Article:"Rewarding Failure Diminishes Sport"

[gkelly]

I haven't done the math, but I assume that under this year's rules Skater A's base mark is more than 12 points higher after sequence multipliers, and therefore Skater A would win, even with -3 GOE on three jumps and three fall deductions, assuming the GOEs on the other jump passes are the same.

And I would agree that that does not seem appropriate, if everything else is the same between the two skaters. The best way, IMO, to prevent that result would be to increase the negative GOE penalties for failed triple axels and quads.

HOWEVER, if Skater A was better in some other significant ways in addition to showing the ability to rotate more revolutions in the air at a time without landing cleanly -- e.g., if Skater A deserved high positive GOEs on the five successful jump passes while Skater B barely squeaked out most of the jumps, or if the spins steps, and/or program components were not equal -- then I do not have a theoretical problem with Skater A coming out ahead with those respective jump schedules/results.

Obviously, those who place high value on not falling for the sake of not falling would disagree.

 

[Mathman]

HOWEVER, if Skater A was better in some other significant ways in addition to showing the ability to rotate more revolutions in the air at a time without landing cleanly -- e.g., if Skater A deserved high positive GOEs on the five successful jump passes while Skater B barely squeaked out most of the jumps, or if the spins steps, and/or program components were not equal -- then I do not have a theoretical problem with Skater A coming out ahead with those respective jump schedules/results.

I quite agree. Add up the points. Whoever has the most wins.

The only area of disagreement is whether we should reduce the points awarded for the three flubbed jumps all the way down to zero or to some number greater than zero.

Obviously, those who place high value on not falling for the sake of not falling would disagree.

In an add-up-the-points system, I think it would be more accurate to say, "those who place a high value on landing the jump..."

 

[Mathman]

Your way, I think, would reward the front runners if they do it right, or would punish the best, the most gutsy skaters the most.

Isn't that the definition of risk? Do or die.

Yes, sometimes the gutsiest skaters will crash and burn. Sometimes they will soar. That's sports. :yes:

 

[gkelly]

The only area of disagreement is whether we should reduce the points awarded for the three flubbed jumps all the way down to zero or to some number greater than zero.

When you figure in the fall deduction, as it stands now the net effect of falling on a single or double jump, or a downgraded triple, is to subtract points from the total score. Only falling on a fully rotated triple or quad actually gains points. Falls on easier triples and downgraded quads earn next to nothing.

So I'd be OK with adjusting the penalties so that the positive value of rotating and falling on those harder jumps is negligible. However, I do think there is some value to giving more points for falling on a hard jump than for falling on an easier jump.

 

[Mathman]

^ Do the same rules apply to competitions at lower levels, where the hardest jumps attempted might be doubles?

Is there a difference between falling before or at the beginning of snow shoveling and after the snow has been cleared? Mathman could have realized how slippery the porch was and aborted the heroic attempt early thus avoiding any fall. Or he might not even bother to begin with.

How many hero points and how much hot beverage would Mathman deserve from various scenarios? Mrs M might be upset at any fall, which caused a commotion from the neighbors, and decide he deserved nothing if he fell. Would she be more upset that he never even wanted to shovel the snow? OTOH, she might reward him according to how much snow had been cleared before the fall, with extra cookies for a good shoveling job done without any fall causing troubles for her to deal with.

To me, that's the whole problem with the CoP in a nutshell. Did you do this little thing before you fell, did you do that? Did you do three-quarters of the jump correctly before you fell or only two-thirds?

Even setting aside the burden this puts on the technical panel and the judges -- doesn't there come a point where you have to say, this is silly?

 

[gkelly]

^ Do the same rules apply to competitions at lower levels, where the hardest jumps attempted might be doubles?

Yes.

I think the fall deduction should be smaller at lower levels, and I believe that is the case in Canada, but it's the same (1.00 per fall) at all levels in the US.

There are different rules about what is required or allowed in short and long programs at different levels, as well as between disciplines. The basic structure of the programs are similar, but some of the details are different, and there are shorter time limits and fewer elements at lower levels.

 

[Bluebonnet]

Here is a test case. Spins, steps, program components, etc., are equal.

