What determines the difficulty of a jump combination?

[Alchamei]

Since we have the low season (playing with words a bit seeing the discusion in stupid questions thread), we simply need to discuss our favourite programs and costumes or speculate about next season. However, seeing the videos before 2000s, some ideas came to my head.

The Surya Bonaly SP from Nagano started this. Scott Hamilton talks about the issue here: https://youtu.be/HFrzG9p_Xs0?t=55s. Surya had trouble with her triple Lutz due to her injury and she replaced the 3Lz-2T with the 3T-3T. The combo is well done, however, Scott says it is the easier combo and the judges will reflect that in their marks. A lot of people rant about that in the comments, saying: How could be a 3-3 easier than 3-2? But other people say that the 3Lz on outside edge is harder to perform, even though only with a double Toe.

However, there is a similar case in 2010. Yuna had a 3-3 in the SP and Mao only 3-2. It would make sense if Yuna scored more points. But Mao had the most difficult triple jump, a triple Axel, rarely performed by ladies. Yuna had also difficult triple Lutz-triple Toe. The base values of these combos were almost the same though, and Yuna's combo even had a higher BV by 0.5.

What is your opinion on that? Is a combo with two easier triple jumps more valuable than a combo with a very difficult jump and a double? Do the BVs reflect the difficulty correctly? Are there things that could be improved?

 

[begin]

This is a nice question for the skaters in this forum. A lot has changed since Surya Bonaly's time, though (ie. every 3-3 is worth more than every 3-2 now).

Ultimately there aren't many complaints about the current BVs and I'm personally okay with them. I've seen people suggest increasing the value of a 3Lo in combination because it's so rare, but I'm not really on board with that either. Most skaters have very strong solo loops but it seems like a matter of risking injury to train it in combination (and we don't need any more of those).

 

[StitchMonkey]

To keep this simple

I think a 3Z+3T combo plus a 3T on its own should be worth more than a 3T+3T combo plus a 3Z on its own. Right now they are worth the same, and no I don't really like that.

I do want to add I don't mind 3T+3T as a combo. I can appreciate the same jump aspect, and if the skater can really rock it, use it. But skaters should be picking it over harder combos because it is a money combo for them that they can nail, not because it makes sense points wise. I don't mind if a great 3T+3T gets more total points due to GOE than a messy 3Z+3T, but there should be some base point incentive/reward for doing the more difficult combo.

Blades of Passion has done a very elegant write up of a proposal for how this type of thing could work. I believe the simplified idea is that combos get a small bonus based on the value of the first jump of the combo. I am fine with that, but also wonder if just listing values for the combos might be almost easier than dealing with math.

 

[OS]

Interesting Question.

I agree 3A and 3A combos should have received higher rewards since Vancouver, but the system also neglected rewarding difficult combos, with appropriate 'proportional' quality rewards.

I have always wondered during the inception of COP, how they come up with such scale value in the first place. On what basis. The way the system have 'evolved' for more than a decade (somewhat federation incentivised manipulation imo), does the 'NEW' value reflect difficulty/risk/reward of the tech themselves?

e.g 3T3T has increased by 0.6 since 4 years ago. What was its BV during inception in 2002? Was it lower or higher?

New Base Value scale (3T= +0.2, 3S= +0.1)

3T3T performed in the second half = 9.46 (first half 8.6)
3Lz3T performed in the first half = 10.30
3F3T performed in the first half = 9.60
3S3T performed in the first half = 8.60
3A performed in the first half = 8.50
2A3T performed in the first half = 7.60

Out of the 6, which is the easiest to perform/easiest chance of getting the highest GOEs, which is harder/greater risk?

When you add second half bonus, what are the likely chance of
Liza/Adelina do a 3T3T + 2.1 GOE (+3) = 11.56
Liza/Adelina do a 3T3T + 1.4 GOE (+2) = 10.86
With halfway 10% bonus Mao 3A + 2 GOEs (+2) = 11.35
Without halfway 10% bonus Mao 3A + 2 GOEs (+2) = 10.50

Which has a smaller % of chances happening?
Which is the most undervalued triple?
Which is the most undervalued triple combo?
Which is the most overvalued triple?
Which is the most overvalued triple combo?

