Exactly as Ant says.
I think the point of the tech panel is to agree on what was done. Occasionally judges will miss something significant, including how many revolutions were performed in a given jump.
This is especially true in pairs when there are two skaters to watch -- each judge with one pair of eyes is bound to miss some details.
If we were to get rid of the tech panel and have judges identify elements and mark each one, then probably the concept of features and levels (for non-jump elements) would have to be ditched. Instead, judges could add bonus points for what they individually consider difficult features of an element (according to guidelines less stringent than those followed by the tech panel) and could add additional bonus points for good execution (= positive GOEs) and they could also subtract points for mistakes and weaknesses of execution.
This would simplify the scoring somewhat. But it would also mean all the decisions now made by the tech panel about whether or not to reward a feature with a higher level would be much more subjective than they currently are. It may be debatable whether more subjectivity = more room for disagreement between judges in this area would be a good thing or a bad thing.
Currently, when the execution of an element falls into what they call a "gray area" there is room for disagreement about what exactly was performed.
This includes added difficulty in spins, spiral sequences, etc., that just barely meet or just barely miss the requirements to receive credit for a given feature. Getting rid of the concept of "levels" and just letting judges subjectively reward extra difficulty along with rewarding or penalizing good or bad quality would eliminate the issue of borderline features.
It includes jumps on the borderline of deserving to be downgraded (and even if the cutoff point were changed from 90 degrees to 135 or 180 degrees, there would still always be some instances that fall very close to the borderline, whatever it is). If there's a base mark for, say, a double loop and a higher base mark for a triple loop, what would the base mark be for a cheated triple loop that half the judges call triple and half call double, even if they all rate it at a -2 grade of execution?
It includes fluke mistakes that might change the nature of the element being performed to something different from what was intended or what is required (esp. in short programs) or allowed, which can have a huge effect on whether the skater should receive credit for this element or for a later element.
If some judges see a solo triple jump with a slight turning hop on the landing and other judges see a triple jump-single loop combination, how might that affect the scoring not only of this element but of another jump combination in the program? That's going to be a problem for the skater under any judging system if it happens in a short program where only one jump combination is allowed (and must include at least triple and double jump at senior level), or in a long program if it results in more than the maximum of three allowed combinations being performed. It will also be a problem for whoever, or whatever computer program, needs to combine the element scores from the full panel of judges if half of them interpret the element one way and half another way.
If some judges see a combination spin with an accidental extra step at the change of foot and other judges see two separate spins, how can that element/those elements be consistently scored? If there's a maximum number of spins allowed (always the case in short programs, and they always need to be specific kinds of spins as well), what happens if some judges score too many spins and call them by several different names? Who decides which spins the skater should actually get credit for?
In the 6.0 system, if those kinds of errors happened in short programs especially, the judges who interpreted the errors as just simple errors on attempted legal elements might take small deductions, and those who interpreted them as resulting in the wrong kind of element being performed might take several severe deductions, which could result in vastly different required element scores from different judges for the same program. No standardization for the skaters, but freedom for each judge to mark what he or she saw however he or she interpreted it.
In IJS using a technical panel, the three members of the panel get to review the videotape, consult each other about what they saw and how to interpret it, and inform the judges of the final determination. There are clear rules about how to handle most kinds of situations and the review process allows the tech panel to develop a consensus on how any given example of a borderline element fits into those guidelines. There isn't really an efficient way for nine or more judges to consult and come to a consensus about what the elements were before finalizing their marks, so with no panel that task would have to fall to the referee or some computer operator position or to an equivalent of the technical controller operating after the fact and it would take that much longer to finalize the scores. Might as well forget about announcing scores after each skater, because the review and data entry process would take even longer than it did the first year of IJS when the tech panels were still getting the hang of it.
