New IJS System Improvements

[bobbob]

I know an improved scoring/skating system will be implemented starting next year, and it is largely decided and they certainly don't care about what we have to say, but, what tangible improvements would you make within the current scoring system?

Here are mine, some are more significant than others. I'm not expecting some major new scoring system so they are assuming the general scoring protocol stays the same.

1. UR/e/level calls are currently basically based on one person's decision while two others with a say...instead of making this the responsibility of 1-3 people, which can sway a skater's score up to 7 points per element (in the case of an UR on a quad) and have major impacts on a skater's score, especially in ice dance, I think we should allow the judges have a say if the technical panel cannot come to a unanimous decision. Perhaps the judges could vote and the skater would get partial credit based on the number of votes e.g. if 5/9 say UR, then 5/9 * UR value + 4/9 * non UR value.
2. UR/e jumps should not automatically be relegated to negative GOE. If we want to penalize them more, than reduce the base value. But it should be possible to get good GOE on a flutz. Perhaps this combined with 1 would make the judges more likely to call out URs, prerotations, and wrong edges.
3. No second half bonus! If there is unbalanced composition of program, take off from GOE! But we should judge the skaters from the quality of the program they put out, not if they demonstrate stamina.
4. Right now combo and solo jump GOE factoring are the same. Perhaps give combos a little bit more GOE factorization and base value based on what the combo is. Also same for step sequences, they are worth too little.
5. Not really about scoring, but reduce the number of elements BUT NOT the length of the program. Give skaters more time to show skating skills and transitions and more time to breathe.
6. Penalize falls more! Only -1? At least -2 or -3.
7. In terms of PCS, tell the judges it is okay to differentiate by component. For instance a 10 SS skater can also have a 7 in transitions, etc. Of course they judges should be able to back that up when asked.
8. Find some way to reduce nationalistic bias by selecting the judging panel in a way to minimize judges from countries who have a vested interest in the event being judged.

 

[chopinskate]

I have formed opinions reading this forum.

1. Combo jumps should have multipliers. Also, 2x-3y combos should have a larger multiplier than 3x-2y combos. They should be scaled with perceived difficulty like that.
2. Rework GOE assignments for jumps. Large jumps with good technique should be rewarded more than small jumps with poor technique. The landings on larger jumps are difficult, too.
3. PCS and TES panel.
4. "Level" assignments shouldn't happen if it destroys choreographic purpose and beauty.
5. Measure pre-rotation if you're going to make UR calls.

 

[GGFan]

I know an improved scoring/skating system will be implemented starting next year, and it is largely decided and they certainly don't care about what we have to say, but, what tangible improvements would you make within the current scoring system?

Here are mine, some are more significant than others. I'm not expecting some major new scoring system so they are assuming the general scoring protocol stays the same.


2. UR/e jumps should not automatically be relegated to negative GOE. If we want to penalize them more, than reduce the base value. But it should be possible to get good GOE on a flutz. Perhaps this combined with 1 would make the judges more likely to call out URs, prerotations, and wrong edges.

8. Find some way to reduce nationalistic bias by selecting the judging panel in a way to minimize judges from countries who have a vested interest in the event being judged.

I agree with several of these but a question and a comment. Is #2 just about making the judges more honest or is it also about making the general fan more happy?

I think #8 would still leave a robust market in big feds "buying" or trading the votes of the countries with no vested interest.

 

[schizoanalyst]

Within IJS:

1 - More replay cameras with better resolution. None of this nonsense of putting the lutz on the far side of the rink where the camera can’t see it.

2 - Professionalize judging. They should have zero affiliation with the federations and the federations should have no say in the matter.

2 - Expand the GOE categories for finer granularity (they are already doing this by increasing it to +-5). Negative GOEs should have greater weight than positive GOEs. A bad step-out or tumble is more harmful to a program than doing a tano or intricate footwork into a jump is enhancing, yet the positive GOE is +1.4 and the negative is -1.4. I would double the negative GOEs, so, for example, a fall would get -4.2 for triples and -6 for quads. I'd get rid of the fall deduction for the first fall.

