of chorus line high kicks, like Michelle did in her Fate of Carmen 1998-99 short program?
But about having a ten-point table like this to help the judges out, again the problem is that you would have to number your list from 1 to 40. And there would have to be language that would allow the judges to distinguish between level 27 ( a PCS of 6.75) and level 28 ( = 7.00).
Last edited:
Yes
A Charlotte in a spiral sequence element must be on an edge to be counted as a spiral position, but it is still part of the element even if it isn't on an edge. A Charlotte in isolation elsewhere in the program does not have to be on an edge and is a transition movement when executed outside a spiral sequence.
Yes. Any movement (that is not an element, of course) that is not crossovers or stoking is a transition movement.
Last edited:
But about having a ten-point table like this to help the judges out, again the problem is that you would have to number your list from 1 to 40. And there would have to be language that would allow the judges to distinguish between level 27 ( a PCS of 6.75) and level 28 ( = 7.00).
I don't really think a subdivision is necessary. When I thought out the chart (not for very long, I must say), I figure that, say, since 2 is Crosscuts in all directions (forward, backward, clockwise, counterclockwise) and 3 is Some transitional moves here and there (like 3 or 4 in the whole program, for an LP), if a skater put in a spear eagle once in the program, that could bring the mark from 2.00 to 2.25, two spread eagles could be 2.50, and two different transitional moves could be worth 2.75. By the time you have 3 moves, well, youève earned yourself a 3.00
This is how I see it... judges could still vary by 0.25 increments, but they still have baselines so that the marks can be somewhat justified and universal, just like spins and jumps - a 3F is worth 5.5, and that's it. In my transitions example, just crosscuts in all directions is worth 2.00... Just an idea.
Kypma
The problem with judging transitions, as I see it, is that you have to take into account not only what is done, but also how well.
The official definition of the component lists just Variety, Difficulty, Intricacy, and Quality for singles skaters (there are other criteria that apply only to pairs and dance). It seems that many judges have been focusing primarily on the quality and giving top scores to top skaters with the strongest skating quality even if the actual content of the program between the elements is quite bare, and lower scores to weaker skaters with more intricate programs.
That is appropriate up to a point. If I do a program with mostly crossovers, a few isolated spirals, and the occasional forward rocker or backward three, and Michelle Kwan does a program with essentially the same transitional content, she's obviously going to perform them much much better than I will and therefore deserve a better score. So what if I add even more complexity to my program? Will that narrow the gap? Even if the extra difficulty makes my quality decline even further relative to hers?
I'm never going to come close. But what about an average junior level skater with junior level skating skills who happens to include all sorts of turns, steps, small/unlisted jumps, spirals, spread eagles, etc. throughout the program including leading into or out of most of the elements? Maybe she would deserve a higher transition score even if it's her highest PCS and Kwan's lowest.
It might be possible to quantify the variety-difficulty-intricacy with something like Kypma's table. So then could the quantity be handled somehow similar to the GOEs, going up or down from the transitions base mark as appropriate? How much should each step up or down be worth?
The official definition of the component lists just Variety, Difficulty, Intricacy, and Quality for singles skaters (there are other criteria that apply only to pairs and dance). It seems that many judges have been focusing primarily on the quality and giving top scores to top skaters with the strongest skating quality even if the actual content of the program between the elements is quite bare, and lower scores to weaker skaters with more intricate programs.
That is appropriate up to a point. If I do a program with mostly crossovers, a few isolated spirals, and the occasional forward rocker or backward three, and Michelle Kwan does a program with essentially the same transitional content, she's obviously going to perform them much much better than I will and therefore deserve a better score. So what if I add even more complexity to my program? Will that narrow the gap? Even if the extra difficulty makes my quality decline even further relative to hers?
I'm never going to come close. But what about an average junior level skater with junior level skating skills who happens to include all sorts of turns, steps, small/unlisted jumps, spirals, spread eagles, etc. throughout the program including leading into or out of most of the elements? Maybe she would deserve a higher transition score even if it's her highest PCS and Kwan's lowest.
It might be possible to quantify the variety-difficulty-intricacy with something like Kypma's table. So then could the quantity be handled somehow similar to the GOEs, going up or down from the transitions base mark as appropriate? How much should each step up or down be worth?
- Joined
- Jun 21, 2003
gkelly, can you give some examples that might be available on You Tube of programs that have great quality in the transitions but little variety or difficulty? Would you say Michelle's Kwan's Aranjuez or Tosca would fall into this category?
What about the other way around? Matt Savoie? (although he was good in all four, IMHO).
What about the other way around? Matt Savoie? (although he was good in all four, IMHO).
- Joined
- Mar 19, 2006
I am not saying this to discredit, but...