Skater A

4T (fall)
4T+SEQ (fall)
3A
3A+SEQ (fall)
3Lz+3T
3F
3L
3S

Skater B

3A
3Lz+3T
3Lz
2A+half loop+3F
3F
3L
2A+2L+2T
3S

Who wins?

To emphasize the difference, let's leave out the five jumps that both skaters did. The remaining three passes are

A: 4T (fall), 4T+SEQ (fall), and 3A+SEQ (fall).

B: 3Lz, 2A+half-loop+4F, and 2A+2Lo+2T

I'm so impressed with skater A who has two quads, two 3As, and a 3-3. He's exciting. Skater B has only one 3A and a 3-3, also no quad. He's boring. Skater B definitely has a much reserved, not-on-par plan, either because he is not able to do more or he is affraid of failing - the opposit mindset of what a sport is all about. Skater B looks like a Junior skater. From the jumping layout, I'm willing to have skater A win over skater B for his gutsy act.

However, I'm also considering the overall effect of a program. Three falls doesn't look good. So I'm in favor of changing. I think that CoP should increase the deductions for the second and the third falls in order to help bringing out the winning program's better overall impression. To make the axel-chasing a little closer to skirt-chasing. Maybe -2 deduction on the second fall and -2 deduction on the third fall. So skater A's final result would be 2 points less than what he would have gotten under the current rules.

If after the larger deductions, skater A still wins over skater B, then be it. Skater B and his coach should blame themselves for the much too conservative jumping layout.

Isn't that the definition of risk? Do or die.

But if not rewarding them in some degree for the effort of trying, it will turn to "Don't Do then Won't Die". That is not sport.

 

[bekalc]

GENERAL QUESTION: How much should execution of elements factor into the final score as compared to difficulty of elements and program components?

I think that poor execution of the elements should be hit hard on the P/E marks. I'd make execution actually mean execution. And I'd also increase the value of P/E vs other components too. I think its vitally important. My issue is it right now I think PCS are far to based on reputation. I know you mentioned how skaters like Denis Ten's PCS don't change much. But one could argue Abbott's does, I've seen Jouberts. However some of the big, big name skaters not so much.

I get that reputation plays a role. But I think its getting to ridiculous levels. I look at results like 2008 Worlds with Kim vs Asada. Yes Kim was the best skater in the world, yes she skates faster than Asada and had harder choregraphy. But on that particular free skate Kim was lack luster, had multiple errors, and Asada was sparkling. Yet the PCS were higher for Kim. Now I've heard arguments that its because Kim had harder choregraphy/faster skater/better transitions and it was all true. But that then goes to suggest that PCS are pretty much at a set point. And that half the mark or even more is predetermined. I think that's a huge issue and its bad for the sport. Daisuke and Patrick at the GPF also illustrates that for me greatly...

Its not about a skater must never fall. I can live with a skater falling. I can live with a skater falling and beating a clean skater if the skater was that much better/had other things. I can even live with higher P/E marks for the skater who fell. But once we get into all kinds of obvious/glaring errors and the skater is getting sky high PCS. I think its a huge problem. Add in now for the men the huge points you get for falling on a quad, and the lesser known/lesser rep skaters can't even make up for their lack of PCS on jumps anymore.

Why bother watching?

And the judges are inconsistent with how they do PCS. Not top skaters like Ten Pcs stay around the same no matter what they do. Abbott who (goes in between) gets higher PCS when he's clean. But Patrick is at sky high no matter what Patrick does on the ice...

Tnere needs to be a system for PCS, but PCS need to be judged to where no you can't get super sky high PCS when your all over the place.

 

[jaylee]

I get that reputation plays a role. But I think its getting to ridiculous levels. I look at results like 2008 Worlds with Kim vs Asada. Yes Kim was the best skater in the world, yes she skates faster than Asada and had harder choregraphy. But on that particular free skate Kim was lack luster, had multiple errors, and Asada was sparkling. Yet the PCS were higher for Kim. Now I've heard arguments that its because Kim had harder choregraphy/faster skater/better transitions and it was all true. But that then goes to suggest that PCS are pretty much at a set point. And that half the mark or even more is predetermined. I think that's a huge issue and its bad for the sport. Daisuke and Patrick at the GPF also illustrates that for me greatly...

