You Be the Judge

[drivingmissdaisy]

Looking at the World championships and the Olympics, how many of them were decided by a huge lead in the short leading to a conservative skate in the long?

As Mathman's Agnes example shows, it doesn't affect only the winner. This same caution can be used to "hold on" for a medal, to "hold on" to a top 10 finish, etc. I'm not sure that we have had any Olympics (of the TWO under COP) where a skater or pair had an insurmountable lead in the short, so in that case the observation doesn't apply. Even Mao could have caught Yuna had Mao gone clean in the long and Yuna made a couple of errors.

[Buttercup]

b) Again, my question is does the "play-it-safe" happen often enough for it to be a real stain on the scoring system? Is skating tentatively a sign of holding back (for example, can't an athlete just be nervous as opposed to strategic?)? Is not doing the full planned content holding back, or can it just be screwing up a landing or something. Looking at the World championships and the Olympics, how many of them were decided by a huge lead in the short leading to a conservative skate in the long?

Joubert in 2007. He had been injured a few weeks prior to Worlds and with a big lead over most of the field, he and Jean-Cristophe Simond went with safe and clean over risky - he didn't get negative GOEs on anything in the LP (this was before the age of the edge calls, obviously). And it worked, because Buttle skated a poor LP and Lambiel and Dai were too far back after the SP. I think that the success of this strategy is why they tried the same thing a year later - but when you're in sixth place going into the free, that's not nearly as good an idea.

Are you sure Joubert's final combo in 2008 was a planned 3S-2T-2T? That's really not very like him, certainly not circa 2008. My guess is that it was a planned 2A-2T (ETA - protocols geeking: the final jumping pass in that program at Euros and 2007 SC was an axel).

I think Plushenko at the 2006 Olympics didn't skate conservatively, but he didn't take major risks, either.

The thing is, in order to play it safe at a major event you need a big lead out of the SP and to be skating towards the end of the final group and have people skate poorly. That doesn't happen too frequently at that level. So the bottom line is that I agree with your contention that it's not a major issue with the system if you're concerned about the Olympics/ISU championships. It does happen occasionally on the GP circuit - Brezina at SA comes to mind. But it's not that easy to build up a big lead in the SP, so if a skater manages it, I don't mind a cautious skate so long as it's quality skating.

OTOH, having a big lead/seeing your competitors bomb can lead to greater risk-taking. Again using Joubert as an example - 2006 CoR, and the only 3-quad program under the IJS.

[seniorita]

^ That was a great skate of Joubert and mostly because he could have played safe and won anyway.
Nowdays, when you have 10 points lead it is still safe but 5 points lead from sp is not so much of a gap now. It also depends if the skater is usually an sp or lp skater, where he excels and can make the gap bigger. I think Chan belongs in Lp category where he can go far away from the rest in Lp and less in Sp and Plushenko and Joubert could do that more often in Sp. Re Plushenko in Torino, it is known that Mishin told him to skate like a robot until his step sequence, doing nothing showy and nothing bravura as he usual does. I remember especially a phrase from his book, not to make an eye contact with anyone and no flirting with audience, lol. I would think ten points lead from sp would let him skate free and enjoy it but it seems he wanted it very much and was afraid to blow it.

[SkateFiguring]

Holding back is a skater going for clean. The skater may not be so confident of his high risk jumps on that day so why take unnecessary risks? After all, going clean is so prized and called for by many. Competitors under-perform or even bomb for all kinds of reasons. They may abort, downgrade, or pop their jumps involuntarily and spontaneously for self preservation. But when a skater does that after a big lead, it may be perceived as holding back.

I disagree that Takahashi held back at Japan Nationals. He is one skater I will not ever accuse of holding back. At Japan Nationals he went for it at both programs. It worked for the SP but backfired for the LP. He went for it after a huge lead and suffered the risks and consequences. It wasn't his day.

I agree Brian Joubert does hold back as a plan sometimes, especially in recent years when he doe not have the same confidence in his quads as before. Brezina was held back by his coaches at SA 2011. He never experienced a huge lead from the SP at an international event before and it was probably unwise with the world Silver medalist competing though Kozuka underperformed so Brezina's strategy still worked.

One may also say a huge lead affords a competitor the cushion and confidence to take risks with the LP. Different skaters react differently to the same situation. It's hard to conclude if the system encourages holding back or going for it after a huge lead.