I never understood why GOEs reward for 3T3T = Same as 3Lz3T? Why not incrementally according to difficulty, otherwise what is the point risking a difficult combo that always had higher risk and difficult to get high GOEs. One thing for sure, next year's jump layout is going to be very interesting. I expect first half lots of slow music, time wasting meandering poseography, minimum choreography, then 2nd half jump element stacks for maximum number crunching.

 

[drivingmissdaisy]

I never understood why GOEs reward for 3T3T = Same as 3Lz3T? Why not incrementally according to difficulty, otherwise what is the point risking a difficult combo that always had higher risk and difficult to get high GOEs. One thing for sure, next year's jump layout is going to be very interesting. I expect first half lots of slow music, time wasting meandering poseography, minimum choreography, then 2nd half jump element stacks for maximum number crunching.

I think PCS can also be helped by doing more difficulty. Yuna worked very hard to upgrade her 3F-3T to a 3Lz-3T but there was no change to her net base value when you factor in that she would do the other jump solo. In fact, there is a disincentive to doing the harder combo because you can potentially struggle to get as good GOE. I'm sure the thought was the 3Lz-3T sounds more impressive and would be rewarded elsewhere in the marks.

 

[Mathman]

I have always wondered during the inception of COP, how they come up with such scale value in the first place. On what basis.

They just ranked the jumps in order of difficulty and added half a point for each one. 3T=4.0, 3S=4.5, 3Lo=5.0,3F=5.5, 3Lz=6.0. Combinations are worth the total of the two jumps, just like now.

The way the system have 'evolved' for more than a decade (somewhat federation incentivised manipulation imo), does the 'NEW' value reflect difficulty/risk/reward of the tech themselves?

I think that was the intent. They decided that a 3S wasn't really any harder than a 3T and so brought them closer together. They decided that a 3Lo was a closer in difficulty to a 3F, while a 3Lz was quite a bit harder than a 3F, especially if edge calls are going to be penalized. Each year they jiggle them a tenth of a point here and there. Federation politics can never be ruled out, but I don't think there is any big conspiracy to tailor the points to the skills of any particular skater.

They go back and forth as to how much 3As and quads should be worth, compared to the lesser jumps. On the one hand, "Whoa! A Quad!" But on the other, you don't want quads, no matter how difficult they are, to score for so much that nothing else counts (spins, for instance).

Personally, I am OK with combinations not getting extra points for being combinations. The extra reward that you get is that you have more opportunities to pick up points elsewhere. In the example given above, true, 3Lz+3T and solo 3T is worth the same as 3T+3T and solo 3Lz. But still the harder combination is better because after the 3T+3T, that skater is maxed out. The best she can do is solo 3Lz. But the skater who goes 3Lz+3T has an opportunity to gain more points by upgrading the solo 3T to a 3Lo or 3F.

Also, if you do a 3T+3T in the long program then you can't also do, say a 2A+3T. But the skater who led off with 3Lz+3T can. So you really do get a bonus, at least in potential, by doing a harder combination, even if that doesn't show up in the base values for the combination itself.

 

[Blades of Passion]

Personally, I am OK with combinations not getting extra points for being combinations. The extra reward that you get is that you have more opportunities to pick up points elsewhere. In the example given above, true, 3Lz+3T and solo 3T is worth the same as 3T+3T and solo 3Lz. But still the harder combination is better because after the 3T+3T, that skater is maxed out. The best she can do is solo 3Lz. But the skater who goes 3Lz+3T has an opportunity to gain more points by upgrading the solo 3T to a 3Lo or 3F.

The problem is this forces the skater to make their program even harder simply to get the reward of the 3Lutz+3Toe. It's only even worth .8 more points in base value to upgrade from 3T to 3Lo now, which is instantly erased via +GOE differential if the 3Lutz+3Toe is executed at "only" +1 GOE quality, as compared to doing an easy breezy 3Toe+3Toe.

Or someone like Gedevanishvili who has done 3Lutz+3Toe and 3Sal as her Short Program setup, so 3 different Triples and a much harder combination, actually gets slightly less points now than someone doing 3Toe+3Toe and a 3Lutz (assuming the solo Triples both come in the second half of the SP). Heaven forbid someone doesn't abuse the silly halfway mark SP bonus either, like Carolina Kostner with a beautiful 3Flip+3Toe and 3Loop in the first half...that setup is worth a half point less than 3Toe+3Toe and a 3Lutz that comes at 1min25seconds in the SP.

