Article by Janet Lynn

[Tonichelle]

I think the 6.0 system was a compromise that worked fine. It gave the top finishers in the SP a chance to "control their own destinies," while at the same time it allowed Sarah Hughes, Rudy Galindo and Paul Wylie to rise to rare glory.

what about the Michael Weiss's who skated as well as ever and yet had lost favor with the judges because he hadn't done well at worlds in a while and he skated 2nd in the entire event? I'm sure there were other skaters that dealt with this as well (I think Todd Eldredge also suffered from an early skate in 02)... CoP levels that part of the playing field easier... it's teh disconnect to the competition as a whole that needs fixing. (for my wants)

 

[Violet Bliss]

It seems like a system could be devised that does not set out deliberately to annoy people and ruin their excitement.

If the present scoring system annoys people and ruins their excitement, IMHO the ISU should have a better response than just to shrug and say, tough cookies.

It's the darn brilliant skaters who set out to deliberately annoy people and ruin their excitement, not the system. They should learn to be nice like those who accommodate and comply with "hipcups" and occasional meltdowns.

 

[ImaginaryPogue]

You know, that Europeans event that gkelly mentioned is actually a brilliant example of how confusing 6.0 could be while pretending otherwise. Mathman actually provided a fairly interesting link to an explanation, but I found myself royally befuddled by the whole thing.

 

[Mathman]

^ http://www.frogsonice.com/skateweb/obo/score-obo.shtml

By the way, next semester I will be teaching a course on the mathematics of "Voting and Social Choice," which is all about the impossibility of designing a satisfactory voting system.

Many of these ordinal-type systems were developed by French mathematicians in the 1790s in the wake of the French Revolution, as they were trying to come up with a viable way to organize a democratic system to govern the new republic. There was also a spurt of activity by economists in the 1940s and 50s who were trying to create mathematical models for national economic policy that didn't cause a depression every twenty years.

So the French got Napoleon and the U.S. economy got a system where we borrowed and spent ourselves into the poor house. Wasting our substance in riotous living, as it were. Oh well.

 

[ImaginaryPogue]

Yeah, that's the article.

 

[evangeline]

You know, that Europeans event that gkelly mentioned is actually a brilliant example of how confusing 6.0 could be while pretending otherwise. Mathman actually provided a fairly interesting link to an explanation, but I found myself royally befuddled by the whole thing.

I googled the event and read the same link as Mathman. I also watched some of the performances on Youtube.

My reaction was relatively similar to yours. Man, those ordinals were insane. At least now I can sort of understand why Cinquanta wanted a system of absolute scoring.

 

[ImaginaryPogue]

intriguingly, here's a comment from the link that is inside the article MM posted.

The argument I make in this article is that the problem is not so much with the scoring system itself -- indeed, from a statistical point of view it is technically superior to the alternatives that have been proposed -- but with the failure of the ISU and the media to adequately educate the public.

The more things change.....

 

[skateluvr]

It seems likely Mathman, that the skating schools/industry will eventually return to now dispensed with teaching methods so that we will see more skater's skaters like Janet Lynn, Yuka Sato. 6.0 was too simple and accountability was lacking, too easy for cheating and favoritism. CoP is the extreme answer to try to quantify everything possible, every step, every move. It's way too complicated even for the great skaters of the past to embrace. So your frustration will be mediated by Olympics 2018. A modified, transparent Cop will be in place and everyone will be happier. These are just trends and evolution, one step forward, two steps back.

I say this with hope that there will be an Olympics and figure skating a part of it in 6 years, though my gut says economic collapse will preclude all this.

In the meantime, I guess we enjoy what there is to enjoy.

 

[gkelly]

I just wish there were a way to have it not be such a huge lead that can't seem to be caught up to... not sure if that's possible.

If you're going to have two or more phases of the event and both of them count toward the final result, then the options are pretty much either

1) discard all information about margin of victory in each phase and use placements only from each phase

This can be done with either ordinals or absolute scores to determine those placements -- the actual scores that go into calculating the results of that phase are going to be thrown out anyway.
Then you can define the relationship between the phases so that each phase is worth the same amount or the last phase is worth the most according to some multiplier (factor). And that will mean a limited number of competitors will "control their own destiny" -- in a two-phase event it will always be the same number, depending on what factors you assign each phase. With long program worth twice as much as short program, then each competitor was in reach of the two skaters placed immediately ahead without help from other skaters' placements, and exactly three skaters could win the title just by winning the freeskate without outside help.

