The Judging Controversy Thread

[Ven]

The most likely scenario is the conspiracy starts at the very top of the ISU. Probably Cinquanta and his minions received large sums of money from Putin, and made sure the tech panel and certain judges would fix the results.

Not just the ladies scoring was fraudulent, but even things like V/T's scores when they made clear mistakes, Plushenko, all the Russian competitors really being overscored. No doubt it started at the very top.

 

[Ven]

Although I think the popular scapegoat tactic has been to focus on one or two judges, the scoring discrepancies are pretty systematic throughout the competition and through multiple judges -- even extending to the tech panel. This is really an indictment of a conspiracy of the entire ISU conducting a fraudulent competition, not just one or two rogue judges.

 

[markmchen]

!! wow.

btw, Caro averaged 1.3 per element in Sochi. Mao averaged 0.9 per element

 

[Meoima]

The most likely scenario is the conspiracy starts at the very top of the ISU. Probably Cinquanta and his minions received large sums of money from Putin, and made sure the tech panel and certain judges would fix the results.

Not just the ladies scoring was fraudulent, but even things like V/T's scores when they made clear mistakes, Plushenko, all the Russian competitors really being overscored. No doubt it started at the very top.

More like they were scared of their lives. Knowing Putin, I am not surprised. Just kidding, lol. As for Plushenko, had he not withdrawn, he would definitely medal. If he had not fell, then it might have be gold.

 

[markmchen]

Regarding the point about the changed GOE weighting post-Vancouver made earlier...

Adelina scored 61.43 base value, 22.78 average GOE, and PCS of 74.40 in Sochi
Yuna scored 60.90 base value, 20.78 average GOE, and PCS of 71.76 in Vancouver

If it wasn't for the scoring change regarding GOE weightings, Adelina would easily have had the new world record

 

[skatedreamer]

I checked Adelina's Sochi performance against Yuna's Vancouver performance. (Too lazy to check against all history, but that is the ladies world record performance) In terms of raw GOEs (before weighting etc.), Adelina averaged 22.78 across 12 elements in her FS. Yuna averaged (in 2010) 20.78 across 12 elements. Adelina's average is even more impressive in that she had one element with negative GOEs. Excluding that negative element, Adelina average a GOE score of 2.2 for the rest of her elements. Yuna's Vancouver performance averaged 1.7 per element.

Holy...um...fertilizer.

Are we saying here that according to the numbers, Adelina's Sochi performance was somehow better than Yuna's in Vancouver? Better in Sochi I could almost (almost) believe but better than Vancouver?

!!!!!!!

 

[Ven]

Holy...um...fertilizer.

Are we saying here that according to the numbers, Adelina's Sochi performance was somehow better than Yuna's in Vancouver? Better in Sochi I could almost (almost) believe but better than Vancouver?

!!!!!!!

The numbers are just placeholders to pseudo-verify made up, pre-determined results. The judging cannot completely save skaters who fall (like Julia) but more or less even with some room for error the sport is fraudulent in its current state and the results are predetermined behind closed doors.

All of this arguing and analyzing about numbers and levels is bunk. None of it matters because the scores are not properly administered anyway. What is 6.25 or 9.5 for SS or CH? What is an edge call or not? Downgrade or not? Level 4 or Level 2? None of it is judged according to the rules, just random numbers are put on the scoresheet to verify pre-determined results.

 

[usethis2]

You are correct that there were fewer divergent judges for Caro and Mao then Adelina. On GOEs, Mao had 1 favorable judge and 1 negative judge. Otherwise pretty tight judging around mean. On GOEs, Caro had 1 favorable judge and 2 negative judges. Judging blocks were not as strong or as clear.

Mao having smaller deviation is understandable because the judges probably assumed (correctly) she was out of the contention anyway due to her disastrous SP. In a sense, Mao received the most honest and correct mark for her FS. An astounding performance which received 142.xx. Had she performed the SP slightly better, she would have been OGM. Then again, in that scenario we might have seen a much bigger standard deviation for her FS marks among the judges.. for reasons other than her skating.

