Can Takahashi Close The Gap On Patrick Chan?

[pangtongfan]

"Can Todd Eldredge close the gap on Alexei Yagudin and Evgeni Plushenko?" That was a question I asked myself when I attended the 2001 Worlds in Vancouver. I was there when Todd skated with a flying speed, faster than his rivals if not everyone in the competition, and with passion and desire throughout his three phases of competition (qualifying, short, free). I was there when the audience gave him a standing ovation and booed at the scores he received for his near perfect free skate wherein the only "error" was his downgrading the opening quad to a clean triple (http://www.youtube.com/watch?v=OEZyj0DM_5o). Alexei Yagudin and Evgeni Plushenko had already skated earlier. "The judges are missing the point about Presentation," said the commentator. I was there. I knew then that a new judging system was needed.

Can Takahashi close the gap on Patrick Chan?
Well, if Takahashi has a near perfect skate as Todd had in 2001 Worlds and is still beaten in PE by someone who has three visible flaws, then I will say another new judging system is needed.

A clean Eldredge couldnt even ever beat Stojko without a quad (if he did all his triples). Of course he was never going to be competitive with Yagudin and Plushenko.

 

[Boeing787]

If Dai skates two perfect programs and Patrick makes multiple mistakes, it is possible for Dai to beat him. The problem with Dai is he hasn't skated two near perfect programs yet. And Patrick tends to peak at worlds. It's going to take a miracle for Dai to beat him.

 

[hurrah]

If Dai skates two perfect programs and Patrick makes multiple mistakes, it is possible for Dai to beat him. The problem with Dai is has hasn't skated two near perfect programs yet. And Patrick tends to peak at worlds. It's going to take a miracle for Dai to beat him.

How many times would Patrick have to fall against a perfect Dai for Dai to win?

 

[Boeing787]

How many times would Patrick have to fall against a perfect Dai for Dai to win?

Seriously, two.

 

[Violet Bliss]

How many times would Patrick have to fall against a perfect Dai for Dai to win?

There is no way to answer this question. Scores are not just determined by the number of falls. E.g. Chan had a random fall with no effect on GOE other than the mandatory deduction, and his falls are usually on fully rotated quads. Daisuke, OTOH, often falls on under rotated and even downgraded quads, and he has popped the all important 3A, even fallen on it thus unable to follow up with the second jump of the combo.

As for a perfect or near perfect Dai, he hasn't shown us yet these last two seasons. He has his perfect SP at NHK and Japan Nationals, but blew it big at other events. He best LP was at GPF with a badly flawed 4T. He received 3 < at his winning NHK LP, including a bad 4T. IOW, he hasn't been able to skate two near perfect programs at one competition. The closest was at NHK with pretty flawed LP.

Chan, OTOH, has shown two near perfect programs at the Nationals, and a darn good, and clean, LP at 4CC. His falls on quads have been due to factors not likely to be repeated, such as hitting the board or that fluke blade click during the entry at SC. He has shown the resolve to get right back to skating and jumping immediately even with mishaps instead of letting the program go. Huge attribute of a champion.

I guess some fans can hope for two perfect programs from Dai while Chan has multiple falls on underrotated quads. Neither of these has happened yet.

 

[Mathman]

Seriously, two.

If we go strictly by the numbers, maybe even three.

Taking Four Continents as a guide, Patrick beat Daisuke by almost 30 points in the LP. Daisuke lost a total of 10 points with his three errors (negative 2 GOE on his quad, popped a 3A. and under-rotated his 3A in combination). Give those 10 points back and Patrick is still ahead by some 19 points.

A fall on a fully rotated jump loses only 4 points (can you believe it!? ), plus whatever positive GOE he might have picked up otherwize. So If Patrick fell four times and was penalized 16 points, he still wins against a perfect Takahashi.

Well, in practice it wouldn't work out this way because the judges would knock off a little on Choreography, etc. Still...

 

[ivy]

Umm, I say that's high risk low return. It still would not have a higher base value than Patrick, and even if he lands both 3As, he will likely get less GOEs and that would cancel the 10% base point increase. If he skates clean, there is no one but Patrick who can beat him, but if he skates badly, he can be overtaken by other skaters.

I think that's a good point. He can't really make a plan based on Patrick making mistakes. It seems a foolish idea not matter who the competition is or when it it. But Patrick has been increasingly consistent as the season goes on, as we have seen in previous seasons.

So then Dai's choice is to play it safe and settle for silver or go all out, threaten for gold, but maybe land off the podium. Personally, if Dai asked me, I'd say concentrate on staying clean and getting silver, accept gold if it happens. Consolidate your place as the strong #2 and look to next season to threaten more.