What about a real contender falling far behind after the SP? Some of them rally back in the LP to medal or even win the event which would be impossible under 6.0. Should a medal be lost as decided by the SP?

[gkelly]

Under the 6.0 system, the short program and the long program each had its own special role to play. In the short program you had to score well enough to "make the finals" -- that is, to put yourself in a position where you can skate for the championship in the long.

But then, no matter how well positioned you were, you still had to skate for the championship.

In other words, if you wanted to be champion you had to go out and skate a good short and then you had to go out and skate a good long.

It depends which version of 6.0 we're talking about. But even leaving out all of the era with school figures, especially the part without factored placements, from 1999 to 2004 under 6.0 the qualifying rounds at ISU championships (i.e., singles events at Worlds and Junior Worlds, and sometimes Europeans) counted toward the final results, meaning that the long program counted only 1/2 of the total score and not 2/3.

So that meant that sometimes more than 3 skaters controlled their own destiny, and sometimes only 2; in theory sometimes only 1.

If you were 3rd or 4th after the short but not in control of your destiny, yes you could win, and you had to skate for it, but you might blow away the rest of the field and still not win depending what order the others finished in.

If you won both your qual round and the SP and the person who was 2nd in the SP had not won the other qual round, then you were in sole command of destiny and guaranteed the gold medal with a 2nd place in the free skate. Plushenko was in that position at 2001 Worlds. Of course he did go out and win free skate as well. But he did have enough of a factored placements lead that he couldn't be beaten overall if he skated reasonably well for him.

Under CoP scoring, a point is a point, whenever and however you score it. There is, in fact, no rational reason, only tradition, for having a two-part competition at all.

This is largely true, but less because of the add-up-points approach to combining results from the two phases and more because of the introduction of so many required elements and limits repeated elements in the so-called free program.

Under 6.0, it seems like falls were more severely penalized. Those who won never fell in the long program unless everyone else did, too (Kristi Yamaguchi is one example I can think of).

Not strictly true. But most of the counterexamples I can think of are from national or GP events, maybe Europeans, or even less important than those. I can't think of examples at Worlds or Olympics. Mistakes in the LP, yes, but not flat-out falls.

There has always been a problem of fallers being held up in the short program. I remember Ina and Dungjen (IIRC) skating clean but coming in behind a falling Russian team in the Olympics one year (1998, I believe). The Russians were the favorites, and did medal despite their fall. So did Evgeni Plushenko. He fell but was kept in 4th in '02. I believe they skated clean in the long program.

And in those cases almost everything else they did in their short programs, aside from the element they fell on, was superior to the performances that they beat.

Often the commentators tended to give the impression that all the top skaters started from a base mark of 6.0 and got deductions for obvious errors, but that was never true. The judges set base marks based on what the skaters did and how well they did it, deductions for errors applied only to short programs, and not all errors requiring deductions were obvious to most viewers (and commentators rarely pointed out the more subtle ones).

There was never a written rule that performances without falls should automatically score higher than performances with falls. Quite the opposite, in fact -- there was a written rule that a fall was no bar to winning.

As a rule of thumb, it was certainly more common for winning performances to be cleaner than the ones they beat, for skaters to lose points when they fell and often to lose expected placements. So if fans operated on the belief that performances with falls should place behind those without falls, and only watched a few top competitors in each event, usually this belief would be borne out by the official results.

But not always. Sometimes judges believed that the skater with the fall was just better and gave them higher scores. In short programs, that might mean starting with a high enough Required Elements base mark and a high enough Presentation mark that even after the mandatory deduction the total was still higher than the total of a rival who had skated clean but with less content or less quality overall. And if a majority of judges thought so, the skater with the fall would come out ahead.

The "problem" wasn't so much that skaters who fell were held up unfairly by the judges, but that fans often didn't have enough knowledge of what else the judges were looking at to understand why it was equally or more fair for the skater with the fall to earn higher scores.

With COP, someone can fall more than once in the short and still be in the running, and fall again in the long and still get a gold or silver.

That could be true under 6.0 as well in a competition with a weak enough field that two falls in the short might still be good enough for 2nd or 3rd place in the short. In fact, that's a bigger theoretical problem with the flattened out margins of victory in 6.0 scoring.

