2019 Four Continents Pairs Free Skate

[CoyoteChris]

M-T/M. Their twist got scary again, but solid on SBS jumps. I still don't care for this program much, but they did a pretty great job here (minus wobbly lift).

Wow...My wife, dogs and I watched the Pairs free tonight....The Canadian 5RLi3 was scary in slow motion! It was clear many judges didnt see her skate hit his head and one even gave the lift a +3. She started the lift with a very bad look on her face...thankfully, no one was hurt but they were clearly overscored on it....
Reports of Sui grabbing the mic were clearly exagerated, darn it. She raised her hand and requested the mic.....I cant beat Rusty with a wet noodle now....
Worlds will be veeerrrrrryyyyyy interestinggggggggggggggggggggggg.

 

[MFarone]

(break from taxes)
First, I will read the USFS guide to the current IJS in the 2018-19media guide, but I doubt it will be helpful.
Then, I will ask my Go to Guru, M, these questions.
So nine judges are chosen for an event,say Ladies Free, read out loud at the start of the event, along with the tech team.
Lets say Judge Number 9 is Ms. Polaski, for the sake of argument.
So on the "Judges Detail Per Skater, " Ms. Polaski is Number 9 on every skater for ladies free.
There is no dropping out the high and low judges for each skater?
There is no random dropping out of judges by the computer at any time?
Now we are gonna say J9 is honest as the day is long. But say Judge 9 is married to the head of a Federation?
Would it not be possible to enter data into a program and see how J9 (and all the other judges doing ladies) score skaters from all countries over a season compared to other judges?
Not saying anything is going on, like Salt Lake or the Icarus debacle, or not calling pass interference and messing with the Saints....just curious....
"Waiting for a Chinook" with apologies to Fredrick Remington

Here is an article Vox about how judging and scoring work in fs: https://www.vox.com/culture/2018/2/14/17004314/figure-skating-scoring-winter-olympics-2018. The high and low scores on each element are thrown out and then averaged to find the "trimmed mean" (I almost typed trimmed meat, :eeking. There are examples of how reaching the trimmed mean is accomplished which I found very helpful.

I'm thinking about the rest of the questions! And you probably buried the cute dogs toys under that big pile of snow...shame on you!

 

[CoyoteChris]

Here is an article Vox about how judging and scoring work in fs: https://www.vox.com/culture/2018/2/14/17004314/figure-skating-scoring-winter-olympics-2018. The high and low scores on each element are thrown out and then averaged to find the "trimmed mean" (I almost typed trimmed meat, :eeking. There are examples of how reaching the trimmed mean is accomplished which I found very helpful.

I'm thinking about the rest of the questions! And you probably buried the cute dogs toys under that big pile of snow...shame on you!

M, my head is swimming...so the high and low scores are indeed thrown out and there is some sort of trimmed mean????????????????
No dog toys outside...you would not believe the plastic tubs full of dog toys we have that my wife wins....
Thanks for the article! I will read it!

 

[CoyoteChris]

Here is an article Vox about how judging and scoring work in fs: https://www.vox.com/culture/2018/2/14/17004314/figure-skating-scoring-winter-olympics-2018. The high and low scores on each element are thrown out and then averaged to find the "trimmed mean" (I almost typed trimmed meat, :eeking. There are examples of how reaching the trimmed mean is accomplished which I found very helpful.

I'm thinking about the rest of the questions! And you probably buried the cute dogs toys under that big pile of snow...shame on you!

OK, I have read the 2018-9 USFS media guide 5 page explaination of the IJS and the USFS website, which is admittedly by them is out of date.
https://usfigureskating.org/story?id=84064

So, yes, the high and low GOE for each element is thrown out. If the trimmed mean means average, mean, or root mean square, I have no idea. I flunked out of astro physics when I hit the third grade. stats did me in.
I can see no evidence unless you can that any judge AT THIS TIME is thrown out....randomly or otherwise.....9 judges exist for major comps...period. An individual judges's GEO may be thrown out for an element, but if there are two judges who give a skater the exact same low GOE for an element, of course only one of the GOEs is thrown out. Factoring of PCSs and "Factoring" of GOEs is necessary and just under the level of magic to understand. But you can understand why it is needed. Mai's StSq4 is worth 3.9 base, but even though she got 3s,4s, and 5s, she was given 1.45 trimmed mean "factored" GOE of 1.45. It has a lower value, meaning easier, than a triple Axel. 8 points.

 

[MFarone]

OK, I have read the 2018-9 USFS media guide 5 page explaination of the IJS and the USFS website, which is admittedly by them is out of date.
https://usfigureskating.org/story?id=84064

So, yes, the high and low GOE for each element is thrown out. If the trimmed mean means average, mean, or root mean square, I have no idea. I flunked out of astro physics when I hit the third grade. stats did me in.
I can see no evidence unless you can that any judge AT THIS TIME is thrown out....randomly or otherwise.....9 judges exist for major comps...period. An individual judges's GEO may be thrown out for an element, but if there are two judges who give a skater the exact same low GOE for an element, of course only one of the GOEs is thrown out. Factoring of PCSs and "Factoring" of GOEs is necessary and just under the level of magic to understand. But you can understand why it is needed. Mai's StSq4 is worth 3.9 base, but even though she got 3s,4s, and 5s, she was given 1.45 trimmed mean "factored" GOE of 1.45. It has a lower value, meaning easier, than a triple Axel. 8 points.

I think we are in agreement. Looking at the protocol for Sui/Han in the FS the 3T/2T J1 was high with 3/J6 was low with -1 so both probably were thrown out. 3STH J1 was high with 4/J3 was low with 3 so those would be thrown out. It doesn't seem like they just randomly say all of J1 and J4 marks will be thrown out. I'm not real sure what they do when all the judges give the same mark as happened with Sui/Han's 3twist which was given 4's. And as you said, then comes the factoring.

 

[CoyoteChris]

I think we are in agreement. Looking at the protocol for Sui/Han in the FS the 3T/2T J1 was high with 3/J6 was low with -1 so both probably were thrown out. 3STH J1 was high with 4/J3 was low with 3 so those would be thrown out. It doesn't seem like they just randomly say all of J1 and J4 marks will be thrown out. I'm not real sure what they do when all the judges give the same mark as happened with Sui/Han's 3twist which was given 4's. And as you said, then comes the factoring.

Since we need a better def: of trimmed mean, I think in your example the computer mearly throws out two marks and spits out the mean, then factors it....Not a bad system. Unless the computer finds out that Judge X continually gets marks thrown out ...