People complain about the judging all the time, but maybe there are so many numbers and so many categories, the judges themselves may have no idea what the final score will be for a skater (unless they can add in their heads), and may not know exactly who will be in first or second place.
As a teacher, I don't (or at least try not to) grade the papers to make certain that a certain final average is reached, or so that student A gets a higher average than student B (to get the subject award, for example.) I grade each test or paper separately, and may not know whose average is higher until I put them in the spreadsheet. That's what the judges are supposed to be doing--just judging each skater's performance, checking those edges and transitions and so on, without regard to anybody else, and then being sort of surprised by the results.
In reality, I don't think they are doing that.
In reality, I think they probably are surprised by the results for the reasons you would expect more than you think they are.
Heck, even if they were trying to manipulate a desired outcome, since they can't control or in most cases keep track of the TES, they would probably often be surprised by their own rankings of the skaters from combining their GOEs and PCS with the tech panels call as if they were the only judge on the panel.
The details of the math behind the TES are too complicated for judges to figure out in their heads. Tthey have no access to knowledge of what levels were called for spins and sequences. They can estimate, but they could easily be off by several points. And then they'd have to keep track of the factors for the PCS as well (easy to do with senior men, not so much with the factors for other disciplines).
And of course individual judges don't know how the other judges on the panel are scoring each skater, whether their own marks are likely to get thrown out or not.
I'm sure it happens quite often that a judge gives low PCS to a skater who didn't impress them and is surprised when that skater is announced in first place on the strength of technical content, or vice versa.
Still, I do not think that the ordinal approach can be completely eclipsed. At the end of the day, somebody wins, somebody is second and somebody third. The judges will give the higher marks to the skater that the judge thinks skated the best, as always.
As I say, the judges can't control the
Total Element Score, although of course they can control the GOE part by giving higher execution marks to the skater they think executed the elements the best.
We even read complaints that the levels are called wrong on steps at times so I will also call that subjective.
If you can say that a certain level call was "wrong," that implies that there is an objective "right" call that should have been made instead. So then discrepancies would not be due to subjectivity but rather to errors in calling.
Or you could say it's subjective in a different way -- different callers perceiving the line between the yes/no decision on certain features at different points in the analogue spectrum of each skater's execution. That might include certain callers being consistently stricter or more generous.
So you could say there is subjectivity of
perception. It's not a matter of
preference or
priorities in the same way that 6.0 scores or to a lesser extent PCS are inherently subjective.
It is possible to rank performances. First, second, third is a ranking. This is what ordinal judging called upon the judges to do.
I do not believe that it is possible to score a performance. In terms of choreography, say, this is a 6.25 performance. Period. It is not a 6.00 performance and it is not a 6.50 performance.
That is the what (nominally) the CoP asks judges to do.
I think the PCS scoring is a hybrid to a degree. The specific scores are more rough estimates rather than precise correlations between a given level of performance and a given number. For each well-trained judge the PCS numbers should be more consistent than 6.0 scores because of the nature of how they're being used, but to some extent they will always be a floating target.
I do think there is a value in reflecting not just who was better than whom, but also by how much -- roughly, if not precisely. I also think there is value in breaking down the areas in which each skater was better or worse than the others to more than just two summary scores (technical merit and presentation).
And short program scoring under 6.0 was similarly a hybrid, with floating base marks and presentation marks that judges set in relation to other skaters in the field, highly influenced by skate order, and the fixed mandatory deductions for specific errors.
If difficulty of elements and certain kinds of errors can be objectively quantified, but quality can only be judged qualitatively and comparatively, then any attempt to score both difficulty and quality will result in hybrid scoring of some sort.