Monday, November 26

The Big Technical Problem with IJS, as Demonstrated by Amodio's TEB Short Program

This short program by Florent Amodio at Trophee Eric Bompard was able to earn him 25.87 points technically.

That score was reached with the following:
- A fall on a quadruple Salchow
- A fall on a triple Axel
- A doubled triple Lutz and subsequently no second jump for a combination
- A sit spin that did not get down low enough so the entire sit/change/sit combination received 0 points

What does that mean? More than half of his elements (four of seven) in the short program were essentially complete failures.

I don't have a problem with the way the judges scored the elements based on what the rules state. In my own scoring, I had him at 25.16, so less than a point difference.

My problem comes with the actual rules themselves. Let's take, for example, a popular program choice for men that don't do quadruple jumps: a triple Lutz-triple toe combination, and a triple flip.

3Lz+3T = 10.10 base points
3F = 5.30 base points
Total for two elements at base value = 15.40 points

Now take Amodio's wipe-outs on both the quadruple Salchow and triple Axel

4S = 7.50 points earned
3A = 6.35 points earned (with x1.1 bonus for being in second half of program)
Total for two elements = 13.85 points

I know there are going to be some people that argue the reason that we see scores so similar for these elements is to encourage quadruple jumps and still reward skaters that are able to rotate them, regardless of outcome (2010 Olympics, anyone?). I don't really agree with that, though.

In diving, for example, if a diver is very poor with their attempt, he still holds a high degree of difficulty but receives scores maybe in the 2's or 3's. So if a dive had a 3.0 degree of difficulty and the diver received final scores of 2.5, 2.0, 2.5, 3.0, 3.0 (after throwing out highest and lowest scores in a seven-judge panel), the diver would earn 23.40 points for the dive (the total sum of scores times the degree of difficulty is then multiplied by 0.6).

Now, take that same 3.0 degree of difficulty dive and say that the diver did very well with it, earning scores of 9.0, 8.5, 8.5, 8.0, 9.0. The diver has now earned 77.40 points for the successful dive. See the difference? Divers have opportunities in later rounds to build back points. Skaters have opportunities in later elements plus program components to earn points. Other sports are also much more severe with failed attempts at elements.

I've always been in strong favor of really penalizing unsuccessful elements in skating. It makes no sense to me that Amodio loses ALL of his points on a sit spin combination because his butt needed to be a few inches lower, but still earns 71% of the points on his failed quad Salchow.

Along those lines, it doesn't make sense to me that someone who triples a planned quad attempt (which is within the rules) still earns quite a substantial amount of less points than a completely failed quad. The triple Salchow is worth 4.20 points. Again, completely failed quad Salchow is 7.50 points.

What this tells me is that skaters who are nowhere close on landing certain jumps cleanly but know they can throw their bodies in the air for 3.5 or 4 revolutions should just do so, because they can still earn big points (obviously Florent Amodio is capable of the quad Salchow and triple Axel, as he did both in the free skate).


I would remove the automatic one-point deduction for falls on elements. Falls on basic skating or in-betweens can keep that deduction. But I would make it so that all falls receive the exact same amount of total points per the element attempted, at least with triples and quadruples.

Take the quad Salchow. If it is fully-rotated with a fall I would give the same value of an under-rotated (<) jump of the same kind, but at the next level down (here, a triple Salchow). This means a quad Salchow with a fall would now be worth 2.9 points, which is about 28% of the base value.

Re-scoring Amodio's technical elements with that kind of system in play, and you get the following:

4S (fall) = 2.9
FCSp3 = 3.44
StSq2 = 3.53
2Lz+COMBO = 1.41
3A (fall) = 2.3
CSSp = 0.0
CCoSp3 = 3.64
Total Element Score = 17.22

He doesn't lose the additional two points on his overall score from the falls on the jumps, so his total segment score would now be 53.48 (compared to the 60.13 he actually scored). I think that sounds much more reasonable considering the performance that was put out.

Going for the big jumps should definitely have rewards, but only when they are done successfully! Take the risk and succeed, it will pay off.


Anonymous said...

I've never read a blog written by someone as inexpert and out of the real world as you. Stop blogging as an ISU official judge.

Tony said...

Where exactly am I out of the real world, 'anonymous'? :)

Anonymous said...

I'm very much in favour of what you are proposing here, Tony. This phenomenon has been bugging me for quite some time.

It's one on the things the ISU didn't think through in their promotion tour for the quads.

In the beggining the first change was ficilitated by guys like Jeffrey Buttle, I think, wo was known for throwing in the Quad but rarely ever landing while still getting a decent amount of points.

To prevent that they implemented the severe - GOEs for quads while the + GOEs remained on the level of triples. That was ridiculous because the risk/reward ration was extremely skewed and it was actually easier to rack up +GOE with good triples than quads. In the end it was very difficult to distinguish yourself or get yourself a decent advantage by doing a successful quad.

Today, the situation was supposedly changed in facour of the quad, but, now it's equally difficult to distinguish yourself through quality quads. And you can't really go wrong in failing any big jumps as long as you rotate it. In fact, you could even argue that in order to play the points game to its fullest, it is a smart move and almost a requirement to put in all that stuff, whether you really have it or not.

The solution of penialising failed attempts while amply rewarding successful ones is the natural consequence to solve this problem. It might, of course, lead to a bit more conservatism in throwing out the high point jumps but I can live with that.

Tony said...

I totally agree about the idea of rotating the quads if you can under the current point system. Skaters can earn so many points by just getting their bodies all the way around for the required revolutions. Makes NO sense. It doesn't mean they are, in fact, anywhere near landing the jump(s) with any kind of consistency.

I am completely fine with skaters really getting rewarded by landing successful quads-- it should be that way. But in my GOE re-working a few posts up, I believe that completely failed jumps should only get about 1/3 of the value that originally started at. It isn't the end of the world then, but also really differentiates between a good jump and a complete miss.

Anonymous said...

>>Anonymous 9:44 AM, November 27, 2012
On the contrary, I sometimes wish that Tony WAS actually an ISU official judge!

Tony said...

Haha, thanks. I really want the original poster to come back and explain where I have all of my ideas wrong. I'm always up for learning new things :)