Under 6.0 if a skater falls twice in the short and is 2nd or 3rd going into the free skate, behind a skater who skated clean with good content and quality, in the free he only needs to be a tiny bit better than the leader to win the whole thing. In IJS, a skater who fell twice in the short and is 2nd or 3rd going into the free behind a skater who skated clean with good content and quality, he will be many points behind the leader and will need to be much better in the free to pass the leader.

Some multiple mistake-filled performances have won top medals. I won't get more specific, but there's one from last year that was infamous. Under 6.0, if someone fell or even made a visible mistake in the long and someone in the top 3 was clean, you knew it was pretty much over. Michelle in 2002, Brian Orser in 1988, etc. That's why everyone thought the Sale and Peltier should've won.

Not everyone. All of those were contested events with mixed ordinals. It so happened that the cleaner performances ended up winning, but there were valid why the long program results weren't unanimous. If you "knew" the event was pretty much over when someone made smaller mistakes as in the Orser and Berezhnaya/Sikhuarulidze examples, then your knowledge was incomplete.

For better or for worse, this sport is complicated enough that there aren't always clearcut winners, and not even a fall much less a smaller mistake necessarily makes it more clearcut.

Here's an easy suggestion: If there is ever again a program where there are no falls, and no hands or feet down, they get a 50 point bonus in the long and 25 in the short.

Ha, that would make it more valuable to do no risky elements, in some cases no elements, or at least no jumps, at all. Most skaters (aside from the top men) do not earn that many points for their jump content in their short or long programs respectively.

The thing is, in order to play it safe at a major event you need a big lead out of the SP and to be skating towards the end of the final group and have people skate poorly. That doesn't happen too frequently at that level. So the bottom line is that I agree with your contention that it's not a major issue with the system if you're concerned about the Olympics/ISU championships. It does happen occasionally on the GP circuit - Brezina at SA comes to mind. But it's not that easy to build up a big lead in the SP, so if a skater manages it, I don't mind a cautious skate so long as it's quality skating.

Agreed.
Skaters will be more strategic at the big championships, and they will face stiffer competition. At GP events, some top skaters are more interested in getting mileage on their programs than in getting another GP gold medal. And there may really only be one "top" (world medal-level) skater in the event, depending how the seeding worked out and who retired and or withdrew. In which case it would not be surprising, with any judging system, for the expected leader to lead and maintain that lead even with mistakes or watered-down content. At Worlds they probably can't afford it.

What about a real contender falling far behind after the SP? Some of them rally back in the LP to medal or even win the event which would be impossible under 6.0. Should a medal be lost as decided by the SP?

It has happened (medaling more often than winning gold) under both systems.

With IJS, you just have to have a larger gap between you and the next-best skater in the LP than that skater had over you in the short. I'd like to see some tweaks to the respective program requirements and/or PCS factors to make this more likely.

With 6.0 and factored placements, you need to win the long program and have all the skaters who were ahead of you in the short finish in just the right order. Which can lead to paradoxes.

[ImaginaryPogue]

Joubert in 2007. He had been injured a few weeks prior to Worlds and with a big lead over most of the field, he and Jean-Cristophe Simond went with safe and clean over risky - he didn't get negative GOEs on anything in the LP (this was before the age of the edge calls, obviously). And it worked, because Buttle skated a poor LP and Lambiel and Dai were too far back after the SP. I think that the success of this strategy is why they tried the same thing a year later - but when you're in sixth place going into the free, that's not nearly as good an idea.

Are you sure Joubert's final combo in 2008 was a planned 3S-2T-2T? That's really not very like him, certainly not circa 2008. My guess is that it was a planned 2A-2T (ETA - protocols geeking: the final jumping pass in that program at Euros and 2007 SC was an axel).

I thought you'd be annoyed if I mentioned 2007 because of the injury.

Source

Lambiel was then followed by Brian Joubert. Joubert correctly realized that if he skated clean the gold could be his. ... So not only did he skate conservatively, he dumbed down his program to a dangerous extent. He had three quads planned in the program but after landing the first one, he left out the next two. On his second element he executed triple Salchow instead of a quad. He replace a second quad toe loop with a triple flip - triple toe loop combination, and his final combination, planned as triple Salchow - double toe loop - double loop, was executed as double Axel - single toe loop. His program was well skated, with the highest PCS of the Free Skate, and component marks in the upper sevens and low eights. Only two elements were scored negative, both due to triple flips which had edge calls.