 

[Boeing787]

If we go strictly by the numbers, maybe even three.

This is how I came up with the number two.

If both Dai and Pat skate a clean SP, it's going to be a tie. The TES will be the same, Patrick doens't have much PCS advantage over Dai in SP or Patrick might lead by 1 point. In GPF, Dai skated a near perfect LP and Patrick's mistakes were equivalent to two falls. Their scores were almost the same. That's why I think Patrick can afford two falls.

 

[demarinis5]

I think that's a good point. He can't really make a plan based on Patrick making mistakes. It seems a foolish idea not matter who the competition is or when it it. But Patrick has been increasingly consistent as the season goes on, as we have seen in previous seasons.

So then Dai's choice is to play it safe and settle for silver or go all out, threaten for gold, but maybe land off the podium. Personally, if Dai asked me, I'd say concentrate on staying clean and getting silver, accept gold if it happens. Consolidate your place as the strong #2 and look to next season to threaten more.

Good advise but Dai won't do that. He has said numerous times that when he skates his mindset is to win and he accepts the risks involved. So imo Dai will not play it safe, which could leave him off the podium.

 

[gkelly]

A fall on a fully rotated jump loses only 4 points (can you believe it!? ),

This is a misleading statement.

A fall on a fully rotated quad or triple axel loses only 4 points, and the "only" applies because the base values for these jumps are so high that even after losing points the skater is still left with more than half of the initial value of the jump.

In other words, the complaint that skaters can get more points by falling on quads and triple axels than by standing up on easier jumps is a valid one, but it is not universally applicable outside of those very difficult jumps.

A fall on a fully rotated triple now only loses 3.1 points including the fall deduction, leaving the skater with less than half the value of the jump -- for a triple toe or salchow, about one quarter.

A fall on a double axel or double lutz loses 1.9 or 2.5 points, respectively, leaving a few tenths of a point net value.

A fall on a single or other double jump, or downgraded triple, loses between 1.3 and 1.9 points including the fall deduction. The points lost are greater than the base value of the jump, so the attempt with fall contributes negative net points to the total score.

I.e., the easier the jump you fall on, the more it costs you.

Even if one of the quad guys falls on an easier jump, he will not get many if any points for that easier jump. Successful harder jumps might compensate, though.

 

[Violet Bliss]

Dai's quads have improved significantly this season. :thumbsup:

Chan's fall on a quad is usually his only flaw on the jump but others, including Takahashi, often incur more than a flaw when they fall. If Takahashi would fall without UR, or UR without fall, or two foot without UR, the penalty would not be so severe.

4T = 10.3 BV
4T< = 7.20
4T with fall = 6.30
4T<< = 4.10
4T< with fall = 3.20
4T<<with fall = 1.00

Negative GOE may reduce the marks of 4T, 4T< and 4T<<. Two footed 4F<< nets Takahashi 3.50.

I may compare Chan's and Takahashi's quads this season later to see how "go big or go home" plays out for each of them.

eta. To illustrate gkelly's point, 4T<< is basically an over-rotated 3T worth 4.10 points. A fall on the 3T (or 4T<<) leaves just 1 point, more than 3/4 of BV lost.

 

[spikydurian]

I guess some fans can hope for two perfect programs from Dai while Chan has multiple falls on underrotated quads. Neither of these has happened yet.

Thanks SF for the further explanation. I am sure many are hoping the above I just hope both skates well as they are both wonderful skaters, and I like them both.

 

[skatinginbc]

Fall on a rotated 4T ==> receiving at least 61.17% of the base mark (10.30 -3 -1)/10.3
Fall on a rotated 3A ==> receiving at least 52.94% of the base mark (8.5- 3 -1)/8.5
Fall on a rotated 3Lz ==> receiving at least 48.33% of the base mark (6.0 -2.1 -1)/6.0
Fall on a rotated 3F ==> receiving at least 41.51% of the base mark (5.3 - 2.1 -1)/5.3
Fall on a rotated 3Lo ==> receiving at least 39.22% of the base mark (5.1- 2.1 -1)/5.1
Fall on a rotated 3S ==> receiving at least 26.19% of the base mark (4.2 - 2.1 - 1)/4.2
Fall on a rotated 3T ==> receiving at least 24.39% of the base mark (4.1 - 2.1 - 1)/4.1
Fall on a rotated 2A ==> receiving at least 24.24% of the base mark (3.3 - 1.5 - 1)/3.3