 

[markmchen]

probably right about mao.

in contrast, Julia's GOE scores have similar pattern as Adelina. 4 judges well above the average. For example, one judge scored Julia 3, 3, 3, 3, 2, 2, (failed jumps), 3, 2, 3. Excluding her failed jumps, Julia's average GOE was 2.1 per element. So well above Yuna in Sochi/Vancouver and just below Adelina in Sochi.

 

[usethis2]

Tell me about it. The similarity between Yulia and Adelina's GOE charts are striking! (minus falls)

 

[Alba]

2. judges seem to relatively agree on Adelina's PCS scores (surprise!)

I saw the protocols and that was interesting and strange, because I'm still convinced that she got more, much more, than she deserved on PCS.

 

[Alba]

Essentially this means you can have a panel where ALL OF THE JUDGES are cheating, and NONE of them get caught because none of them tripped the corridor that THEY THEMSELVES CREATED THROUGH THEIR OWN AVERAGE CHEATING SCORES?!

Please tell me that's not right

I wouldn't be surprise, and the reason is not necessary because all the judges are cheating.
If the ISU procedure is to punish those judges who are "out of line" in scoring - and many fans seems to agree with that, forgeting that maybe these judges are right in giving those scores - than I certainly would be afraid to find myself in a "outline" position, If I were a judge.

Have you notice that even when an element it's perfectly done you rarely see +3 from more than 1 or 2 judges, at max?
Lakernik in an interview in 2012 (very interesting one) was criticising the judges for this. I mean for not giving the right +3 GoE's when deserved.

 

[gotoschool]

I am with Ven on this, but I have already thrown in my two cents. Especially PCS scores were likely set ahead of time because if have to be as readily justified. Basically, if the scoring is rotten for several different skaters in comparison to each other- what's the use in trying to find out the part that's justified.

 

[Mathman]

[SUP][/SUP]

If I may ask, what are the possible benefits of using absolute value over sum of squares?

For the rectangle pictured here

http://upload.wikimedia.org/wikipedia/commons/f/fa/Rectangle.svg

suppose that the point A represents the average scores and point C represents the scores of an individual judge. How far is C from the average?

The sum of the absolute values is the length of the path that goes from A to B and then up to C. This length is |a| + |b|.

The sum of squares method goes straight along the diagonal. By the Pythagorean Theorem this distance is SQRT(a[SUP]2[/SUP] + b[SUP]2)[/SUP].

(There is also an "n" that sneaks in there, but never mind that.)

 

[ILuvYuna]

Adelina's Scores: Suspicious vs. Not-so-much

TES

Copied Yuna's, deleted all data first, and then re-built it using Adelina's numbers.

First, the individual column scores that actually counted for each GoE (hi and lo X'ed out, cheating block in red.)
This shows the hypothetical strategy (if 7 votes high, all scores from 1,2, and 5 will count. Keep in mind that a lot of the judges gave the same hi & lo mark. Since we're assuming the 1,2,5, and 7 were cheating, I dropped the hi & low from the other columns whenever possible. This is simply to illustrate the feasibility of the strategy).

Row #1.......1....2....x....4....5....6....x....8....9
Row #2.......1....2....3....4....5....6....x....x....9
Row #3.......1....2....3....4....5....x....x....8....9
Row #4.......1....2....3....4....5....6....x....8....x
Row #5.......1....2....3....4....5....6....x....x....9
Row #6.......1....2....x....4....5....6....x....8....9
Row #7.......1....2....3....1....5....x....x....8....9
Row #8.......1....2....3....4....5....6....x....8....x
Row #9.......1....2....3....4....5....x....x....8....9
Row #10......1....2....x....4....5....6....x....8....9
Row #11......1....2....3....4....5....x....x....8....9
Row #12......1....2....3....4....5....6....x....8....x

Now the scaled GoE breakdown. Set up the same way as Yuna's, starting w/the "curve" setter in the cheating block (Column 7),
and starting w/the fence-sitter/questionmark judge in non-suspicious columns (Column 6)