[Mathman]

As Mathman's Agnes example shows, it doesn't affect only the winner. This same caution can be used to "hold on" for a medal, to "hold on" to a top 10 finish, etc. I'm not sure that we have had any Olympics (of the TWO under COP) where a skater or pair had an insurmountable lead in the short, so in that case the observation doesn't apply. Even Mao could have caught Yuna had Mao gone clean in the long and Yuna made a couple of errors.

To me, the question of playing it safe to protect a lead is not the main issue. In 2006 Arakawa left out both of her planned triple-triples (one on purpose the other by accident) after Sasha fell twice. This happens in every sport.

An example of an Olympic skater having such a big lead that no on can catch him would be Plushenko in 2006. Give Plushenko a ten point lead over Johnny Weir and you can pretty much guess the outcome.

For an Olympic example where someone won the short, then held on to finish second in the long and first overall, that would be Shen and Zhou in 2010.

As for Takahashi at Japanese Nationals, no one (certainly not I) accused him of loafing. But he did have a substandard skate in the free program, yet won overall because of his lead in points carried over from the short program.

I like the model of the semi-finals, then the finals better than carrying over points. In this model you have to deliver the goods in the final no matter what happened before.

I like even better the finals only model. Use the short program as a stand alone mini-competition in its own right, rather than a half long program with accumulated points.

[ImaginaryPogue]

Fair enough.

[Mathman]

It depends which version of 6.0 we're talking about. But even leaving out all of the era with school figures, especially the part without factored placements, from 1999 to 2004 under 6.0 the qualifying rounds at ISU championships (i.e., singles events at Worlds and Junior Worlds, and sometimes Europeans) counted toward the final results, meaning that the long program counted only 1/2 of the total score and not 2/3.

So that meant that sometimes more than 3 skaters controlled their own destiny, and sometimes only 2; in theory sometimes only 1.

If you were 3rd or 4th after the short but not in control of your destiny, yes you could win, and you had to skate for it, but you might blow away the rest of the field and still not win depending what order the others finished in.

If you won both your qual round and the SP and the person who was 2nd in the SP had not won the other qual round, then you were in sole command of destiny and guaranteed the gold medal with a 2nd place in the free skate. Plushenko was in that position at 2001 Worlds. Of course he did go out and win free skate as well. But he did have enough of a factored placements lead that he couldn't be beaten overall if he skated reasonably well for him.

Now that's a format that is both exciting and fair.

It's fair because every skater knows what the deal is, and every skater "controls his own destiny." Win the qualifying round, win the short program, win the final.

It's exciting because often no skater can do all that, so lots of cool edge-of-your-seat scenarios unfold. Like the ladies event in 2001 worlds.

[gkelly]

I like the model of the semi-finals, then the finals better than carrying over points. In this model you have to deliver the goods in the final no matter what happened before.

So how many skaters advance from the semi-finals to finals? 24 singles skaters at the big championships? Or only those with a chance of a medal?

If the points don't carry over, then someone could barely qualify in last place of those who advance, win the final, and win the whole thing. The mediocre skaters who just want to make the final will go all out in the semis. The top skaters who can afford to will take it easy.

I don't hate this idea, but I bet the TV networks would. Especially if the same programs are skated in semifinals and finals.

I like even better the finals only model. Use the short program as a stand alone mini-competition in its own right, rather than a half long program with accumulated points.

So the prestigious championship is the well-balanced freeskating competition, one program only.

At big championships with 50 entries, how would they be whittled down to a manageable number? semifinal/qualifying round skating the same program, perhaps with some prequalifiers?

Then there could also be other competitions with their own medals. And not necessarily the same competitors.
I'm not sure that the short program as it currently stands would be the best format for a separate stand-alone competition, but by inertia that would probably be the ISU's first choice if they decided to have more than one type of event for solo men, solo ladies, and mixed pairs.

[seniorita]

Whats the difference of the semifinal model than the one we have now? I understood that in semi final model, in the semi final skaters will skate the short program? And not two times the same?(semifinal and final). They just wont transfer the marks and will be used as a Q round?

And the short program in final only model will be not a technical one, as a seperate competition and will berenamed as Kwan trophy room - entrance 1 dollar?

[Olympia]

And give Kwan the retroactive gold medal for 1998's short. That seems fair.