Fall on a rotated 2Lz ==> receiving at least 0.2 point the base mark (2.1 - 0.9 -1)
Fall on a rotated 2F ==> no worse than 0 (1.8 - 0.8 -1)
Fall on a rotated 2Lo ==> no worse than 0 (1.8 -0.8 -1)
Fall on a rotated 2S ==> no worse than -0.2 deduction (1.4 - 0.6 -1)
Fall on a rotated 2T ==> no worse than -0.2 deduction (1.4 - 0.6 -1)

Conclusion 1: The higher the difficulty level, the less execution or quality weighs. Difficulty outweighs quality in Senior competitions where high-level jumps (2A or higher) are expected.
Conclusion 2: The CoP places quality and difficulty on a more equal footing to differentiate the rankings among those who "may compete but cannot win".

 

[Mathman]

I.e., the easier the jump you fall on, the more it costs you.

I think we could just as well look at it this way:

If you fall on a quad you lose four points.

If you fall on a triple you lose 3.1 points.

Four points is more than 3.1 points.

In the CoP you add up the points. You don't add up the percentage of base value that you won or lost. My statement should have been, "if you fall on a fully rotated jump you lose at most 4 points."

Example: A skater's first two jumps are 4T and 3Lz. He rotates both jumps but falls on one of them.

Scenario A: He falls on the 4T. His score is 6.3 + 6.0 = 12.3

Scenario B: He falls on the 3Lz. His score is 10.3 + 2.9 = 13.2

His fall on the quad was more costly. The business about percent of base value is a red herring. In the CoP, points are points, whatever they are a percentage of.

But any way you slice it, if Chan's actual performances score 20 points above Takahashi's theoretically best program, then Chan can fall a lot of times and still be ahead on points.

 

[gkelly]

Conclusion 1: The higher the difficulty level, the less execution or quality weighs. Difficulty outweighs quality in Senior competitions where high-level jumps (2A or higher) are expected.

Thanks for the breakdown. This does seem to be the case the harder the jumps become. However, there are other gradations of quality besides fall/-3 GOE or base mark, and most jumps don't end in falls, so quality will be a distinguishing factor over the course of each program.

Conclusion 2: The CoP places quality and difficulty on a more equal footing to differentiate the rankings among those who "may compete but cannot win".

Not sure what you mean by "may compete but cannot win." Whether a skater can win or not all depends on the level of competition and the strength of the field. Or by "win" do you mean only "win the world championship"?

E.g., in US juvenile competitions (not sure if the same holds true in Canada or elsewhere), triple jumps are not allowed, so all the skaters will be doing jumps up to but no higher than double axel. In any competition at that level, quality and difficult may be on a more equal footing to differentiate rankings, and also someone is going to win with that content.

 

[gkelly]

In the CoP you add up the points. You don't add up the percentage of base value that you won or lost. My statement should have been, "if you fall on a fully rotated jump you lose at most 4 points."

OK.

Keeping in mind, of course, that the skaters who start with higher base marks may still have more points even if they lose more points.

Example: A skater's first two jumps are 4T and 3Lz. He rotates both jumps but falls on one of them.

Scenario A: He falls on the 4T. His score is 6.3 + 6.0 = 12.3

Scenario B: He falls on the 3Lz. His score is 10.3 + 2.9 = 13.2

His fall on the quad was more costly. The business about percent of base value is a red herring. In the CoP, points are points, whatever they are a percentage of.

OK.
Looking at it this way, this principle would be even more true if falling on a jump meant 0 points toward the total score, which I think is what you would prefer. Falling on a harder jump would lose you more points than falling on an easier jump.

But the planned total that those points are lost from is higher for a skater who attempts harder content to begin with. So if two skaters fall on different jumps and lose different amounts from their planned base mark, they still might come out the same.

With a fall = no cigar (no points) system, then these two skaters would earn the same total for their jumps, assuming they got the same GOEs and the same distribution bonuses:

4T+3T, 4T (fall), 3A+2T, 3A, 3Lz, 3F+2T+2Lo, 3Lo, 3S

4T, 3A+2T, 3LZ+3T, 3A, 3Lz (fall), 3F+2T+2Lo, 3Lo, 3S

The first example attempted a harder repeat jump. They both fell on the repeat jump and got no credit for it. So the jump content will not be the deciding factor here.

The same would be true if skater B popped the lutz before falling:
4T, 3A+2T, 3LZ+3T, 3A, 1Lz (fall), 3F+2T+2Lo, 3Lo, 3S

No credit for jumps with falls means that falling on a 1Lz is worth the same as falling on a 4T.