JUMPS...............BV..... .....7.......5.......2......1...........GoE/Actual/GoE..........6.......3.......4.......8.......9
3Lz+3T............10.10............3.......2.......2.......2...........1.58/1.00/0.42..........2......-1.......1.......0.......1
3F+2T+2Lo.......9.24x...........-1.....-1......-1.....-2...........-.88/-.90/-.98.........-1......-2......-1.....-2......-1
2A+3T.............8.14x............3.......3.......3.......2...........1.93/1.80/1.68..........3........2.......3.......2.......2
3F....................5.30............3.......2.......3.......2...........1.75/1.50/1.26..........3.......1........2.......1.......2
3Lo...................5.10............3.......3.......3.......2...........1.93/1.60/1.40..........2.......2.......2.......2........2
3S...................4.62x............3.......2.......2.......2...........1.58/1.20/0.98..........1.......2.......1.......1........2
2A...................3.63x............3.......2.......3.......2...........1.25/1.07/0.90..........2.......2.......2.......2........1



Distribution I:
.................-2GoEs = 3..........................-1GoEs = 7 ..........................0 GoEs = 1
(0%)Cheating Block = 1 (33%)..........Cheating Block= 3 (43%).........Cheating Block = 0 (0%)
(0%)Other Columns = 2 (67%)..........Other Columns= 4 (57%).........Other Columns = 1 (100%)

..........+1GoEs = 8 ................................+2GoEs= 29 ..........................+3GoEs = 15
Cheating Block = 0 (0%).................Cheating Block= 12 (41%).........Cheating Block = 12 (80%)
Other Columns = 8 (100%)..............Other Columns= 17 (59%)..........Other Columns = 3 (20%)

Distribution II:
................Total GoE's............%=-2........%=-1........%=0........%=1........%=2...........%=3
Cheating Block........28..............4%............10%.........0%..........0%..........43%.........43%
Other Columns........35..............6%............11%.........3%..........24%.........49%...........9%





SPINS..........BV..........7.......5.......2......1............GoE/Actual/GoE............6.......3.......4......8.......9
CCoSp4.......3.50.........3.......3.......3......2............1.36/1.21/1.10.............3.......2.......2......2.......2
FCSp4........3.20..........3.......3.......3......3.............1.50/1.36/1.20............3.......2.......3......2.......2
LSp4..........2.70...........3.......2.......2......2.............1.13/1.07/1.10............2.......2.......3......2.......2
(Spins are A's strength, but notice Cheating Block still scored her *consistently* higher than Non-suspicious Columns. My guess* is non-suspicious columns were more reserved because Yulia had already skated at this point, and her spins were more impressive. *uninformed guess, as I haven't look at Yulia's protocol yet)

Distribution I:
..........+2GoEs = 15........................................+3GoEs = 12
Cheating Block = 4 (27%).......................Cheating Block = 8 (67%)
Other Columns = 11 (73%).......................Other Columns = 4 (33%)

Distribution II:
......................Total GoE's.......%=2........%=3
Cheating Block........12..............33%........67%
Other Columns........15..............73%.........27%
(Same as Jumps: the Cheating Block awarded mostly 1's, while the Other Columns awarded mostly 2's)






STEP SEQ......BV..........7.......5.......2......1.... .......GoE/Actual/GoE........6.......3........4.......8.......9
StSq4..........3.90..........3.......3.......3......2...........1.93/1.70/1.40.........3.......1.......2......2.......2
ChSq1....... ..2.00..........2.......2.......3......2............1.56/1.50/1.54.........3.......2.......2......2.......2
(Notice most of the 3's for the step sequence that was called wrong are coming from the cheating block = they're trying to capitalize on the higher multiplier of L4. Also, notice that the only +2 (from the cheating block) for this wrong-call element, came from Judge 1. This is because it was Judge 1's job to score "safely" on everything - he/she scored 2's practically straight down the line. By "safe" I don't mean lowering the chances that they'll get caught, I mean increasing the likelihood that it will not be dropped, thus, ensuring that a decent high score will count towards the final GoE points for every element.)

