Another unexpected week on the Grand Prix is in the books. The big news was the performance of two-time World champion Javier Fernandez, who stumbled to sixth after struggling with a stomach illness that kept him from full form. On the other hand, pre-competition favorites Gabriella Papadakis/Guillaume Cizeron and Wenjing Sui/Cong Han delivered stunning performances to take the dance and pairs titles. Alina Zagitova overcame a fall in the short and vaulted to take gold in the ladies' event.
A reworking is in order
Reflecting on the first part of the season so far, there's a consistent theme emerging as the sport is getting more difficult and there is more risk on the line every time a skater steps on the ice. I've written about this before - since its inception, the ISU Judging System has had to deal with how to properly balance risk and reward. As refinements to the IJS have been made, base values have been reshuffled (see: quads), elements have been redefined (see: step seqences), and features have been reimagined to reward originality over repetition (see: Biellmann variation).
But the IJS has always relied on Grades of Execution (GOEs), a more-or-less rigid, discrete-value system that provides same sets of deductions and additions for large swaths of elements. For example, a +3 GOE for every triple jump except for the axel maps to a +2.10 addition to the base value, whereas a -3 GOE maps to a -2.10 deduction; a +3 GOE for quads except for the axel maps to +3.00, where as a -3 GOE maps to -4.00.
I have never particularly understood the reasoning there. Essentially, falling on a cleanly-rotated quad lutz nets a skater 9.60 points (13.60 base value minus 4.00 for execution due to the fall), and even after subtracting the extra mandatory one point for the fall, the skaters is still left with 8.60. As a point of reference, the base value of a triple axel is 8.50.
This past weekend, some were miffed at Alina Zagitova's score in the short program after she fell on her combination. Program components aside, it really had a lot to do with what she got for that combo. The IJS is saying that, by putting her lutz-loop in the bonus and fully rotating it, she got 10.11 points for the element (11.10 base value x 1.1 bonus = 12.21 base value minus 2.10 for execution due to the fall).
Compare that relative to Gabrielle Daleman's toe-toe, which received almost perfect execution marks - that element got 10.60 (8.60 base value plus 2.00 for execution). Or compare that to Elena Radionova's averagely-executed lutz-toe, which got 10.50 (10.30 base value plus 0.20 for execution). It's tough to say that those three combos should have been given about the same score, even when taking the lutz-loop's difficulty into account.
When a difficult element is performed well, it should be rewarded well. Decreasing the base value of difficult elements is only going to disincentivize skaters from doing them, because they wouldn't be worth the risk. But the IJS currently doesn't disincentivize mistakes the way that it should. If a skater knows that they can fall on a quad and get a *higher* score than doing a triple well, then the incentives will point to the quad because there really isn't much to lose.
The judges here aren't doing anything wrong - it's the GOE system that creates the wrong balance of risk and reward.
A reworking is in the works
As I write this, the ISU is preparing to test out a revamped GOE system at Tallinn Trophy later this season. (The test will be behind the scenes with a different panel of judges.) It uses a wider GOE range of -5 to +5, where each "grade" is tied to a percentage add or drop in the base value, rather than a blanket set of deductions and additions that are given to a vast range of differently-valued elements. The percentage add or drop is still in debate, but it's supposed to be in the range of 10-15% per grade.
So that's to say that if a skater falls on a quad lutz, the element would receive a -5 GOE, which drops the base value of 13.60 to potentially 6.80 (10% of BV per negative GOE) or as low as 3.40 (15% of BV per negative GOE).
Of course, this system of -5 to +5 will still need to rely on judges giving positive GOEs when they are appropriate, not when there's a well-known skater on the ice. But if this GOE system is instituted, it is a huge step in the right direction for the IJS.
N.B.: I would go one step further and eliminate the late-program bonus for falls. After all, the idea of the bonus is to reward skaters for doing jumps when it is more physically straining to do them. But if they aren't landing the jumps, why incentivize it?
But will we revert to 2010?
So how do we not go back to an era of conservative skating after the recent advances in technical boundaries? Well, don't decrease the base values, that's the first step. The advances we have seen since 2010 were the direct result of the increased reward for the risk of harder elements. Skaters, for the most part, aren't throwing elements out there just cuz. They are throwing them out there because they have the percentages behind them in training.
But a stricter system of execution marks will necessarily decrease the instances of skaters throwing jumps out there just because it's a win-win situation (see the quad lutz example above).
Will this fix everything? Of course not. It will never be perfect, but incremental improvements are welcomed to none at all.