PDA

View Full Version : Gifts to Chan keep American men out of GPF



Pages : 1 2 3 4 5 6 7 8 9 [10] 11 12 13 14

AliasJohnDoe
12-10-2010, 12:57 AM
I'd just go back to -1, -2 and -3 points for -1, -2 and -3 GOE. And reduce the quads value and GOE very slightly.

As in, redo most of the changes that were done because of people whining about lack of quads.

Now those changes are in effect, they resulted in fully rotated jumps (especially quads) being worth more, even with a fall. And hey, now people have something new to whine about. OH JOY!

My problem with -2 for falls is that it penalises somebody who falls on a quad the same as somebody who falls on a single jump or somebody who just trips on their crossovers.

But the penalty for a fall deductions aren't just for jumps. The fall penalty in the deduction column are for falls whether they are on jumps, footwork, spins, a caught toepick or one of those crazy out of the blue slips on a crossover. A fall of any kind is a penalty...always has been. And it doesn't just come off the TES. It comes off the TSS score which includes the TES and PCS combined score.

The penalty for a fall in the deduction column has always been there, and noone ever complained about the deduction being there before. I'm just trying to come up with a solution to fall penalties based on the rules already put in place. And falls have "always" been penalized in the deduction column. I'm just saying raising that penalty may be an alternative solution. Everything else has been tweaked...why not the deduction column?

The deduction column is for penalties on the program as a whole, whether the penalty is a fall, extended lift, costume deduction and over the time deduction.

And Thank You for responding Ziggy and Rock2. I was beginning to think I wasn't explaining myself very good. I only want to improve the system with the rules that are already in place. And the penalty/deduction column has always been there for a reason, but never tweaked.

bek
12-10-2010, 03:04 AM
As in, redo most of the changes that were done because of people whining about lack of quads.

Now those changes are in effect, they resulted in fully rotated jumps (especially quads) being worth more, even with a fall. And hey, now people have something new to whine about. OH JOY!

How much does the idea that just because I wanted to see quads worth more, it doesn't mean I wanted to see people rewarded for falling. I wanted them to reward the quys who landed the jump cleanly. Not hand someone MORE points than they'd get for a triple lutz, falling on their arse. I'm not sure why reward quads more = reward falls more? Its hardly truly rewarding jumps if you give people tons of points for falling on their butt.

I think that they should move away from the -1, -2, -3 Goe anyways, and move to percentage of base value for plus and minus GOE. But have greater penalties for mistakes (Since those with plus goe are already getting full base value).

I.e how about -1= 1/3 of the jump value taken off, -2 equals 60% of the base value taken off, -3 equals-90% of the jump value taken off. Than it could be also +1 GOE= 10% of the base value added on, +2=20% of the base value, +3 30% of the base value.

There could be a rule that if someone falls on the second half of a jump contention that the fall can get -2. (so said person gets at least 60% cred for landing the other jump

professordeb
12-10-2010, 03:36 AM
AMEN. Well, said. I wonder if Patrick where a bit more humble, if some of the ire wouldn't be so great. Gotta love ubers and anti-ubers.

And perhaps some people should put what he says into the context of what was asked and answered instead of picking out parts that make him look more young and foolish than he already is. Sheesh, I pretty sure there have been a number of skaters who have said something and it sounds one way when you hear only one part of the answer. Then later, someone comes along and explains what was asked and what was actually and completely said (Joubert regarding Buttle, anyone remember?) and most of us accept it as OK.

I don't think any of us who just follow skating (as opposed to knowing skaters quite well and personally) really know if Patrick is humble or not. Yeah, we've seen what he has said in print - the right or wrong of it notwithstanding. Is he PC enough for some - apparently not and most times, I care not. But then, I'm not the most PC person either. :D

gkelly
12-10-2010, 04:42 AM
But the penalty for a fall deductions aren't just for jumps. The fall penalty in the deduction column are for falls whether they are on jumps, footwork, spins, a caught toepick or one of those crazy out of the blue slips on a crossover. A fall of any kind is a penalty...always has been. And it doesn't just come off the TES. It comes off the TSS score which includes the TES and PCS combined score.

The penalty for a fall in the deduction column has always been there,

Actually it was around 2005 that the fall deduction was introduced. At least the very first test run of IJS in the 2003 Grand Prix did not include fall deductions. As I recall, judges were encouraged to reduce the P/E component, but it wasn't obvious that they were doing so, if indeed any of them were. So the automatic fall deduction off the TSS was introduced.