In an event with three or more phases, depending how are consistent or inconsistent in the earlier phases you might have one competitor who is completely out of reach going into the final phase, or you might have as many as 5, 6, or 7 competitors in reach of the title without help. And of course, "help" (in the form of other skaters finishing in just the right order) is always a possibility but never within the skater's own control.

Or

2) Maintain data about margin of victory in the earlier phase(s) and also in the final phase.

A: This means that if many skaters perform approximately equally in the first phase, their scores will be close enough going into the final phase that for all practical purposes they all control their own destinies (unless the margin of victory is even smaller in the final phase).

B: But if one skater significantly outskates everyone else in the short program, then the lead going into the long may be so large that it's highly improbable any other skater could catch them.

A is the most exciting scenario. But B is the least exciting. And can't have one without the other if you carry over the scores.

It seems like a system could be devised that does not set out deliberately to annoy people and ruin their excitement.

If the present scoring system annoys people and ruins their excitement, IMHO the ISU should have a better response than just to shrug and say, tough cookies.

Well, I don't think the intention was to deliberately ruin people's excitement (scenario A). I think the intentions were to provide the possibility of maximum excitement (scenario B) and also to prevent flipflops and to provide all the useful information that breaking down scores to individual elements provides. However, B does often occur as an unwelcome byproduct.

OK, so suppose that the ISU decided that excitement and suspense for spectators in the final phase is more important than rewarding deservedly large margins of victory in the first phase. What could they do to avoid frequent occurrences of scenario B?

A few possibilities:

1) Change the number of elements in the short vs. long programs so that there are more than twice as many point-earning opportunities in the LP (this is almost already the case -- the number of elements in the long is slightly less than twice that in the short, but the number of jump elements is more than twice, and at the senior level those tend to be by far the biggest point-getters), and also change the factors for the program components so that they are worth more than twice as much in the long as in the short.

or

2) Round the short program scores to the nearest full point, or the nearest 2- or 5-point mark, so that many skaters will be literally tied going into the LP. But do keep those decimal values on file somewhere in case two skaters tie absolutely in the freeskate and the short program difference is needed as a tiebreaker.

or

3) Use IJS scores to determine placements in each program, but forget about maintaining margin of victory. Use factored placements to combine results from the two phases. Make the factor for the long program large enough compared to the short so that more than 3 skaters -- maybe 5 or more -- will control their own destiny to win without help.

or

4) Get rid of the short program entirely in events with X or fewer skaters. For larger events, use the short program only as a qualifier to determine the top X finishers who will advance to the LP. Throw out the scores and let everyone who advances start with a clean slate for the long.

Those are the options I can think of for increasing the excitement of the long program.
But then someday you'll have a situation where a beloved skater blows away the rest of the field in the SP and just barely loses the LP by the tiniest of margins, and the fans will be outraged again.

The skaters in the lower ranks could still skate the LP for placements, even though their chances of getting first place might be slim (as in the 6.0 system) to none (as in a true playoff model).

wouldn't that make Sarah Hughes Olympic Moments and Rudy Galindo Nationals moments (or Paul Wylie's Olympics for that matter) impossible then? not sure I like that idea. Especially if we're trying to give the Mathman audience element something to feed off of.

I guess it depends how many skaters you allow into the final skateoff -- my X in option 4 above. If only top 2 after the short make it to the gold medal round and third and fourth battle for bronze and everyone else is out of the tournament after the short, then that would be true. But that model really only makes sense for sports where competitors battle each other one on one (or one team against one other) in a series of matches.

Since skaters perform one after another, they can compete against many other skaters at the same time. Or few others, but more than one. So if the "playoff" or "gold medal round" is the final long program with a warmup group's worth of 6 competitors, and short program scores and placements served only to determine who qualifies for the final round, then all 6 competitors who make it would have an equal shot at gold

Or X could be more than 6 -- e.g., 12 or 24. And the reward for placing higher in the short program could be a more advantageous spot in the skating order for the long. But otherwise, if you bomb the freeskate, your placement in the free is your final placement -- a good short would not hold you up in overall standings.

Hughes was 4th after the SP at 2002 Olympics (and thus needed "help" to win gold); Galindo at 1996 Nationals and Wylie at 1992 Olympics were 3rd in the short so they were always within reach of gold on their own merits. In a gold medal round with X of 6, not 2, then they would still have had their moments.