I think you mean 2010 Worlds, which was nearly 2 years ago, so I'm not sure if it's the best example. Kim's PCS advantage vs Asada was much lower than it was compared to what it had been at the Olympics and at TEB, so that goes against your statement that PCS was predetermined. And Asada wasn't perfect; she had a downgraded 3A and her other 3A that wasn't downgraded barely got any +GOE. Are you going to argue that Asada's 3A shouldn't have been downgraded? Because if it hadn't been, Asada would have easily won the FS. But if it deserved to be downgraded (a huge blow to her TES, since she didn't have a lutz or a 3/3), then how could Asada have won the FS? The only other way would have been to say, "Well, she should've gotten a bump here and a bump there in PCS to overcome the deficit of the 3A downgrade to make sure she won the FS." But if she didn't deserve a bump in PCS based on what she put out there on the ice, that's corrupt. Remember that this season, the judges didn't know whether or not a jump had been downgraded by the technical panel. I also think that another factor that influenced the PCS was that Asada and Kim skated in different groups. Had they skated back to back, perhaps the PCS would've been slightly different, maybe enough to make a difference in the rankings of the FS, but not enough to make a difference in overall rankings (as Asada and Kim were rightfully 1st and 2nd overall).

There's nothing to stop a skater, who, in getting low marks for TR, CH, and IN, to add transitions and choreography that better reflects the music to their program. That's what competing early in the season is all about, so you can get feedback from judges and adjust your programs accordingly. And on that note, Asada had both 3As ratified at 2010 4CCs and no major errors, and her PCS was less than 60. That was a huge, huge red light that the judges didn't like her FS. They also didn't like it during the GP. But...she didn't change it. Maybe it was too late, but you can't say that PCS is predetermined because skaters have the option of changing their choreographer and transitions from competition to competition. Michelle Kwan dumped her SP in the 2001-2002 Olympic season. It's been done before.

I like the fact that the IJS punishes cheated jumps and rewards rotation (and I do agree that there needed to be two types of downgrade rules, < and <<). But I've noticed that skaters with errors that aren't obviously visible to the audience or casual fan watching at home (i.e. underrotated jumps, or edge calls) often get more appreciation than their scores reflect; skaters who fall or have obvious step-outs but rotated their jumps are pilloried but may get higher scores than one may expect. While the scoring system is in the favor of those who rotate (and land) their jumps, audience appreciation tends to be in favor of those who land jumps, cheated or not. And that's where most of the debates about placement come from.

It would be helpful if, when the skater is in the kiss & cry and clips of their performance are being played, calls by the technical panel could be displayed over a clip of a jump or spin, showing +3 or -2 (edge call or downgrade).

 

[ImaginaryPogue]

bekalc, I did read your post and will respond to it in a new thread as I think it's an interesting subject worthy of better analysis than I can do right now.

 

[Mathman]

But if not rewarding them in some degree for the effort of trying, it will turn to "Don't Do then Won't Die". That is not sport.

Let's look at the actual effect of jollying the skaters along with partial credit.

In reality, not talking hypothetically about what might happen, all this partial credit and hand-holding for half-way attempts has created a group of skaters with a CoP mentality.

Let;'s compare the 6.0 mentality with the CoP mentality directly. Here (again) are the jump layouts for the top six skaters in the SP in the 2002 Olympics (6.0), versus the top six in the 2010 Olympics (CoP).

6.0

Yagudin 4T+3T
Honda 4T+3T
Goebel 4S+3T
Plushenko 4T (fall, intended 4T+3T)
Abt 4T+3T
Li 4T

Dambier also hit a solo 4T and Liu did a 4T+2T

CoP

Plushenko :4T+3T (you can't keep an old 6.0-er down )
Lysacek: No quad
Takahashi: No quad
Lambiel: 4T attempt, but flubbed
Chan: No quad
Weir: No quad

I am sure the CoP has many virtues, but encouraging quads by giving partial credit for failed attempts does not seem to be working.

 

[Mathman]

But if you get no points for something, the risk-reward benefit skews towards the "not trying it at all" side if it's not near 100%. Currently, the quad-fall risk-reward is worth enough points for the men to be trying it...

After all, you of all people should know, that IJS is a MATH game.

That is what I mean by the CoP mentality. The CoP gives you 5 hundredths of a point for raising your left hand and 6 hundredths of a point for raising your right hand.

OK skaters. Everybody raise your hand.

(OT. By the way I just looked at Plushenko versus Goebel in the 2002 Olympics. Plushenko still would have been second overall even if Plushenko had been placed in fifth behind Abt in the short program. So in this example it didn't matter.)

 

[skatinginbc]

Everybody raise your hand.

Is that a holdup? Am I getting robbed?

 

[Poodlepal]

You know what I think? I think, for whatever reason, the male skaters we have today (and in 2010) just aren't as good as they were in the early 00's.

Evan is good, but not as good as Alexei Yagudin, who did quads and was usually clean.
Evgeni in 2010 was 8 years older than he was in 2002 (as were we all). He landed a quad, but wasn't as good in other ways as he was in his peak.
I thought Takahashi did try a quad, but fell. Either way, he was no Timmy the quad king (although he's a lot better than Goebel in other areas).

In general, the 2010 medalists were not as good as the 2002 medalists.

One question I would have is that, since the CoP rewards edges, knee bends, transitions and footwork a lot more than the 6.0 system seemed to, would that have an effect on the jumps? Are skaters either practicing these elements more so that they don't focus on their jumps as much? Are they putting so many extra things in their programs to add points that they are too tired to land the jumps consistently? Did the CoP cause the reduction of reliable quad-landers in the late 00's? Or is it just a different population of skaters?

 

[bekalc]

I think you mean 2010 Worlds, which was nearly 2 years ago, so I'm not sure if it's the best example. Kim's PCS advantage vs Asada was much lower than it was compared to what it had been at the Olympics and at TEB, so that goes against your statement that PCS was predetermined. And Asada wasn't perfect; she had a downgraded 3A and her other 3A that wasn't downgraded barely got any +GOE. Are you going to argue that Asada's 3A shouldn't have been downgraded? Because if it hadn't been, Asada would have easily won the FS. But if it deserved to be downgraded (a huge blow to her TES, since she didn't have a lutz or a 3/3), then how could Asada have won the FS? The only other way would have been to say, "Well, she should've gotten a bump here and a bump there in PCS to overcome the deficit of the 3A downgrade to make sure she won the FS." But if she didn't deserve a bump in PCS based on what she put out there on the ice, that's corrupt. Remember that this season, the judges didn't know whether or not a jump had been downgraded by the technical panel. I also think that another factor that influenced the PCS was that Asada and Kim skated in different groups. Had they skated back to back, perhaps the PCS would've been slightly different, maybe enough to make a difference in the rankings of the FS, but not enough to make a difference in overall rankings (as Asada and Kim were rightfully 1st and 2nd overall).

There's nothing to stop a skater, who, in getting low marks for TR, CH, and IN, to add transitions and choreography that better reflects the music to their program. That's what competing early in the season is all about, so you can get feedback from judges and adjust your programs accordingly. And on that note, Asada had both 3As ratified at 2010 4CCs and no major errors, and her PCS was less than 60. That was a huge, huge red light that the judges didn't like her FS. They also didn't like it during the GP. But...she didn't change it. Maybe it was too late, but you can't say that PCS is predetermined because skaters have the option of changing their choreographer and transitions from competition to competition. Michelle Kwan dumped her SP in the 2001-2002 Olympic season. It's been done before.

I like the fact that the IJS punishes cheated jumps and rewards rotation (and I do agree that there needed to be two types of downgrade rules, < and <<). But I've noticed that skaters with errors that aren't obviously visible to the audience or casual fan watching at home (i.e. underrotated jumps, or edge calls) often get more appreciation than their scores reflect; skaters who fall or have obvious step-outs but rotated their jumps are pilloried but may get higher scores than one may expect. While the scoring system is in the favor of those who rotate (and land) their jumps, audience appreciation tends to be in favor of those who land jumps, cheated or not. And that's where most of the debates about placement come from.

