Actually, I don't disagree with anything that you are saying here. Maybe we are just misunderstanding each other.
Probably. We can get closer to the same page by using more specific examples, but then this thread and our posts will get even longer.
And, yes, the number of discrete variables means that the number of permutations is also large, but again, they are mathematically finite.
…Certain variables/aspects are only bounded by the limitations of human capability and physics (e.g. how "deep" an edge can go, or how fast they can skate, or the number of rotations, "hang time"/delay, and distance in a jump), but, again, these can be mapped to a numerical range of scoring effect.
…This is what was meant by the proposition that the technical elements can be viewed though a pragmatically Platonic lens.
Still not sure how this would work.
Is one of these closer than the others to what might be considered the platonic ideal of a layback spin?
http://www.youtube.com/watch?v=qnoRJ8gZpZU&t=0m55s
http://www.youtube.com/watch?v=v_Ikxi6Zx4A&t=0m30s
http://www.youtube.com/watch?v=K5p3zrEI7N0&t=2m25s
http://www.youtube.com/watch?v=m4IFWjMlFVk&t=4m30s
-Sure, there are cultural differences, but those differences don't prevent viewers from learning to understand and appreciate other (human) cultural cues if they make a good-faith effort.
True. That's where education comes in -- educating judges and educating skaters/coaches and educating audiences.
There will still be some cultural and personal preferences, though. We're not going to get everyone to agree 100% on everything.
The scores that the judges get to give (as opposed to the averages) aren't that precise, though.To re-emphasize, I am not against numerical scoring of itself. Any competitive endeavor, almost by definition, must have some numerical basis, however simple. Where I question COP is in the implied precision of the scoring for the artistic components (PE, CH, I), because the scoring increments cannot be persuasively shown to correspond to a set of discrete variables with fairly precise gradations of incremental fulfillment (in contrast to the "technical" elements).
On a scale of 0 to 10 that covers the whole range of skating ability, if 0 means "you didn't make the slightest effort," 5 means "about the minimum we'd expect at a senior international level," and 10 means "wow I can't imagine anything better -- different maybe, but not better," giving only integers with no decimal places would mean there's no way to distinguish between all the average-for-their level competitors in the same event. So what if we allow judges to give .5 scores in between those benchmarks? That gives them a little more control. .25 gives a little more than that, but there it has stopped. Any increments smaller than 0.25
are the result of averaging across the whole panel.
I think what you're proposing is to change the meaning of 1, 5, 7, 10 etc. from each competition to the next. I don't know how that would help.
Thanks for thinking things through here -- it's interesting.
As per my caveat, this is just a preliminary idea off the top of my head. But for the sake of discussion, how about this:
-Each "artistic" component (that is to say, this only applies to PE, CH and I, because I consider skating skills and transitions to be, in concept if not always in actual practice, "technical" and hence more amenable to precision) is scored separately. For each of these components scores, the skater receives a rank relative to the field. So if there are 30 competitors, there are 30 rankings.
-Each ranking number (e.g. #1, or "the best") can have, say a maximum of 2 or 3 names in that slot. In other words, two or three skaters can be tied in the #1 slot. Similarly for the #2 slot, and so on. This doesn't mean that it's mandatory for judges to put in the maximum number of names in each ranking slot. This is only if the judge feels that there is no clear choice for that ranking slot. So (assuming a maximum of 3 skaters per ranking slot) you could end up with, for example, 2 skaters ranked #1, 3 skaters at #2, only 1 skater in the #3 slot, and so on. What this means is that there may be 30 rankings, but they may only go down to, say, #20 (since there are multiple names in slots in some cases).
OK.
Is it acceptable to skip slots? E.g., if the best skater is miles ahead of everyone else on that component, is it acceptable to award that skater #1 and the next best skater (who really isn't very good) #3 or 4 or 5?
Or maybe there are three good contenders (all #1, or #s 1, 2, and 3 respectively) and then the best of the rest gets #5 or lower.
The big-picture theme, though, is that the increments are pre-determined and are not discretionary on the part of the judges.
This, I have a big problem with. It takes away the power of the judges to judge.
In 6.0 judging, it didn't really matter if, for example, a judge gave the #1 skater at a national competition 5.9s and the #2 skater 5.7s or if she gave them 5.8s and 5.3s, respectively. The second place skater still got a second-place ordinal. But the numbers did tell us a tiny bit of information about what the judges thought about the programs, and that was the only information we got. Meanwhile, the judges had complete control over their own placements.
In IJS judging, if the judge gives the #1 skater 8.5s and the next best skater 6.5s, that will make a difference in the results if #1 doubled a couple of jumps, for example, but was still significantly superior in the components. But the judges can't completely control the placements because they don't know the levels of the spins and they haven't memorized the point values of every element and don't have time to add them up.
-I am currently envisioning that the artistic components scores would be given by each separate judge for each separate component, as is the current practice. The key difference is that the judge is only responsible for determining greater than/lesser than/equal to, strictly relative the the field of that particular competition, and with no discretion for determining the precise magnitude of the scoring differential.