 

[Violet Bliss]

So far this season, of the three event winners with leads from SP, only Takahashi had what could be called an "insurmountable" lead and I didn't read a single complaint about it. On the contrary, there was so much joy, excitement, and gushing over it. Why so much lamenting now? In skating history, I don't think there have been complaints about any dominating competitor. Rather the winningest skaters have been the most celebrated ones, selected into the top 10 lists and called legendary, etc. I always thought sports celebrate the best, the trailbrazers, the greatest, the ones who break one record after another, that leave others in the dust. Here some are complaining and calling for devising the scoring system to prevent someone from excelling and winning too big, at least at the initial stage. I can't think of any sport that has, or would have, competition rules for such purposes.

OK, the only changes mada to prevent big wins that I can think of was in Table Tennis which the Chinese dominated. So about ten years ago, the size of regulation pingpong ball was increased to slow down the speed, and the game scores were reduced to 11 from 21, as the Chinese players were tenacious and often able to come back from early deficit. Before that there was a rule change to make serves completely visible so there wouldn't be sneaky hiding of intention and the kind of spin to be used. Guess what, the Chinese still dominate Table Tennis competitions. Of course arguments can be made that it's not healthy for the sport to have one dominating nation. But preventing individuals from big wins? I don't see any logic about it.

 

[Mathman]

In tennis if you win your quarter-final match, your reward is to get to the semi-finals. Margin of victory is thrown out.

In the worlds series of baseball, if you win the first three games by 20 points and lose the last four by one point, you lose the series.

Not exactly relevant to figure skating, but there are certainly many instances in sports where early scores are not carried into later segments of the competition. Not saying whether this is good or bad, just that figure skating has a choice and should approach it thoughtfully.

Skaters can still win the short program big. Then they can go out and win the long program big.

 

[fscric]

In tennis, the one who lost is eliminated from the tournament, he'll disappear from the next stage of the tournament, so I don't think the comparison is suitable.

 

[Violet Bliss]

In some sports there are qualifying rounds the results of which will be thrown out, such as the round robin or the regular season vs the playoffs. In figure skating, the qualifying results are discarded too. The actual competition involves two programs although at Worlds or the Olympics, the SP also serves as a qualifyer to cut down the LP competitors somewhat, to 25 or 20. In the GP series, the GPF is the Skate Off and previous wins and scores are only qualifyer and not counted. This is not disimilar to other sports when there are eliminations leading to a grand finale.

At least in figure skating, we don't have KO, home runs, and sweeps. Many sports allow early ending of a match, instead of discounting early scores.

 

[Barb]

I think Janet is wrong, the sport wasnt popular before in the US because the system, instead because they had Kristi Yamaguchi, Michelle Kwan, Tara, Sasha and of course the famouse scandal. this sport really always has had few really fans, like us, the others spectators are for some particular skater, like Michelle being super good skater, wining a lot. This sport is compared to ballet and really is similar, is so beautiful but a lot of people just dont like, (I come to think that the most of people are just stupid), the figure skating is so beautiful, is not obvious?. The US is in financial crisis too, then I guess that not so much parents pay for a expensive sport, then there are not very stars, In Japan everybody are skating, obviously they will found someone super good. And I prefer Dai´s sp 2009-10, than any other program in the 80´s or 90´s.

sorry my english.

 

[Mathman]

BBut if one skater significantly outskates everyone else in the short program, then the lead going into the long may be so large that it's highly improbable any other skater could catch them.

A is the most exciting scenario. But B is the least exciting. And can't have one without the other if you carry over the scores.

To me, the problem with B is not that one person runs away from the field. As SkateFiguring said, it is also cool to watch a dominating performance when one contestant clobbers the competition.

The bad scenario is the one where a skater gives a fine short program and ends up with a comfortable lead. Then he flubs the LP, but manages nevertheless to hang on by the skin of his teeth, finishing third or fourth in the LP but first overall. In this situation it's not that the event is boring, but rather that you go away with misgivings that maybe the wrong guy won. The skater, too, has to feel a little sheepish.

It seems like to be the champion you ought to put up a championship performance when the chips are down in the final stretch. It might not always happen that way, but the rules could be adjusted to encourage it.

 

[Violet Bliss]

Some skaters are SP specialists with unreliable LP due to stamina or nerves issues. I imagine they would want to build as big a lead as possible in the SP to hang on to any medal hope. If they do end up on top of the podium, that would be because other contenders falter. It is the disappointing performances of the other contenders that contribute to the bad taste, not the winner with SP lead, who still have to have skated the LP reasonably well even if far from the SP excellence. In any case, such scenario is highly unlikely in a major event with most top competitors participating, not all of whom will underperform badly even if some do.