[skatinginbc]

As for Takahashi at Japanese Nationals, no one (certainly not I) accused him of loafing. But he did have a substandard skate in the free program, yet won overall because of his lead in points carried over from the short program.

In their conscious minds, some may argue that it is fair while others may think otherwise. But do they actually judge the overall competition based on what they think they believe? Since not many people are willing to play my "Be the Judge" game, I give up. The Japanese Nationals were chosen for a reason. Had gkelly and more experts given the gold to Kozuka or even Hanyu, I would have known that being able to carry a huge lead from the short is a legitimate concern of CoP, not just a chronic complaint from a casual fan.

[Mathman]

So how many skaters advance from the semi-finals to finals? 24 singles skaters at the big championships? Or only those with a chance of a medal?

I was thinking more along the lines of the former system of carrying over placements rather than points. You wouldn't be disqualified from skating the final, it's just that if you were in tenth place after the short it would take a miracle for you to make the podium.

If the points don't carry over, then someone could barely qualify in last place of those who advance, win the final, and win the whole thing. The mediocre skaters who just want to make the final will go all out in the semis. The top skaters who can afford to will take it easy.

Again I was comparing the 6.0 model with the CoP model. All skaters have to go all out in the short program. The top skaters would have to go all out so that they would enter the free skate with a high enough placement to give them a realistic chance to win. The lower-rank skaters would go all out because -- well, what are you there for, if not to go all out?

Passing on to plan B...

So the prestigious championship is the well-balanced freeskating competition, one program only.

At big championships with 50 entries, how would they be whittled down to a manageable number? semifinal/qualifying round skating the same program, perhaps with some prequalifiers?

Then there could also be other competitions with their own medals. And not necessarily the same competitors.
I'm not sure that the short program as it currently stands would be the best format for a separate stand-alone competition, but by inertia that would probably be the ISU's first choice if they decided to have more than one type of event for solo men, solo ladies, and mixed pairs.

Whats the difference of the semifinal model than the one we have now? I understood that in semi final model, in the semi final skaters will skate the short program? And not two times the same? (semifinal and final). They just wont transfer the marks and will be used as a Q round?

And the short program in final only model will be not a technical one, as a seperate competition and will berenamed as Kwan trophy room - entrance 1 dollar?

And two dollars to get your picture taken with Michelle.

I was thinking more of the Grand Prix events than the World Championship.

So let's says at the Eric Bompard Trophy, the first event (replacing the short program) is a...well, a short program. But the judging will emphasize audience appeal, hot dog tricks, the Wow! factor, and the Ain't That Purdy factor. Winner gets a big trophy, called the 2012 Eric Bompard Trophy Trophy (the winner of the long program gets the Eric Bompard Trophy), plus a cash prize of $50,000. Skaters can choose to omit this phase of the competition if they wish. (If that happens a lot, up the prize to $75,000.) That's Thursday.

Then on Saturday the actual competition begins. Skaters skate a long program with CoP scoring. Highest score wins the gold medal and collects points towards the Grand Prix Final, the ISU season best list, etc.

[Buttercup]

I thought you'd be annoyed if I mentioned 2007 because of the injury.

Source

Why would I be? It's factually correct that he skated an easier program than at several past events, and IIRC he said it was in part a strategic decision. I think there was good reason for it, and I'm still impressed that headcasey Brian Joubert managed the last undefeated season in singles skating - but it was a conservative skate for him.

As for 2008: I am not sure how accurate that info is. Joubert and JCS didn't really know what they were doing when it came to CoP, and saving a 3-2-2 for his final jumping pass strikes me as unlikely. Also, Joubert talked a lot about doing three quads, but I don't think it was ever part of a concrete plan after the 2006-7 season. I will be extremely surprised if he ever manages it again. OTOH, it's not like anyone else has managed it, either

Because this is not a Brian Joubert discussion, I would like to add that transitions are over-rewarded and that this is the first thing I'd change if Speedy/David Dore asked for my advice, because as it is the judges have to reward them both through GOEs and PCS and that's too much. Unless it's the SP out-of-steps jump or extreeeeeemly telegraphed, GOEs for jumps should reflect the quality of the jump (height, distance, air position, landing) and not the steps preceding it. I would change this before tinkering with penalties for falls, adding qualifying rounds or even (gasp) ridding the world of catch-foot positions.

OT: I miss our TV discussions.