My Sister-in-law (works for the Sheriff's Dept.) can do this and has exhibited her ability to in very random situations. I have shown her a pic of my friend alone that was in front of a sky and clouds background only. Her comment was, "I thought he would be taller than you Sean." How she knew I am 2 inches taller????? Also the fact there is going to be some tell tail object to tell. I see your "white background with no markings on it," but this is not a realistic comparison. "One" knows how high the boards are around the rink, "One" knows how big the rink is, and as a matter of "ratifying" "One" knows which direction they are facing prior to the jump. So there are things around to gauge these things - although none of my examples had to do with hight.
The ratio / percentage of "mistake" is understandable. But I think that it is in question (the scoring) more often than 10-15%.
And as the thread goes, IMO have not seen or heard any good reasoning on why CH and IN should be separate marks. The reasons I have heard are more along the lines of "capability to work with that which is" and the disassociation of WHY CH is what it is, is the exact same reason there is music to interpret in the first place. You CAN NOT have one without the other. Is there such a thing as a skater interpreting music WITHOUT CH????
On the GP topic, I was thinking more along the lines of all events are treated like the GP. And have MANDATORY comps. Don't show you get 0 that is factored into your seasons score. I don't necessarily like the idea, but that is what I was thinking in response to some other idea only. GP the way it is, fine with me. etc...
Last edited:
How about the Plushenko program that I linked much earlier in this thread . . . or most of his programs, for that matter?
Yes.
How about these?
http://www.youtube.com/watch?v=ZMwXYdkSK4g
http://www.youtube.com/watch?v=vj62HfrfIfA
http://www.youtube.com/watch?v=NXHdm4fGeYE
In appearing to disagree with me, I think you are in fact exactly making my point. The more frames of reference you have in the rink, or wherever, the better one's assessment will be. With only a few reference points you do worse. The person against a a white background is just an example of taking away most frames of reference to contrast with other more realistic situations.
Some random comments.
Crossovers (crosscuts) are not transitions. Neither is stroking. They can, however, play into consideration in the skating skills mark.
Coming up with a transitions marks is in fact five dimensional.
What fraction of the program (outside the elements) contains transitions?
Are a variety of transition movements included?
How difficult are the transition movements?
Do the transition movements have intricacy (complexity)?
With what quality are the transition movements executed?
I don't like the idea of making up a list of transitions and assigning a ranking to them. If you do that, you might as well have the Technical Panel call them and give them base marks and have the judges score the quality. What a nightmare that would be (IMO).
Last edited:
- Joined
- Jul 11, 2003
Anyone to reply who has the answer:
(The best way to get speed is with crossovers, ok?) so a skater does a number of crossovers (you know a jump is coming up) so in order to satisfy the regulation to do some transition into an element, he/she does (and this is quite common after all those crossovers) a Rocker - cross free foot, forward step into mohawk into a Flip. For me that's just a big deal
and transitions didn't really happen, except in their most primitive form.
My question: does this satisfy the question of transitions?
There is another approach: from what I saw in Tokyo.
It seem to me that Mao did do a series of mixed steps which finished with a choctaw into a Triple Axel. Quite rare in competition where skaters usually just take off with a long (amost too long) back outside edige into a 3A.
My question: Are the judges aware of this or are they just interested in the 3A?
Joe
(The best way to get speed is with crossovers, ok?) so a skater does a number of crossovers (you know a jump is coming up) so in order to satisfy the regulation to do some transition into an element, he/she does (and this is quite common after all those crossovers) a Rocker - cross free foot, forward step into mohawk into a Flip. For me that's just a big deal
and transitions didn't really happen, except in their most primitive form.My question: does this satisfy the question of transitions?
There is another approach: from what I saw in Tokyo.
It seem to me that Mao did do a series of mixed steps which finished with a choctaw into a Triple Axel. Quite rare in competition where skaters usually just take off with a long (amost too long) back outside edige into a 3A.
My question: Are the judges aware of this or are they just interested in the 3A?
Joe
Then in your notes you would record there were transitions and they were simple. You would make a mental note of how well they were executed. If they did the same thing for every jump you would make a note the transitions lacked in variety.
Yes, what you describe would be considered transitions preceding the element.
Same as your first example, you would note that transition movements preceded the jump. You would note its difficulty and quality. You would compare it to all the other transitions at the end and decide how much variety there was in the transitions.
Note also the transitions into the jump do not have to be turns or steps. Other examples: a spread eagle or Ina Bauer into a double Axel; hydroblade move into a jump (or spin); posing and arm movements preceeding an element.
We are supposed to be interested in everything. If the jump had a difficult entry that would be taken into consideration in the GoE, maybe bumping up the GoE 1. If transitions were lacking we would take that into consideration in the transitions mark. If the skater rode a long edge in preparing for the jump we would knock the GoE down 1.
Everything is supposed to be accounted for someplace, and some things are accounted for more than once and in more than one way.
Last edited:
