Tuesday, March 19

Even More Analyzing Chan and the IJS

I'm on a writing spree lately, and I like playing with numbers.

Imagine that each of the eight jumping slots in Chan's free skate were marred by errors with the severity of a -2 GOE (think a big step out or fall out on everything). Here is what his score looks like if you consider all of the jumps to be fully rotated and keeping the exact same GOE as he received on all of the non-jump elements in his Worlds free skate.

4T+3T 12.40
4T 8.30
3Lz 4.60
StSq4 5.60
CCSp3 3.73
3A* 7.35
3Lo* 4.21
3F+1Lo+3S* 9.60
FSSp3 3.24
3Lz+2T* 6.74
ChSq1 3.40
2A* 2.63
CCoSp4 4.29

Total Value of Eight Jumping Slots = 55.83 Points
(Total Value of Non-Jump Elements = 20.26 Points)

New Technical Total with Straight -2 GOE for Jump Elements = 76.09
(Program Components Total = 89.28)

Segment Total = 165.37

Chan received an actual segment score of 169.41 at Worlds, which included falls on two jumps (one that was called under-rotated), a double of third jump, and a major fall-out of a fourth. The difference between making a severe error (non-fall) on *every* jumping pass and the combination of mistakes he made is just 4.04 points.

Now, people have been writing to me and sharing their opinions on social media about how increasing the penalty for mistakes on the bigger jumps would result in the skaters not even trying the jumps (in other words, we'd go back to non-quad attempts circa pre-Vancouver).

Let's say a skater has every intention of completing a quadruple toe loop, but he actually doubles the jump. Go back up to the element listing for Chan's free skate and we will say that he doubled his second quad but then received a 0 GOE on every other element. Look at his score now:

4T+3T 14.40
2T 1.30
3Lz 6.00
StSq4 5.60
CCSp3 3.73
3A* 9.35
3Lo* 5.61
3F+1Lo+3S* 11.00
FSSp3 3.24
3Lz+2T* 8.14
ChSq1 3.40
2A* 3.63
CCoSp4 4.29

Total Value of Eight Jumping Slots = 59.43 Points
(Non-Jump Elements Total remains at 20.26 Points)

New Technical Total with Straight 0 GOE for Jump Elements and Doubled Quad = 79.69
(Program Components Total = 89.28)

Segment Total = 168.97

The above demonstrates to me why I don't really buy the whole argument that going for a tough jump, such as a quadruple toe loop, and missing it results in the skaters being penalized oh-so harshly. Actually, I think completing wiping out on the jump because you are dedicated to getting those four rotations in actually gives you a real benefit of the doubt compared to doubling the jump.

-2 GOE on every single jump in a very sloppy program results in just a 3.60 point difference from the skater who completes every jump at a base (average) quality aside from a doubled toe loop- which is still a clean jump.

A double toe loop is worth 1.30 points. Once again, under my proposal that a fall on any jump results in 25% of the base value, a fall on the quad would be 2.58 points. Almost double the amount you'd receive for popping the jump. That's enough credit for taking the 'risk' in my opinion.

Anyways, going back to that 3.60 difference in points between the two hypothetical programs. Now, we have to remember once again that even with all of these errors, there is no criteria for lowering program component scores because of the jumping mistakes-- all eight of them.

Some people are going read this and write to me saying something to the extent of, "Well if he's making eight mistakes in a program, he doesn't have any skating skills!" I know this is coming. But in reality, Chan still does have the skating skills, the transitions, the choreography that goes beautifully with the music, and he probably is still interpreting the music pretty well, too.

I think oftentimes people let audience reactions get in the way of the 'interpretation' component mark. Just because the audience is going wild, it doesn't mean that they fell in love with the way the skater interpreted their music (see Rogozine). Just because the audience is disappointed that the skater didn't hold it together technically doesn't mean that they skated right through the music.

My first article discussed the possibility of reducing the performance and execution component score by 5% or so each time the skater fell. But what about this situation where a skater gets -2 for every single element? No falls-- does that mean the skater should still get a 9.00 for performance and execution?

As I write this and think about my post the other day, I actually think that it would be better to still keep the components scores, for the most part, separate. Scratch the automatic lowering of the performance and execution mark after falls. Of course, if there is a direct correlation between all of the mistakes and the performance, then mark it down. Ideally, as I mentioned, I'd like to see different judging panels for the elements and the components, anyways.

After the above experiment, though, I do think that the severity of mistakes in general needs to be more greatly reflected in the technical mark.

Denis Ten received 56.85 points for his jumps in the Worlds free skate. This included a quadruple toe loop with over a +2 GOE, a triple Axel-triple toe with nearly a +2 average, another triple Axel that received +2, and five of his eight jumping passes in the second half. The only mistake was a doubled flip, which cost him 3.85 points off of the base mark and the world title-- but that's a story for a different time.

See the problem? 56.85 for seven jumping passes being fully-rotated and receiving a majority of +1 and +2 GOE's-- meaning they were good and very good jumps. Go back to Chan's jump layout with all -2's across the board. 55.83 points would have been earned.

1.02 points difference. Sure, Chan would have hypothetically been going for two quads and a triple-single-triple sequence which puts his *planned* content higher than Ten's, but the difference in the quality of these jumps would have been drastic! The marks and system say to fans and anyone else reading the protocols that the jumps only had a 1.02 point difference between the two skaters. That's the problem. Who cares about planned difficulty when everything becomes a train wreck technically?

After this little experiment, I strongly believe there needs to be a harsher penalty for mistakes on the GOE and technical side of the sport.


Anonymous said...

I'd like to hear your opinion to the comments given on your last post suggesting changes in the factors of the second mark to lessen the weight of the PCS in the total scoring. They sound pretty logical and could be easily be implemented, while your adjustments (while I don't disagree) would require a whole lot of tinkering with an already complicated system. What do you think?

Anonymous said...

Which of the followings should be penalized more severely?

1. a) A planned quad popped into a single
b) A planned quad popped into a double

2. a) A planned quad downgraded and fallen
b) A planned quad downgraded and stepped-out / two-footed
c) A planned quad downgraded but landed

3. a) A planned quad under-rotated and fallen
b) A planned quad under-rotated and stepped-out / two-footed
c) A planned quad under-rotated but landed

4. a) A planned quad fully-rotated but fallen
b) A planned quad fully-rotated but stepped-out / two-footed
c) A planned quad fully-rotated and landed

Currently, it is scored in the following order. Personally, I am fine with the current ordering, but would like to see severer penalties on 3-a & 3-b and 4-a & 4-b, respectively, by going back to the previously used scale in factoring negative GOEs on quads (multiplying by 1.6). (Note: There were no BVs for UR jumps at that time, but let's say something like 1.4 may be good).

Anonymous said...

I know it makes me a freak but I think falls, meaning on your ass falls, should count for zero points. I wonder how the competition would have turned out if that were the case. And as to how that would effect quad attempts, they could make the base points much higher, like 20 points. Therefore when you go up for that quad you're really rolling the dice but maybe it's worth it. But it would also stop people who really can't land them from trying them. I know this makes me more of a freak but I'd also rather watch a program full of popped jumps than seeing them crash to the ice repeatedly. I think that would happen too. But these splatfests full of ugly elements are created by this point system.

Tony said...

I don't think that the weight of the PCS should be tinkered with. Ideally, it works great when you think of the 50/50 idea. Some top skaters have come very close to 100 points for their TES, while 100 points is obviously the highest for the PCS-- if a skater were to get 10.00's across the board.

I think the PCS isn't the problem, but it's more having to do with an accurate technical score. I'm writing a new blog post as we speak with my newest idea.

Anonymous said...

"I don't think that the weight of the PCS should be tinkered with. Ideally, it works great when you think of the 50/50 idea. Some top skaters have come very close to 100 points for their TES, while 100 points is obviously the highest for the PCS-- if a skater were to get 10.00's across the board".

Tony, but consider the following.

Imagine, we need to find out how much money (in American dollars) is the sum of:

¥1000 (Japanese yen) + € 200 (euros)

What are we going to do first? We'll check the exchange rates for both currencies, turn yens and euros into American dollars, then add the numbers together.

IJS's answer to the problem is:

¥1000 (Japanese yen) + € 200 (euros)= $ 1200 (American dollars)

Just as:

¥1500 + € 100 = $ 1600

I'm saying
TES100 =/= PSC100
There's no 50/50 and there's never been.

Gemma said...

Wow this is really interesting. What do you think of the argument that marks for transitions and choreo should be less (or even less) after a 'fall on your ass fall' - surely getting up off your ass isn't the ideal transition out of the jump. Some people have suggested that Chan's choreo and transitions marks were still too high. What do yo think?

Anonymous said...

I disagree with the concept that the elements have absolutely nothing to do with the performance. I think it does in the audience's mind and in the mind of others.

And I'm not talking in terms of negatively. When Mao Asada landed a triple axel/double in the short program at the Olympics, how does that not positively affect the overall understanding of her performance, I think it does.

And I think having -2 on every jump or doubling just about every jump like Lepisto did, should effect the overall understanding of the program.

