Am I correct that the skating system would result in the following final placement order: 81, 82, 83, 84, 85, 86?
If so, it seems a tiny bit unfair, because if you add up the total points for each couple, the final placements would be 81, 82, 84, 83, 85, 86. It also seems more fair for 84 to place ahead of 83, since they got three 1st place marks and 83 only got one. And it also feels a little unfair that 84 got more 1st place marks than every other couple, but only wins 4th (or 3rd) place.
But I’m not going to question an established scoring system or propose any changes. I’m just thinking out loud. (Umm...typing out loud.)
Yes, 83 will beat 84, and 84 will be 4th.
I understand why it looks unfair, but IMO on deeper analysis it actually makes a fair amount of sense--and it has a hidden benefit.
********
The reason it makes sense to me is that the skating system seeks to identify
consensus. As an analogy, if this were a mathematical operation, it would be most similar to the
mode as opposed to
median or
mean. Mode is the least-loved measure of central tendency in general, but it is useful when you're looking for an explicit pattern in a dataset where the number of unique values is relatively small.
Anyway, with respect to couple 84, the reason they do not win is that
the judges can't agree on it. Three judges think they are the best, but the
majority of judges find them to be fourth or worse.
By comparison, nobody thought that Couple 81 was best, but the
majority felt that they were in the top 3--and it is only Couple 81 for whom they agreed on such a high placement. The next highest place they could agree on was couple 82, whom they unanimously agreed were in the top 4. Everybody else got polarized reviews that include two or more "bottom 2" marks.
*********
This is a system with plenty of apparent oddities, to be sure, but here's the hidden benefit:
it degrades the impact of individual bias. In order for biased marks to be decisive, you need a majority of judges to be biased in the same way. Couple 84 is the perfect example. If we take away their first-place marks, the rest of the judging panel thought they should be fourth at best. Even with three judges giving a (potentially biased) "first place" mark, they couldn't place better than 4th. In order to win, they would have needed at least one other judge to agree that they were in the top 2, or two other judges placing them in the top 3.