It would be helpful if, when the skater is in the kiss & cry and clips of their performance are being played, calls by the technical panel could be displayed over a clip of a jump or spin, showing +3 or -2 (edge call or downgrade).

I'm not arguing that Asada's jumps shouldn't have been downgrading. I'm arguing that she should have had the PCS mark over Kim for that particular competition. Asada's mistakes weren't that visible and she was sparkling. Kim on the other hand was lethargic at one moment after her last fall it looked like she didn't even want to get up. But still while the marks were closer Asada didn't even win the Performance and Execution mark. They tied that. IF that's not a joke what is?

And Asada always had some strengths over Kim artistically even when Kim was clean. I.e better lines being one of them. Once again its not about hating Yu-na she's a great skater and I was rooting for her to win OGM. Its about though making it a competition. The concept that Kim should always get higher PCS than Asada (maybe not as high as she's clean) but always high PCS isn't right. The concept that Asada should always get higher PCS than KIm isn't right either. It should be based on the performance. For one half of the mark to preset is dangerous for the sport. And its hard for me to imagine that the mark isn't preset when you get results like that competition.

Once again I'm not arguing about the TES score in that competition. Asada won that barely and knowing she had those issues I have no problem with Asada barely winning the TES. What Asada should have won was the PCS. She should have been way ahead of Kim on P/E based on what both girls did that night on the ice.

 

[jaylee]

I'm not arguing that Asada's jumps shouldn't have been downgrading. I'm arguing that she should have had the PCS mark over Kim for that particular competition. Asada's mistakes weren't that visible and she was sparkling. Kim on the other hand was lethargic at one moment after her last fall it looked like she didn't even want to get up. But still while the marks were closer Asada didn't even win the Performance and Execution mark. They tied that. IF that's not a joke what is?

Have you ever sat down, watched a live competition, and tried giving out full PCS marks for every skater back to back? I highly recommend that you do so. And you might be surprised that even though you preferred a performance for a skater from Group A over a performance by a skater in Group B, you actually gave a skater in Group B higher or equal PCS marks, even though looking back that would not have been how you ranked them. Kim and Asada did not skate in the same group at 2010 Worlds, and the judges would likely have been surprised looking back that they had given them both the same PE mark. But that is just one category in PCS and it did not impact the overall result.

You confirm my point about how skaters benefit in the eyes of the audience from making invisible mistakes like landing a cheated jump vs falling on a rotated jump. So if the scoring system and PCS followed your line of thinking, the skater who makes barely visible mistakes and being "sparkling" gets higher PCS over the skater who makes visible mistakes (though they avoided invisible mistakes like underrotating). Do you know what that leads back to? Sarah Hughes, 2002 Olympics. She was sparkling and her mistakes weren't visible to most of the audience there.

And Asada always had some strengths over Kim artistically even when Kim was clean. I.e better lines being one of them. Once again its not about hating Yu-na she's a great skater and I was rooting for her to win OGM. Its about though making it a competition. The concept that Kim should always get higher PCS than Asada (maybe not as high as she's clean) but always high PCS isn't right. The concept that Asada should always get higher PCS than KIm isn't right either. It should be based on the performance. For one half of the mark to preset is dangerous for the sport. And its hard for me to imagine that the mark isn't preset when you get results like that competition.

You're a bit contradictory; so because Asada has better lines, she should what...get higher PCS than Kim? But you say you're not saying that either. I'm not arguing that Kim should always get higher PCS than Asada. Kim wasn't at her best at 2010 Worlds and her PCS dropped accordingly. Asada was at her best at 2010 Worlds and her PCS increased accordingly. So it was based on the performance. It just happens that Kim's PCS still came out on top, barely.

As I said in my previous post, why didn't Asada tweak her programs? Actually, if you go back and watch her 2010 Worlds FS vs. 2009 TEB FS, you can see that Mao took out choreography and transitions. This is not a thing that helps your PCS go up, no matter how much effort you put into a performance.

Once again I'm not arguing about the TES score in that competition. Asada won that barely and knowing she had those issues I have no problem with Asada barely winning the TES. What Asada should have won was the PCS. She should have been way ahead of Kim on P/E based on what both girls did that night on the ice.