I'm just saying raising that penalty may be an alternative solution. Everything else has been tweaked...why not the deduction column?

The deduction column is for penalties on the program as a whole, whether the penalty is a fall, extended lift, costume deduction and over the time deduction.

I think the problem that people are complaining about now is that triple axels and quads still retain a large percentage of their value even after the deduction and -GOE penalties for a fall. No one's complaining that skaters who fall on double jumps are getting away with too much credit for their failed jumps. In fact, between the -GOE and the fall deduction, double jumps with falls end up with negative net value.

So if the deduction for falling were increased to -2 instead of -1, or to -2 for the second fall, -3 for the third, etc., it would have a disproportionately severe effect on skaters at lower skill levels who are doing mostly double jumps and a comparatively negligible effect on skaters who are doing triple axels and quads.

And that problem can't be easily solved just by saying, OK, the fall deduction will be -3 for senior men, -2.5 for junior men and senior ladies, -2 for junior ladies, etc. Look at the junior men's field. The best jumpers are doing multiple triple axels, sometimes two in the long program, and a tiny handful may be trying quads. The weaker jumpers may be doing more doubles than triples. The same rules have to apply to the guys who are competing for medals and the guys who are shooting for top 15 in the same JGP event.

If the problem that needs solving is that falls on triple axels and quads are worth too much, then I think the solution needs to be in increasing the -GOE value for those jumps (as was done a couple years ago and then undone this year), not in increasing the fall deduction across the board.

I think of it this way. Most falls have a negative effect effect on the impression of performance/execution, and some falls that are due to momentary or pervasive lack of edge control also have a negative effect on the impression of skating skills. But not all falls are equal. Some are very disruptive, some barely a blip. Out of thousands of skating falls I've witnessed over the years, I can think of about three that actually had a positive effect on the overall impression of the program.

So I think there needs to be some leeway in how judges reflect falls in the PCS. Let them give PCS harsher penalties for more disruptive falls and none for falls that they don't perceive as disruptive at all, even if some fans do.

Keeping in mind that long programs contain more risk elements (primarily jumps) and more risk elements after the skater is fatigued, it's less realistic to expect completely clean long programs and each individual fall represents a smaller proportion of the total program length. Also, PCS values for long programs are double what they are in the short, so that 0.25 difference in a men's long program translates into 0.5 in the final score. Therefore, as a rule of thumb, I'd expect judges to give no PCS penalty to one minimally disruptive fall in a long program, and maybe 0.25 off Performance/Execution for each moderate disruption or each two minor disruptions.

In short programs, values of the PCS are smaller compared to the value of each element score or each fall deduction, and also a fall in a short program represents a larger percentage of the program. So maybe judges should be harsher on falls in short programs in their P/E scores.

Judges could also be encouraged to consider the overall impression of the element execution as reflected in GOEs, so that programs with minuses on most elements would systematically tend to score lower on P/E than programs with mostly pluses. That wouldn't apply only to falls that are subject to the TSS fall deduction, but also sloppy execution of any kind.

But I still think those should only be guidelines, not something to be required or enforced.

If judges follow those guidelines, then the PCS would tend to be lower than what the skater would have gotten for a clean performance of the same program. But observers who don't appreciate the good qualities of their non-favorites as much as the judges do, or vice versa, might still think the nonfavorites are overscored and favorites underscored.

I think this is already the case. It's just that individual judges make up their own rules of thumb as to how to penalize disruptions or sloppiness in the P/E mark -- there aren't official guidelines.

And meanwhile, when skaters whose PCS are deservedly still high even when reduced by 0.5 here and there from what they'd get for a clean program also get significantly positive scores for rotated triple axels and quads with falls, it looks like they're being overscored.

Do we want judges to be thinking "I don't like the way falls on superhard jumps are scored this year. I give -3 as required, and the tech panel gives the fall deduction, and the skater still gets too many points for the failed attempt. Falls on medium-hard and easier jumps are already adequately punished. So I'm going to counteract the discrepancy by dinging the PCS more harshly for skaters who fall on quads or triple axels than I will for skaters who fall on easier jumps."?




There could be a rule that if someone falls on the second half of a jump contention that the fall can get -2. (so said person gets at least 60% cred for landing the other jump

You mean -2 GOE for falling on the second jump in a combo? That is already true. Of course if there's anything else wrong with the element the GOE may end up at -3 anyway.

sk8er1964
12-10-2010, 04:55 AM
It seem to me that according to some of our posters everything wrong with the scoring and the standings is ALL CHAN'S FAULT. Yep, I got the message loud and clear. :lol:

I disagree. I see Chan haters interpreting the discussion in that way, but many, many people are arguing that the system that rewards programs like Chan's SC SP is flawed and would be flawed no matter who was rewarded like that.