Do you want to take away the possible excitement of come from behind wins for the prevention of win by SP? Since COP is not an ordinal system, the finale placements have been strange and unpreditable combinations of SP and LP placements. A second place in both can end up in first or third, or a combination of two fifth places may win a Silver - all real and recent examples. Is such unpredictability not exciting, if a predictable win is considered boring and disconcerting?

 

[Spun Silver]

This sport is compared to ballet and really is similar, is so beautiful but a lot of people just dont like, (I come to think that the most of people are just stupid), the figure skating is so beautiful, is not obvious?.

I like the way you put that.
About Daisuke... I'd be curious to know how Janet Lynn feels about him and his programs. Surely she wouldn't describe his skating as a "measured fall from grace." Her critique may be of the worst fruits of the IJS without taking account of the best ones.

 

[Olympia]

I think Janet is wrong, the sport wasnt popular before in the US because the system, instead because they had Kristi Yamaguchi, Michelle Kwan, Tara, Sasha and of course the famouse scandal. this sport really always has had few really fans, like us, the others spectators are for some particular skater, like Michelle being super good skater, wining a lot. This sport is compared to ballet and really is similar, is so beautiful but a lot of people just dont like, (I come to think that the most of people are just stupid), the figure skating is so beautiful, is not obvious?. The US is in financial crisis too, then I guess that not so much parents pay for a expensive sport, then there are not very stars, In Japan everybody are skating, obviously they will found someone super good. And I prefer Dai´s sp 2009-10, than any other program in the 80´s or 90´s.

sorry my english.

I agree with you, Chapis: the sport is like ballet, which is why I love it so much. I am also tempted to think that people who don't like it are just stupid...but we won't tell them! You're probably right that what makes skating less popular in the U.S. today is that we just don't have any star ladies.

I understand what you mean about Daisuke, though there are some programs in the 80s and 90s that I also love--for example, almost anything by Kurt Browning.

I suppose we can't go back to the 6.0 system, but it is hard to enjoy the CoP system in its current form. Maybe the system will be modified in some way that makes it more understandable without losing its ability to quantify a performance. The fact is, I'll continue to love skating in whatever form I can get it, and I suspect most of the rest of GS will also, even as we complain about the scoring system.

 

[seniorita]

Not so. Looking at last season, high PCS did not win competitions for Verner, Contesti, Takahashi, etc. when they faltered in TES. They have however saved Joubert and Takahashi from embarassing placements on occasions when they bombed their programs.



Choreography is closely tied to a skater's skating ability. Most skaters are not able to perform the top skaters' programs even without the difficult elements. A good choreography showcases and highlights the skater's strength and an over- or under-challenging one will not put the skater in good lights. It is really a colaboration between a skater and his/her choreographer. It takes two to work out the best design and bring out the best of the skater and then the skater has to perform it well to optimize the marks.

So a skater should be judged
I) on the quality of his choreography,
II) or on the quality he has been performing and how he is committed to the choreography he has been given- even if the choreo itself is not good?

Examples,. Song who obviously didnt afford a Lori Nichol kind of choreographer so far but looks really engaged in his sp or Leonova programs who have a bad choreography but she is really commited to it.

 

[gkelly]

To me, the problem with B is not that one person runs away from the field. As SkateFiguring said, it is also cool to watch a dominating performance when one contestant clobbers the competition.

The bad scenario is the one where a skater gives a fine short program and ends up with a comfortable lead. Then he flubs the LP, but manages nevertheless to hang on by the skin of his teeth, finishing third or fourth in the LP but first overall. In this situation it's not that the event is boring, but rather that you go away with misgivings that maybe the wrong guy won. The skater, too, has to feel a little sheepish.

It seems like to be the champion you ought to put up a championship performance when the chips are down in the final stretch. It might not always happen that way, but the rules could be adjusted to encourage it.

You can try to make the final phase worth more, and I suggested several ways in which that can be done. But you can never get rid of the possibility that a lackluster performance will win the title.

Whether you carry over factored placements or total scores, sometimes the only people who have a good day were too far behind in the short program to catch up with a good long. The only way to guarantee that will never be the case would be to throw out the short program results entirely. Which would throw out almost all incentive for good skaters to push themselves in the SP -- all they need to do is enough to qualify. And maybe earn a good skate order for the LP if that's the only reward.

Sometimes everyone has a bad day in the long program. So even if only the long program counts, you still might end up with the least-bad skater winning and feeling sheepish.