Indicators and Factors for Predicting Worlds.

[Mathman]

Sorry, Joe. I misunderstood your first post. I thought you were asking which factors are the most reliable in making prodictions about worlds. I didn't realize that you were asking us to apply these factors to particular skaters.

So I said, for instance, that the ability to handle pressure is an important factor, and that last year's results are as good an indicator as any of which skaters to watch.

 

[Hsuhs]

I think a bunch of old scores (received by a given skater in the last 2 or 3 years) could produce more or less accurate predictions. There always will be an error in those calculations though.

 

[Kypma]

I think a bunch of old scores (received by a given skater in the last 2 or 3 years) could produce more or less accurate predictions. There always will be an error in those calculations though.

That usually works to some extent, but various skaters have drastically improved from one season to the next, or some have regressed. Think of Shixuka Arakawa, 2004 World Champion and 9th in 2005 to become Olympic Champion the following year. I don't know of her other scores, though. And Miki Ando finaly broke through this year...

All in all, I think for some skaters it is posibly to predict to some degree what their results will be like, but there are always unpredicted things that come about - injuries, sudden improvements, the skate of their lives... etc. But in general, oine can see a pattern...

Kypma

 

[sorcerer]

Golden Skate Forum Monthly Standings

I think most of the readers took Joe's first post like Mathman did.

So now I got to know and love Joesitz's intentions, but also think that such "best tried probability at the present time" would be a defacto "timely ranking" that takes account of all the substantial skating factors.
A newcomer like me would like to take it this way: why not turn this into "Golden Skate Forum Monthly Standings" !

Because ISU standings, which simply adds up the skaters points that are proportionally distributed only to the relative placings of their past games, doesn't reflect the reality, we know that by seeing Nakano standing above Ando, etc.
......
If this "GS Standings" is considered worthy, then we'll need Mathman to eventually establish the most adequate formula, the rest spending the whole summer trying and assessing.
You'd need quite an amount of discussions going on to select the appropriate variables and adequate coefficients.
Variables would certainly include many of the "indicators" and "factors" suggested here.

Like Mathman I would insist that the last years results does have lots of significance; many variables can be pulled out from this, including the PB, and the high-low difference and diviations can be formulated to show consistency, so on.
And age is a MUST variable.

 

[Joesitz (RIP)]

Good idea, Sorcerer. I could never understand the rankings of the ISU. Irina was still listed in first place even after she took leave of skating for two years.

Hsuhs - That has more possibilities for a Longevity Award based on body-of-work. BTW, would you give Kwan a Longevity Award for Excellence based on her body-of work?

Kypma - Good addition: Improvements as opposed to Declinings. I was thinking of Kwan losing much of her contents as the years went on due to her injuries

MM - Handling pressure is a big one. Problem is we don't see many of the international skaters to see if they do handle pressure, and those that we do see quite often have no pressure in the SP but do have it in the LP (Sasha, for example). In Psychology there is a factor of 'fear of winning'.

Joe

 

[Hsuhs]

I would give Kwan a Kwan award. Readers choice or whatever they call it. They can fax it to her.

I believe in numbers. Predictions based on more than 1 trial, are more accurate. To prove that one or other thing is a valid predictor of smth, is in fact a huge problem by itself.

 

[Mathman]

If this "GS Standings" is considered worthy, then we'll need Mathman to eventually establish the most adequate formula...

I already figured out the absolutely 100% fool-proof figure skating prediction formula. According to my calculations, Irina Slutskaya is the 2006 Olympic champion, Sasha Cohen is the 2006 World champion, and on the men's side, Kevin van der Perren won 2007 worlds in a squeaker over returning Matt Savoie.

I would give Kwan a Kwan award. Readers choice or whatever they call it. They can fax it to her.

Actually, they faxed it to her so many times (7) that they broke their fax machine and had to tell her she couldn't have it any more.

About a rating system, though, there is a difference between a ranking (based on what skaters have actually accomplished) and a prediction (anticipating what you think they might do in the future.) It would be kind of cool to try the latter. :think:

 

[sorcerer]

Well then your formula must be fool-proof-rational, no doubt, but your input values for men must have been too subjective or something.

Anyway, predictions are of course always doomed to fail, needless to refer to Murphy's law.

However "standings of skater's potential" can be a piece of truth, at least come close to it; the difference it will show compared to "standings by skater's accomplishments" should be very interesting. The latter has no variables to discuss on.