Its pretty clear in the various threads who the ubers are, and who the haters are. The rest of us are looking for an explaination/discussion/solution to improve the scoring so there is balance.

bek
12-10-2010, 05:07 AM
You mean -2 GOE for falling on the second jump in a combo? That is already true. Of course if there's anything else wrong with the element the GOE may end up at -3 anyway.

Oh I don't think that falling on your combo back end is punished enough in this system. I was talking about if they moved to GOE being rewarded by percentage.

For example if the system was

-1=25% off base value
-2=50% of base value
-3=75% of base value.

I think percentage for positive GOE should be only 10% incriments seeing as they are already getting rewarded by the base value something like
+1 =10% added
+2=20% added
+3=30% added

I was afraid what I was suggesting would be very harsh for the person who did a 3lutz/3toe and actually landed the 3lutz, so thats why I said that maybe given only 50% in that case.

The benefit to what I'm suggesting is that harder combos would actually be rewarded more under a base value system.

I can't remember what a quad is worth now, but going by last year's base value for a quad, a fall on a quad would only net someone 2.45 points, thats a pretty huge difference in comparision to our current rules.

If they want quads to be worth more too, they could always introduce bonus points for CLEAN quads. Key word is clean.

I am cool with the new rotation rules though.

judgejudy27
12-10-2010, 05:22 AM
I love how people use the word 'clean' as if they should have won just because there were no falls. There is plenty of other things skaters a judged on. The base value between one or another could be the differnece between one skater having enough points if they fall.

I never said Sasha deserved to be placed higher at any of the events. I was simply pointing out the myth of her so called 2 fall margin on other top skaters, other than maybe on say Fumie Suguri (and even her not always). She is not a valid example.

judgejudy27
12-10-2010, 05:25 AM
The best way to ensure this happens is to score nothing but jumps and falls. Forget PCS, forget spins and footwork, forget the rest. Just jumps and falls. Then you will always be able assert the completely irrational argument that winning and medals should ONLY be determined by falls, nothing else. This is what you're trying to do.


What BS. :rolleyes: If people thought medals should just be awarded on who falls less than we would be saying Chan deserved about 9th place at his 2 Grand Prix events. Yet most of us are saying he should have been 3rd at the first, and 2nd at the second (just much further behind winning). This may come as a shock to you but Chan is not the all time greatest spinner, footwork, artist, performer, and everything else in the history of figure skating. And considering his winning with so many falls is unprecedented he would pretty much have to be the best ever at everything else to justify the 6 fall margin he is given by judges so far over the rest.

Rock2
12-10-2010, 06:19 AM
He is overscored in jump GOE (a huge amount even on days his jumps are shaky), spin GOE, and overscored alot for interpretation and performance. One would think he had the jumps of Oda, the spins of Lambiel, and the performing ability and musicality of Takahashi and Lambiel, neither which is true even on a good day.


Well at least you're someone who tries to sort this out using the rules.
Hmmm. In the case of skate Canada Patrick got little GOE on jumps, mid-range GOE on spins (which are fast and centered with accurate positions, while he has straightened his free leg which is a bonus for him) and strong GOE on footwork where I don't think anyone can touch him as it is, so 2s and 3s it is. No overscoring on TES as I see it. All deductions in GOE and for falls taken per the rule book.

PCS is likely where the concern is. Can debate that one as much as we want as he did get 10 pts more than Oda. Can go through that one in as much depth as you like, but I have to caution you that neither performing ability nor charisma impacts PCS. Again that's show skating. There is some relevance for performing ability in the IN mark, but it all depends on the music you choose in terms of its appropriateness.

I do think that in terms of who's skating now, no one can touch Chan other than Dai in SS, TR and IN, but I"m sure that statement will start a whole poopstorm so I'll wait for that to happen before I outline why...again in the context of the rule book. Once more, pls don't talk to me about musicality and performing ability, the latter especially. Little relevance I'm afraid.

Fortunately, Chan wasn't up against Dai or Lambiel or Scott Hamilton or whomever at SC, so I guess he got away with one, being up against a more favorable field.



On top of that the whole system is wrong in many ways and there should be bigger deductions for multiple falls in the first place, which would hurt "skate on my ass" Chan more than anyone else.