If skater B popped the lutz so badly he earned -3 GOEs but rotated the single and stood up, he'd get more points for jumps than the guy who rotated the quad but fell down.

And at my level of competition, if it were scored by IJS, if I did a 1Lz that deserved -3 GOE (highly likely if I tried one), I'd darn well want those 0.3 points.

Under 6.0, there was also a rule that jumps landed on two feet should not be scored. I think this referred to jumps landed with weight about equally distributed between both feet, not to an incidental touchdown or step out after landing on one foot. So if we want to change the IJS to give no points for jumps with falls, should there also be no points for jumps landed on two feet? And leave it up to the tech panel to decide which jumps qualify as "landed on two feet" and hence ineligible for any points at all, vs. jumps landed with errors in which the other foot touches the ice and hence eligible for base value according to amount of rotation, minus the applicable negative GOE?

But any way you slice it, if Chan's actual performances score 20 points above Takahashi's theoretically best program, then Chan can fall a lot of times and still be ahead on points.

True. If we were talking about the best skater in the world and the 200th-best skater, there would be no outrage here. The problem is in expecting that the 2nd-best skater in the world should always be within easy reach of the best (assuming we can define Chan and Takahashi as definitively the best and 2nd best -- which is always subject to change as they and other skaters continue improving or start to decline).

 

[skatinginbc]

by "win" do you mean only "win the world championship"?

Of course. I'm talking about major international Senior Competitions, the World Championship in particular. This is a thread discussing whether "Takahashi can close the gap on Patrick Chan". It is inconsequential for us to argue that falling on the lower-level jumps receives zero or a negative point because not many elite male skaters who are eying for the gold would attempt them anyway. Even if Takahashi falls on a downgraded triple flip and receives 0 for that element, it just proves my argument that difficulty (full rotation) matters the most.

there are other gradations of quality besides fall/-3 GOE or base mark, and most jumps don't end in falls, so quality will be a distinguishing factor over the course of each program.

But the worst possible GOE is -3. Even if one has a wrong edge entrance, double footing, hand-down, and then falls forward and backward and upside-down and falls again while crawling up halfway, the worst penalty is -4 (fall penalty included). Though quality is one of the deciding factors of the final scores, it is the difficulty level that matters the most.

 

[gkelly]

But the rules have to work for everyone they apply to, not only the world medal contenders.

There are different rules for different competition levels and different disciplines. E.g., senior ladies aren't allowed to do quads in their short programs. And how quads are scored doesn't have much effect on the ladies' long program/overall results either.

But all senior men compete under the same rules. And at a senior men's competition where no one is successfully landing quads, someone is still going to win.

Mathman and others have used general terms to discuss the principles of the scoring, and that's what I'm addressing. If you object specifically to how quads are scored, then make sure you specify "quads" and not just "jumps." Because the objection might not be generalizable to all jumps, might not apply to the IJS as a whole.

 

[Violet Bliss]

But any way you slice it, if Chan's actual performances score 20 points above Takahashi's theoretically best program, then Chan can fall a lot of times and still be ahead on points.

Your theoretically best program from Takahashi did not include positive GOE he would likely get. From your calculation

Daisuke lost a total of 10 points with his three errors (negative 2 GOE on his quad, popped a 3A. and under-rotated his 3A in combination). Give those 10 points back and Patrick is still ahead by some 19 points.

His 3As from NHK were 10.64, 9.48 points more than the 1.16 for the 1A at 4CC. His 3A+3T from GPF was 15.43, 5.18 more than 10.25 from 4CC, while his quad from Japan Nationals was 11.90, 3.6 in GOE difference. Give him the same mark for the 4T in his SP and it would increase the score by 8.7. Thus you can add 26.96 to Takahashi's 4CC total TES if he had jumped as well as he could. With a little higher PCS from the increased TES, Chan probably can't afford a fall against a near perfect Takahashi, unless maybe he is actually perfect other than a fall.

I see a lot has been written since I started this post, with interruptions, so I need to read them. However, I want to point out that values of GOE go both ways. The flip side of proportionately lower penalties on low quality is lower rewards for higher quality. If GOE value for a quad is increased for the purpose of penalty, it will also increase the difference between a front runner with quads of high quality from those attempting to close the gap. An unintended consequence?

 

[skatinginbc]

But all senior men compete under the same rules.

But the same rules produce different effects on different levels of skaters. Under the same rules, women may hold a different strategy from men, senior different from junior, podium-hopeful different from top-twenty seekers. It lacks a consistent principle or value within the same rules.