It sounds like what you're proposing is that the judges are not allowed to share with the skaters and the audience any information about how good they thought the skaters were in terms of general skill level, only who they thought was better than whom, or approximately the same. Nor, it seems, is "much much better" allowed to have any more weight than "just better enough that they don't deserve to share a ranking."
There's no way for the system to know before the competition takes place whether the skaters will be spread about evenly in ability
1---2---3---4---5---6---7---8---9-
or, if not, how they'll be clumped
1--------------2-3-4----5-66666-77-
111-222-33-44-5-6--8888-
1-2-3-4--5--6----7--8---9-
Fixing the increments in advance means that the judges can't really judge these skaters in this event in a meaningful way.
Ahh, you've got me, here . You're absolutely right that the logistical problem of handing out purely relative scores in a competition is probably the biggest issue.
It was an issue with 6.0 rankings. It will be even more of a problem if the judges have to rank skaters against a large field on multiple parameters.
My very preliminary thoughts as to a possible solution (and I say again for the eighth time that this is just an idea that I've been playing around with for a pretty short time):
-Have a separate panel of judges (who can be members of the general judging pool, but who are designated to judge only the artistic components at this particular competition).
I like this in theory. In practice, if it means more total officials, it would probably only be financially feasible at the most important events.
The artistic rankings (and, hence, the artistic component scores) are punched in and tabulated at the end of the SP and the FS. Thus, the "overall" score will only be known at the end of that day's competition.
Television will hate this. Essentially it takes coverage of the sport back to the days before computers when no one knew the results until everything was tabulated by hand.
This involves some incremental cost. One could theoretically have this done by one panel of judges responsible for both the technical as well as the artistic scores (as currently done), but this would leave open the charge that the judges can "manipulate" the artistic scores to achieve a desired result. How to achieve this as a practical matter is a separate issue; I'm just focusing on the concept for the moment. It would actually be even better, I think, if the "artistic" panel didn't know what the skaters received in their technical scores, but I haven't fully worked out the mechanics of that .
As I alluded to earlier, judges don't know what the skaters received in their technical scores now. They may hear the numbers announced in the arena, they may even write them down if they want to keep track, but while they're assigning PCS for skater Q they have no idea what technical score skater Q will receive. At best they can make a rough estimate if they have a good sense of how much each jump is worth in the scale of values.
I think it's even conceivable that this type of scoring procedure might heighten the audience thrill factor, particularly in cases where the technical scoring is close, and the overall results very much hinge on the artistic aspects, which the audience generally feels they can more readily relate to.
So would technical scores and/or technical rankings be announced after each competitor, but the audiences need to understand that they're only hearing half the scoring in real time and the results might swap around dramatically when the other half of the numbers are added in? Cinquanta will hate that.
My own personal viewpoint is that it might actually be easier to correctly rank the skaters on artistic components if done after all the performances are completed, and with more than a few minutes to make an assessment per score (the subject for scoring is both complex and holistic).
Again, I think this might be true with 6 or fewer skaters. The larger the group, the harder it will get. Each additional warmup group probably multiplies rather than adds to the difficulty of comparing.
Scoring skaters against external benchmarks rather than against each other makes it much easier to be accurate with larger groups.
There's a reason why US regional and club competitions split events into qualifying rounds if there were more than 18 skaters in the 6.0 era and increased that number to 24 after the adoption of the IJS.
It might, for instance, cut down on the cases where a skater received low marks simply because the skater wasn't in the final group.
Quite the opposite, I expect.
Give it a try -- sit down and watch a competition of 12 or 18 or 24 skaters in a row and try to rank them, giving scores as you go along. Then try with another group of the same size taking notes but waiting until after the last skater to come up with your rankings.
How can you legislate the mental processes of the judges. If you tell them to keep track by any means necessary but not show the audience, they're absolutely going to be thinking things like "That looks like a 5.0 performance." How do you think they knew how to score the first skater in a 6.0 event? (There was a pause for the referee to share the median mark for the first skater with the panel so they could make sure they were all on the same page before proceeding to the second competitor.)The judges will still be able, unofficially and privately, to keep a rolling tabulation as the competition moves along, and he'll still be able to utilize almost all of the "outside" principles of assessment that he/she has always done. Just not any reference to any skate (either personally ideal or historical) outside of that competition in assigning ranks,
It depends on whether judges are allowed to reflect margin of victory within each of the components or whether they're locked into increments in the numbers the sizes of which have nothing to do with the sizes of differences in the performances.Under the above, I don't see how this would mean that base values would more often determine the results.
5 triple jumps
8 triple jumps
You're going to get different results if the judges don't have the discretion to say that the former was at least one triple's worth better than the latter on each of the artistic components.
Of course they probably wouldn't have put the latter 2nd or 3rd on those components, and on Skating Skills and Transitions as well, at Worlds. But, since they're from the same country, what about at their nationals? What if the latter skater really is the second best in that field at most or all of those components?
I'm not in any way going all Tom Cruise in proselytizing for this idea. It's just a musing that I thought would make for interesting discussion, possibly highlighting some of the issues along the way.
Interesting discussion.
I definitely recommend that you give your system a try yourself to work out the details.