Yup the system needs tweaking still. But you can give up on the multiple fall deduction pipedream. Will never happen. Maybe Speedy can explain why better than I can. I'm running out of angles to come at this.

ks1227
12-10-2010, 06:41 AM
I do think that in terms of who's skating now, no one can touch Chan other than Dai in SS, TR and IN, but I"m sure that statement will start a whole poopstorm so I'll wait for that to happen before I outline why...again in the context of the rule book.
Well, I think Kozuka is in the same class. But I agree that it's a small group.

theshrew
12-10-2010, 08:36 AM
Sometimes Patrick is overscored because he gets credit for things he doesn't do, according to the rules. Sometimes he is overscored because he gets the credit he deserves, but other skaters don't get the credit they deserve.

And he is overscored because, according to the rules, he gets rewarded thrice for the same thing he does very well: high transition mark because he has lots, very well executed transitions, high mark for choreography because having lots of transitions gives the impression of a very elaborate program and high GOE for jumps because of the transitions /difficult entries.

Artifice
12-10-2010, 09:53 AM
Yes, the PCS system tends to allow multiple rewards for the same thing, which is not supposed to be expected.
One thing IMO should be done to reduce the number of components, there are really too many (5) criterias to judge in a too short time, also given that the definition and undestanding of components is not perfectly clear for judges. Not that they are not competent (eventhough sometimes they are bad, but it's how it is everywhere), but the system requires too much capacities for us human beings.

Once the components are limited to 2 or 3 scores, skaters like Chan won't be rewarded too many times for one quality. He will be rewarded for one thing and it will be clearer to see that he doesn't necessarely have to be rewarded for the rest.

That being said there is still a problem with programs with multiple falls that can win over possible less risky or quality performances. It's ok to encourage risky elements like quads, it's therefore ok to accept the consequence of the risk that is falls, but it's less ok to allow a 4 falls program to win. Extreme performances shouldn't be allowed to score high, wether it is a risky performance ala Chan or a poor athletic skate ala Lepisto.

I don't know if one should set a calculation rule that would prevent such thing to happen because whatever the accounting is done there will always be some cases when it doesn't work and when it won't prevent of happening what we actually tried to avoid.
Maybe something like a clear rule that would say something like a program with more than 3 falls won't be allowed to win the section, and a program with less than 3 validated triple jumps (for exemple, it can be more or less) won't be allowed to win the section.
It's just an exemple of a rule that can be set independantly of adding calculation rules. It's just based on an overall thought that is we don't like to see a splatfest win, nor do we want to see a technically empty program win. Therefore the rule can be simplified in a sense to directly address to skaters that if they falls too much whatever the risk they take, whatever skating skills they have, they can't expect to win. Same things for skaters who don't attempt anything in order to gather points that would eventually lead them to win by defaut.
In other words it is a message that says skaters shouldn't play too much with the limits of the system : either by attempting too much things or by limiting the risk factor and gathering points almost only with components.

Dragonlady
12-10-2010, 02:48 PM
In other words it is a message that says skaters shouldn't play too much with the limits of the system : either by attempting too much things or by limiting the risk factor and gathering points almost only with components.

So skaters should just try to stick to mediocrity and not go for too much? That's saying, don't show up the less talented, the less ambitious - don't go for the win! How exactly can people set out to win the event if they're not to do "too much" and what exactly constitutes "too much".

Rock2
12-10-2010, 04:08 PM
That being said there is still a problem with programs with multiple falls that can win over possible less risky or quality performances. It's ok to encourage risky elements like quads, it's therefore ok to accept the consequence of the risk that is falls, but it's less ok to allow a 4 falls program to win.

I do get the theory and why this causes people angst, believe it or not. But, you can never design an objective/transparent scoring system that can ensure this happens. If you do, by hyper-penalizing falls, you will have show skating on your hands. Might as well bring back Peggy Fleming; she could win!

So...sorry folks, really. It will never happen. As I said, though, I'm ok with -2 for falls.

professordeb
12-10-2010, 05:09 PM
I disagree. I see Chan haters interpreting the discussion in that way, but many, many people are arguing that the system that rewards programs like Chan's SC SP is flawed and would be flawed no matter who was rewarded like that.

Its pretty clear in the various threads who the ubers are, and who the haters are. The rest of us are looking for an explaination/discussion/solution to improve the scoring so there is balance.

Sorry, I was attempting to by pithy and funny amongst the angst. Sorry it didn't come across that way.

I would agree, that the scoring system we have still has areas that could use work on them and I appreciate when actual discussion is taking place. It makes it easier for me to understand why one over the other.