 

[Mathman]

Here’s what I think. No system of laws can possibly be sufficiently rich in detail to take into account every possible variation of human behavior. That’s why we have human judges.

Same with figure skating. You and I (human judges) can see the merit in Cedevanishvilli’s jump layout. Robots consulting a formal list of point equivalents cannot. Even if the most wonderful point system in the universe (yours, for instance ) were in place, there would always be someone who did something cool that we hadn’t considered and that maybe deserves an extra tenth of a point just on general principles and in defiance of the letter of the law. And there would still be skaters who would sacrifice composing a great program if instead they can game the system and squeeze out another point.

What if a triple loop right there suits the choreographic theme and the musical structure better than a triple flip? Too bad. You have to do a triple flip if you want to win.

For this reason I cannot get too exercised over whether a triple toe should get 4.1 points or 4.2, or whether a difficult combination should get more base value than the two jumps separately. I do not think that either choice would result in better programs or in more satisfactory judging.

 

[OS]

With all due respect Mathsman, while what you said is valid in principle of defending 'no system is perfect', but ultimately that is not what we are talking about here.

There are reasons why there will always be the need for system engineers under any human/informational/social/political system to ensure flaws, bugs, misuses get identified and get fixed as quickly as possible. It comes with the territory since the invention of any code/computing, there are hackers. In any case, we are not talking about some amazing complicated scientific algorithm that nobody get affected by, cared for or understand. It is more like there are some blatant disregard to major flaws that many people knows about for years, that get exploited, but it is only the new 'fixes' (such as + 0.66 for 3T3T after half way from 4 years ago) get done that made little sense. Is 3T3T all of a sudden more difficult? More rare? More risky? Undervalued in quality? Who decide? Why did all the other jump combination BV losses 0.66 in value all of a sudden on a relative basis? Especially given how GOE is rewarded, the difficult combos looses more than just BV but also lost on risks rewards.

Perhaps a more valuable question is NOT what determines the difficulties of a jump combination, but who determines the difficulty?

If the objective of Code of Points is to provide true fairness/impartiality/improvement to the sport as a whole by deriving appropriate points to quantify quality, difficulty, intricacy, then it has failed based on past 4 years of record where little actual change has been paid on these 3 areas other than: depreciate the importance of Quality (Reduce GOEs 30%), Difficulties (disproportionate values to jump difficulty, easiest jump combination got the biggest boost), Intricacy (PCS judging often questionable, abused, but goes unchallenged)

If the Code of Points is to allowing some way for ISU influence certain outcome with plausible deniability, by over rewarding certain elements, repress certain elements, the code of points system ends up provide a very powerful tool to meet their agenda.

If the objective is the first reason, then it would have been done by a panel of specialists/experts in a matter of open auditing system taking view from all sides much like a focus group, full disclosure with reasons why, how, when, what. By those has no conflict of interest with those who benefit and loss from such rule changes.

If the objective is the second reason, then it would likely to be done 1 or 2 people to carry out ISU agenda/incentive more than anything else behind close doors. The answer is in the pudding.

 

[leafygreens]

actually gets slightly less points now than someone doing 3Toe+3Toe and a 3Lutz (assuming the solo Triples both come in the second half of the SP). Heaven forbid someone doesn't abuse the silly halfway mark SP bonus either, like Carolina Kostner with a beautiful 3Flip+3Toe and 3Loop in the first half...that setup is worth a half point less than 3Toe+3Toe and a 3Lutz that comes at 1min25seconds in the SP.

Does anyone think that eventually, all skaters will be putting their jumps in the second half, and then the IJS will adjust to ban a second-half bonus?

 

[LRK]

Does anyone think that eventually, all skaters will be putting their jumps in the second half, and then the IJS will adjust to ban a second-half bonus?

Or they might devise some rule as to how many jumps you are allowed in each half...

Note: I'm not saying they ought, just that they might.

 

[solani]

Does anyone think that eventually, all skaters will be putting their jumps in the second half, and then the IJS will adjust to ban a second-half bonus?

Like Medvedeva ... for me that's poor choreography, it should be penalized. It's maybe ok if the music builds, but only then.
What I like about CoP is that we're seeing a variety of combos now. Before CoP the skaters had to do a certain combo to win, like f.e. 3Lz-2T for ladies and 3A-2L or 3A-3T for men.
What I don't like about CoP is, that we haven't seen a quad-triple combo from the last to male world champions. They're certainly capable of doing them!