3 - Get rid of time bonus. If back-loading or front-loading is appealing choreographically, reward it in PCS (and if it isn’t, punish it in PCS). If an element is placed late in the program and a judge deems it difficult to perform it so late in the program, then it should be awarded in GOE. Difficulty of placing an element should be a GOE issue.

4 - Award difficulty of combination and elements. Generically, +3Lo combos are undervalued, 4T/4S/3T/3S values are too high (relative to 4Lz/4F/4Lo), 3A value is too low (relative to 3Lz/3F), spin GOEs are too low.

5 - Allow more freedom for skaters by allowing them to substitute for jumps or other elements with a spiral sequence or great additional spins or choreo sequences (or if they want, do more jumps). Adjust the values to make these choices competitive since programs are sooooooo boring now. Skaters need freedom to make it more interesting.

My real fix:

General impression scoring. There is nothing objective about the current system, so dispense with the arbitrary base-values and multipliers and additives. Give everyone a score out of 100. Weight the short as (1/2) and don’t weight the long, add the two scores together and whoever is highest wins. It carries over the beauties of the 6.0, but allows for the fact that differentials in scoring matter instead of segment placements.

 

[chopinskate]

Within IJS:

General impression scoring. There is nothing objective about the current system, so dispense with the arbitrary base-values and multipliers and additives. Give everyone a score out of 100. Weight the short as (1/2) and don’t weight the long, add the two scores together and whoever is highest wins. It carries over the beauties of the 6.0, but allows for the fact that differentials in scoring matter instead of segment placements.

This would be great, but I'd also like the judges to provide specific reasons for their scores. No 80s and 70s from two judges when the others are giving 50s. Or at least justify those 80s and 70s.

Some standard must be set.

 

[Shayuki]

2. Rework GOE assignments for jumps. Large jumps with good technique should be rewarded more than small jumps with poor technique. The landings on larger jumps are difficult, too.

This much is obvious. But what I'm curious about is what people value more - Big jumps with bad technique or small jumps with good technique? Because personally, I greatly prefer seeing small jumps done beautifully. In general, I wish more attention was paid to the technique. I don't think it even is a bullet point.

 

[GGFan]

General impression scoring. There is nothing objective about the current system, so dispense with the arbitrary base-values and multipliers and additives. Give everyone a score out of 100. Weight the short as (1/2) and don’t weight the long, add the two scores together and whoever is highest wins. It carries over the beauties of the 6.0, but allows for the fact that differentials in scoring matter instead of segment placements.

I may marry you just for this one comment! All of this fake math is killing me!!

 

[chopinskate]

This much is obvious. But what I'm curious about is what people value more - Big jumps with bad technique or small jumps with good technique? Because personally, I greatly prefer seeing small jumps done beautifully.

I think all four classifications can come into play.

4. small; bad
3. big; bad
2. small; good
1. big; good

 

[Shayuki]

Ah, also - I'd increase the value of quads and 3A for ladies specifically but reduce them for men. I get reducing them for men because everyone's doing them and it becomes more about trying to do as many as possible somewhat passably rather than doing all the jumps with high quality.

For ladies on the other hand, the risk very rarely is worth it to even perform a 3A. I really think that this makes the sport stagnate for ladies and is what lead to all the backloading in the first place - It's not like there are many other ways to improve one's TES. The step up to 3A is too steep to usually be worth the trouble and the reward isn't great.

I think all four classifications can come into play.

4. small; bad
3. big; bad
2. small; good
1. big; good

So they're valued equally? Alright, that's fair.

 

[chopinskate]

So they're valued equally? Alright, that's fair.

Hmm? No, that's how I'd rank them. I think small jumps might be simply because you don't have the most leg muscle strength, and so you shouldn't be penalised for small jumps, as long as you're doing them properly. But they are easier to land, so it's not as impressive to land them as a big jump done properly.

Personally I think such skaters should try to make them larger, by for example going into them with more speed, but maybe it's not always possible.

2 and 3 might be virtually tied, depending on what happens during the performance.

 

[Shayuki]

Hmm? No, that's how I'd rank them. I think small jumps might be simply because you don't have the most leg muscle strength, and so you shouldn't be penalised for small jumps, as long as you're doing them properly. But they are easier to land, so it's not as impressive to land them as a big jump done properly.

Personally I think such skaters should try to make them larger, by for example going into them with more speed, but maybe it's not always possible.

Ah, I see. So you actually value good technique more. Then we even agree there.

I also think there's an "acceptable" height and there should be diminishing returns afterwards. For example, Miyahara's jumps are below acceptable height. But apart from that, when the jump is adequately high I don't really see a huge need to make it even bigger.

 

[Koatterce]

whole post

1. The issue with this is that the judges are likely more biased than the technical panel, so the judges' biases/preferences will just come out in that too.

2. I think the reason they become negative GOE is because the skater did the jump incompletely/wrong, so they didn't actually execute it well (thus negative GOE). UR's already reduce the base value, and the final GOE does not necessarily need to be negative, but is definitely reduced (so if there are more positive elements, the jump can still get positive GOE). If you can get good GOE on a flutz, some skaters could just start doing flips but saying that they're doing a lutz so they can still get good marks without needing to properly learn another jump (and there'd have to be a significant decrease in BV to compensate for this)

3. I agree with this one. The program should be marked on the execution of the elements. And programs should be constructed so that the elements and music go together (e.g. not all music is suitable for front/backloading), and if it doesn't, should be reflected in the composition mark.

4. I agree with this one too. Doing jumps in combination is more difficult than as solo. Maybe the first jump retains its normal BV, and the second/third jumps have a small multiplier. Increasing BV of step sequences etc. should be increased too because they're becoming extremely overshadowed by jumps, so messing them up is not too bad as long as there are good/difficult jumps.

(left out 5 because I have no comment)

6. I'm not completely sure about this one. Because the fall deduction itself is only -1 but with the negative GOE on the element as well, it becomes a much larger deduction. With the upcoming +/-5 GOE, it will penalize falls more (but we'll have to wait and see exactly how). Also, in the case of fluke falls, penalizing them by more seems a little excessive.

7. THIS. is so necessary. plus all the increasing PCS due to arbitrary reasons like consistency or whatnot. I get that experience is needed to increase PCS, but it does not guarantee it. I'm pretty sure that hypothetically, if a skater skated the exact same program exactly the same way throughout the course of the season, their PCS score would rise (which makes no sense). If they improve their program over the course of the season, then great, an increase in PCS is warranted. but only if they actually improve.

8. I've been thinking about this whole judge bias thing, but it's an issue that's almost impossible to solve. Nationalistic bias is definitely a thing, but judges can also be biased towards/against certain skaters/countries even if they're not from the country (or the competitor's). But that's almost impossible to tell because it wouldn't be such an obvious reason as nationality. Plus, every judge is going to be biased regardless (e.g. maybe towards particular styles, or political reasons/country alliances). And you can't get neutral judges because every judge has to come from a country.

I think all four classifications can come into play.

4. small; bad
3. big; bad
2. small; good
1. big; good

I agree with this too. Technique is better than jump size but size should also play a role.

 

[bobbob]

I think all four classifications can come into play.

4. small; bad
3. big; bad
2. small; good
1. big; good

Like levels for jumps? That would not be a bad idea except we would need to make sure it looks for something independent of GOE. If height and technique is already in GOE, no point to count it twice.

 

[chopinskate]

Like levels for jumps? That would not be a bad idea except we would need to make sure it looks for something independent of GOE. If height and technique is already in GOE, no point to count it twice.

Technique, at least, isn't being accounted for in jump GOEs currently.

 

[bobbob]

Ah, also - I'd increase the value of quads and 3A for ladies specifically but reduce them for men. I get reducing them for men because everyone's doing them and it becomes more about trying to do as many as possible somewhat passably rather than doing all the jumps with high quality.

For ladies on the other hand, the risk very rarely is worth it to even perform a 3A. I really think that this makes the sport stagnate for ladies and is what lead to all the backloading in the first place - It's not like there are many other ways to improve one's TES. The step up to 3A is too steep to usually be worth the trouble and the reward isn't great.

So they're valued equally? Alright, that's fair.

I don't think we should do different base values for men and ladies, simply because not all skaters which the scoring system applies to is a top tier skater, and there are just as many men who will try hard and never get their triple axel as women. The risk is the same, the reward should also be similar.

And I don't know, the reward for a triple axel is pretty good, in my opinion. Up to 6 points compared to a double, and if you fall and rotate you still get more than a 2A. Obviously if you fall AND underrotate you should not be getting more than a double axel.

 

[Shayuki]

I don't think we should do different base values for men and ladies, simply because not all skaters which the scoring system applies to is a top tier skater, and there are just as many men who will try hard and never get their triple axel as women. The risk is the same, the reward should also be similar.

And I don't know, the reward for a triple axel is pretty good, in my opinion. Up to 6 points compared to a double, and if you fall and rotate you still get more than a 2A. Obviously if you fall AND underrotate you should not be getting more than a double axel.

If you fall and rotate, you get less than for a good double axel. That pretty much is where the problem lies. It's not worth the risk.

A 3A for a lady is much, much bigger a deal as well as more rare than a quad for a man, yet the reward is smaller.


Balancing should be done for the top levels. The lower end skaters will make do with what they get.

 

[QuadThrow]

There is the same thread every year when the end of the season has come.

I think changes should come from the Problems we have with the system.

1. Problem 1. T/M got incredibel high components for this bad candymen FP. How can we change that?

2. Medvedeva gets some GOE +3 for her Axel. Totally laughable. How can we change that?

3. There are just too many falls at mens' programs. I think the problem is solved a bit by reducing the numer of jumps next season.

4. Ice Dance competitions are currently decided by levels for the Step Sequences. Is this a real issue or not?

5. I think there should be no bonus for 3Lo combination. Zagitova won with this element without any bonus. She scored over 13 points.

 

[Shayuki]

1. Problem 1. T/M got incredibel high components for this bad candymen FP. How can we change that?

Very little about GOE has anything to do with the jump itself. Only like 2 bullet points do. The rest is all stuff like arm variation, timing to music, transition after, transition before. Only "good height and distance" and "effortless throughout" have anything to do with the actual jump, and it's possible to hit +3 GOE without hitting either of those bullet points.

The solution is to give more emphasis on proper technique for the jump in GOE allocation.

5. I think there should be no bonus for 3Lo combination. Zagitova won with this element without any bonus. She scored over 13 points.

She won on many things, but 3Lo combination over a 3T combination is only worth 0.8 points, or 0.88 when backloaded(And within the context of her program, 1 points or 1.11 when backloaded). You don't think it's too small a reward? "She scored over 13 points" is one thing, but she would have scored over 12 points doing +3T instead, and would likely have had better GOE for that jump combination because of how much easier it is - especially when it comes to maintaining flow afterwards(Many +3Lo combos get stuck on landing and get lower GOE).

 

[chopinskate]

Very little about GOE has anything to do with the jump itself. Only like 2 bullet points do. The rest is all stuff like arm variation, timing to music, transition after, transition before. Only "good height and distance" and "effortless throughout" have anything to do with the actual jump, and it's possible to hit +3 GOE without hitting either of those bullet points.

The solution is to give more emphasis on proper technique for the jump in GOE allocation.

The effects should also be compounded. What I mean is this:

FIRST see if the jump is done with proper technique. If yes, that opens your +GOE account.

Good air position? +GOE if yes.
Flow?
Musicality?
==========================
Effortless? +GOE if yes.
Is the jump large? +GOE if yes.
==========================
Are the turns completed in air before settling down? +GOE if yes.
Is the jump delayed? +GOE if yes.
==========================
Is the jump done through a difficult entry? +GOE if yes.
Is the jump done with some variation? +GOE if yes.

Going from the more important qualities to the less important ones.

 

[schizoanalyst]

If you fall and rotate, you get less than for a good double axel. That pretty much is where the problem lies. It's not worth the risk.

A 3A for a lady is much, much bigger a deal as well as more rare than a quad for a man, yet the reward is smaller.

I'd rather see a nice double axel than a fall on the triple, but I think it would be easier to just scale the 2A down (along with all the triples below the axel) than make the system even more convoluted. I've long thought the values of 2A/3T/3S were too high anyway for both sexes. I'd just scale the values for 2A/3T/3S/3Lo/3F/3Lz/3A as 2.5/3.3/3.4/4.1/4.3/5.5/8.5 (or something similar).