In general I think the problem with IJS is its too much math and it gives the judges very little opportunity to evaluate the whole of the program. and I think there needs to be a way for the judges to do so. But to be fair it shouldn't be just falls taken into account.

And yes I think if your falling on your footwork, IMO it should affect skating skills scores. If your falling on a bunch of your jumps than how were the transitions going into your jumps successful, surely transitions should go down as well.

Tony Wheeler said...

We have to remember that crowd excitement has nothing to do with performance/execution.

By ISU criteria, it is 'the evaluation of the skater's/couple's ability to exhibit a pleasing appearance through body awareness and projection.' Included are carriage, style, body alignment, variation of speed for singles skaters.

So a skater lands a triple Axel. Another skater wipes out on a double toe. Body alignment refers to the shape of movement through the music's phrasing-- not how much time is spent perpendicular to the ice (ie not falling). None of the criteria have a correlation with mistakes, unless the carriage somehow starts to suffer or become lazy, or if the skater gets moody and skates right through his/her music-- which I've never seen someone like Chan do.

I really think people are just taking their own interpretations to the components marks and looking for reasons to hold skaters down when they do have mistakes. Shouldn't we be looking at the first set of marks and figuring out why there is still so much credit given for complete failures or near-failures of jumps?

Same with skating skills-- it has nothing to do with 'amount of time not spent falling which is then factored into a total percentage of a program', which is the vibe I'm getting from some people and their belief of what it is.

physics girl said...

No no no, popped jumps are no good either and should not be rewarded. It is much more difficult to rotate a triple than a double, so even if a fallen triple may be more disruptive than a landed double, the skater has to get credit for the increased difficulty. Otherwise any pre-novice kid that can do doubles would suddenly have a chance at the senior level.

Tony said...

The only way a pre-novice skater with only doubles is going to have any chance is if the senior level skater either makes an error on *every single jump* or pops all of the jumps-- and then they are doing the same technical content anyways.

I used Denis Ten as the example of a skater falling 5 times in a program. How many times do we actually see that happen? Not much. And even with his 5 falls, he is only scoring EIGHT points less with my proposal than what the current system gave him.

I am still getting the idea that most people think that falls are going to make the skater end up with half of what they are receiving now, but in reality it's only a few additional points.

Anonymous said...

By ISU criteria, it is 'the evaluation of the skater's/couple's ability to exhibit a pleasing appearance through body awareness and projection.' Included are carriage, style, body alignment, variation of speed for singles skaters."

And once again, how pray tell is someone who is falling all over the place presenting a pleasing appearance, good coverage, and good body awareness.

To me Tony, P/E in some areas is one place where the audience absolutely CAN have a much better idea if a skater is hitting the mark or isn't hitting the mark. They may not be able to judge the difficulty of choreography, the strength of skating skills, but they can at least tell when someone presents a pleasing perfromance. And they can absolutely tell in a lot of cases very bad execution.

That's the problem falls, stumbles they are MESSY. Chan was extremely messy in that skate, there was nothing "pleasing appearance wise" about that skate.

Anonymous said...

Multiple glaring errors out to be published precisely because they are glaring enough that the audience can see them. They are disrupting and break the flow of the program. You can still have a great performance with one fall, and some minor errors.. But multiple glaring mistakes, they effect the tone of the program.

For the sport to brush aside things that are obvious mistakes even to 3 year olds is ridiculous. Nobody is saying that one fall your automically out, after all nobody's complaining about the Russian pairs win. But people are screaming about errors that pile up and are getting brushed aside. I still think Chan gets high SS marks when he falls, but I think his SS marks should be higher when he lands his jumps. I could live with him still getting around a nine/high 8 in SS, but Perormance/Execution. I'm sorry I draw a line

Anonymous said...

And for the record its not about Chan only. I don't think any skater should be getting high P/E marks when they fall even my favorites.

Anonymous said...

And I'm going to add this too. Everyone keeps on saying well Patrick's choreography is so difficult, its so transition filled. Well here's the thing Patrick's programs are clearly TOO difficult for him. He has done his long program cleanly or even really close to cleanly ONCE this season. What every happened to quality of quantity. I appreciate good choreography, good transitions, and great difficult jumps. But to me skaters should also be looking at what they can realistically do well. And that goes with everything, speed is great, deep edges great, but you still need to be able to control it.

I'd rather see a skater win with less quantity, but still having content that is done with quality. Than a skater winning with programs that are unwatchable because they are peppered with errors. Execution clearly matters a hill of beans in this sport.

It was wonderful seeing Kostner slowing down, even doing lesser content. She was 100% more watcheable.