 

[Mathman]

Is 3T3T all of a sudden more difficult? More rare? More risky? Undervalued in quality?

Possibly the ISU felt that a 3T+3T combination was undervalued in the past, and the current value is what it should, in fairness, have been all along. It used to be 4.0. Now its 4.3. Which is right? Which is wrong? Which requires justification and which is, oh well, it has to be something so why not this?

To tell the truth, I am not alarmed by the increase in base value for the triple toe. A skater who does 3T+3T gets 0.4 points more this year than last.

So does the skater who does 3Lz+3T and solo 3T.

So does the skater who does 3F+3T and 2A+3T.

Possibly the purpose of the change is to encourage more triple-triple combinations altogether. All of this is fairly arbitrary anyway, IMHO. – by what calculation to we decide that a triple Lutz is exacty 50% harder than a triple toe, rather than 40% harder?

 

[satine]

Does anyone think that eventually, all skaters will be putting their jumps in the second half, and then the IJS will adjust to ban a second-half bonus?

For the SP maybe. Not for the FS. Who can do all of their jumping passes in the latter half of a free program? :shocked:

 

[drivingmissdaisy]

We have to also remember that, if there are quirks in the scoring system that allow a skater to earn more points than we feel they should, any skater is able to take advantage. Caro is capable of doing a 3T-3T and a 3Lz so if that scores more than her other layout she should consider it. I agree with MM that it is somewhat arbitrary but there will never be complete agreement on the values. I agree with BoP that the halfway bonus in the SP is silly; the same jump done at 1:30 in the SP is worth more than it is at 1:30 in the LP.

 

[gkelly]

Like Medvedeva ... for me that's poor choreography, it should be penalized. It's maybe ok if the music builds, but only then.

If it's poor choreography, then it could be penalized with a lower Choreography component score. As you say, it depends on the program.

 

[Mathman]

For the SP maybe. Not for the FS. Who can do all of their jumping passes in the latter half of a free program? :shocked:

All is a stretch. But I think the standard will be what Hanyu won the Olympics with. Three big jumps right at the beginning. Fool around until the second half begins. Then 5 jumping passes in a row. Then the rest of the program.

The whole point of the second half bonus was to get skaters to spread their jumping passes out throughout the program (for purposes of program values and choreography), rather than bunch them all up at the beginning. It didn't work out exactly how the ISU hoped.

 

[Mathman]

We have to also remember that, if there are quirks in the scoring system that allow a skater to earn more points than we feel they should, any skater is able to take advantage.

This is not a positive feature of the scoring system, however.

 

[OS]

I agree 0.4 is meaningless had it not work on a relative system. The problem of boosting the easiest triple combo by 0.66 from 4 years ago, it also removes 0.66 advantage from everything else. Or 0.3 /0.33 lesser for every other triples. The fact 3t3t in the second half now have a high chance of outscoring a decent 3A, 3lz3t, 3f3t make the sport anti competitive for anyone without reputation points and strong federation pull, even if they do have the tech and artistic goods.

By equating the easiest tech with the hardest. It makes it virtually impossible to compete if you are outside the top 6 even if you bring difficulty, intricacy (PCS not work as it suppose to, takes a few competitions until judges get it right. Often it is like tossing a coin on the day or depends on what judging panel you get, where you are competing, and who against.), quality (pray. remember to smile and wave).

By depreciating GOEs to reward quality variability, bridging gaps of rewarding difficulties, ignoring PCS uncertainties, It all points to iron out traditional competitive variables such as difficulty, quality and intricacy that ladies have used to 'compete' in the past, that was the main determinant factor when judging 6.0 (basically the stuff outside judges 'endeavour' to predetermine outcome.) It doesn't affect the men because they have quads and 3A adjusted scale of value. The point system for the ladies based on their current scoring setup has became less about what rewarding what skaters can bring onto the ice than empowering the judging panel to better determine who they want through PCS and GOEs. This is a conflict of interest considering ISU governs the judges, and major federations all have their own judges and judging interests jousting at these top elite competitions, all of which has less to do with the skating or the progression of the sport (to push for faster, harder, stronger, evolution, revolution).

 

[solani]

We should not forget that the temptation to use performance enhancing substances or methods has certainly increased because of the bonus for jumps in the 2nd half of the program. I don't think that the ISU has ever tought about that.