Actually, what it seems like you arguing is that what Asada should've won was P/E. But in your previous post, you acknowledge that Yu-Na had better transitions, skating skills, and choreography. That's already 3 of the 5 components. But now you're saying Asada should've won PCS. How can she win PCS based on one component, PE?

For all the time that you spend discussing 2010 Worlds ladies LP (where the overall results were correct), you ignore the fact that Asada has been a far bigger beneficiary of judges giving higher PCS to a performance filled with errors that helped save her from a far worse overall result. It goes both ways.

See, 2010 NHK FS: http://www.isuresults.com/results/gpfra2010/SEG004.HTM
or 2010 TEB FS: http://www.isuresults.com/results/gpfra2010/SEG004.HTM
2010 NHK SP: http://www.isuresults.com/results/gpjpn2011/SEG003.HTM
2011 Russia FS: http://www.isuresults.com/results/gprus2011/SEG004.HTM

Look at where her TES ranks and where her PCS ranks relative to other skaters. Mao consistently, regardless of the actual performance, gets higher PCS than skaters who earned higher TES than her, not just by one spot, but a few spots above her. And unlike 2010 Worlds, it actually affects the overall result. If you change the system so that Mao gets higher PCS for the clean performance at 2010 Worlds, thus winning the FS, then it means she should also gets lower PCS for the competitions where she made more errors versus other skaters who made fewer. It's not just Mao, either, but Daisuke has benefited similarly, as well as Kostner, Czisny, or any top skater who has a bad day.

Ideally, I like to see skaters win based on both TES and PCS. I have no problem with a skater who gets 2nd highest TES and the highest PCS or slight differences like that. But it gets to be glaringly obvious that judges are propping up a skater when their TES plummets and yet their PCS remains at the top, and I find it most unfortunate when it affects final results.

 

[mskater93]

IJS

Plushenko :4T+3T (you can't keep an old 6.0-er down )
Lysacek: No quad
Takahashi: No quad
Lambiel: 4T attempt, but flubbed
Chan: No quad
Weir: No quad

I am sure the CoP has many virtues, but encouraging quads by giving partial credit for failed attempts does not seem to be working.

But this example (2010) was with lower BV for the quad, the same carrot for a good one (up to +3) and a much bigger stick for mistakes (negative GOEs were more than just -3 total points off for a mistake, I think on the 4T a -3 was actually -4.2 or -4.5). A fall on an under-rotated 4T (one that was 91 degrees or more under-rotated, not 181 or more) after deductions in 2010 was worth 1 point. Yep, 1. Why take the risk when it's about adding up the sum of parts? This is the biggest problem when you allow a speed skater to devise your scoring system for a performance sport - he creates some esoteric mathematical system that needs to be tweaked constantly to get the desired result - enough risk so that people are working on and trying (with varying success) the hardest elements but success on those hardest elements don't necessarily equal a win if the rest is average...

 

[Mathman]

^ That is very well reasoned, IMHO. I don't really know what to think about the whole thing.

 

[mskater93]

Thanks, Math. We're on the same page more or less, we are just approaching proving the same theorem with different mathematical equations....

Competing under IJS, I can tell you there are things I LIKE about the system (instant technical feedback of what was good and what the judges didn't like) and things I think don't work (risk-reward at lower levels/adult levels, fall penalties and how/where they should be applied - personally I think a fall should be based on a percentage of tech base value and there should be an automatic 0.25 deduction in PE mark that gets applied like the automatic -1 right now for each fall with judges' discretion for additional penalty in PE if the skater takes a long to pick up where they left off - PE seems to be a static score at the highest levels even when the skater is flat/falls multiple times). I think we all agree this still isn't QUITE it from a scoring perspective, but putting together a scoring system that "gets it right" in both the judges' and fans' perspectives is what the ISU should focus on.

 

[Bluebonnet]

I agree with Poodlepal and mskater93 reguarding to quad in men at 2010 Olympics. I think if we wait until we could use 2014 Olympics info for this question, the conclusion on the effect of partial credit might be in its favor.