Distribution I:
.........+1GoEs = 1...................................+2GoEs = 11........................................+3GoEs = 6
Cheating Block = 0 (0%).................Cheating Block = 4 (36%).........................Cheating Block = 4 (67%)
Other Columns = 1 (100%)..............Other Columns = 7 (64%)..........................Other Columns = 2 (33%)

Distribution II:
......................Total GoE's.......%=1............%=2............%=3
Cheating Block........8.................0%............0%... ...........0%
Other Columns........10.............100%............0%. .............0%







PCS

First, the individual column scores that actually counted for each component (hi and lo X'ed out, cheating block in red.)
This shows the hypothetical strategy (if 7 votes high, all scores from 1,2, and 5 will count)

SS.......1....2....3....x....5....6....x....8....9
T/L......1....2....3....x....5....6....x....8....9
P/E......1....2....3....4....5....6....x....x....9
Ch/C....1....2....3....x....5....6....x....8....9
Int.......1....2....3....4....5....6....x....x....9

Averages:

Adelina's PCS Group Average...................................Yuna's PCS Group Average
Cheating Block = 9.59............................................Cheating Block = 8.96
Other Columns =9.02.............................................Other Columns = 9.49

Breakdown:

Components.......7...........5..........2............1.........................................6..........3..........4.........8..........9
SS.................9.50.......9.50......9.50........9.50.....................................9.00......9.25......8.75.....8.75.....8.50
T/L................9.50.......9.25......9.25........9.25.....................................9.00......8.75......8.50.....8.75.....8.50
P/E................9.75.......9.75......9.50........9.75.....................................9.50......9.25......9.00.....9.00.....9.25
Ch/C..............9.75.......9.50......9.75........9.75.....................................9.75......9.25......8.75.....9.00.....9.50
Int.................9.75.......9.75......9.75........9.75.....................................9.25......9.00......9.25.....8.75.....9.25

Ind.Avg...........9.65.......9.55......9.55........9.60.....................................9.30......9.10......8.85.....8.85.....9.00

Adelina's Actual TES: 74.41
And if you apply the the 1.6 factor to each component, for each judge:
Judges........7...........5..........2............1.........................................6..........3..........4.........8..........9
TES........77.20......76.40......76.40........76.80.....................................74.40......72.80......70.80.....70.80.....72.00

Yuna's Actual TES:74.50
Yuna's scores w/1.6 factor applied:
Judges........2...........4..........7............8.........................................1..........3..........5.........6..........9
TES........67.20......72.40......74.80........73.60.....................................73.20......78.00.....75.20.....76.40.....76.80




If you were to drop the cheating block completely. Remove high and low from the non-suspicious columns, and take the average PCS, for each skater:

Adelina PCS: 71.87
Yuna PCS: 76.13

A margin of +4.26 PCS in Yuna's favor.

I'd like to do TES the same way - Adjust BV of Step Sequences for both skaters, Adjust Adelina's 1st combination to show edge call and underrotation downgrade, drop the cheating blocks completely, remove high and low from the remaining columns, use multipliers and average to determine GoE.


And if that makes more sense in terms of what the skater's actually delivered, and what kind of pcs they were capable of, well... then at least it'll be here for everyone to see hwell:


***Disclaimer*** There might be mistakes, if you find any,just point them out and I'll fix it..
**ETA fixed some colors, and wording under spins.
**ETA2 - Judge1 and 7's scores in the individual pcs scores were in each other's place by accident. fixed now.

 

[ILuvYuna]

[SUP][/SUP]

For the rectangle pictured here

http://upload.wikimedia.org/wikipedia/commons/f/fa/Rectangle.svg

suppose that the point A represents the average scores and point C represents the scores of an individual judge. How far is C from the average?

The sum of the absolute values is the length of the path that goes from A to B and then up to C. This length is |a| + |b|.

The sum of squares method goes straight along the diagonal. By the Pythagorean Theorem this distance is SQRT(a[SUP]2[/SUP] + b[SUP]2)[/SUP].

(There is also an "n" that sneaks in there, but never mind that.)

Aoohhh, so in other words, sum of squares gives you less wiggle room, while absolute values give you more wiggle room, right?

 

[chloepoco]

The most likely scenario is the conspiracy starts at the very top of the ISU. Probably Cinquanta and his minions received large sums of money from Putin, and made sure the tech panel and certain judges would fix the results.

Not just the ladies scoring was fraudulent, but even things like V/T's scores when they made clear mistakes, Plushenko, all the Russian competitors really being overscored. No doubt it started at the very top.

Are you seriously accusing Cinquanta of accepting monetary bribes?

 

[ILuvYuna]

Yuna's Scores: Suspicious vs. Not-so-much

TES

JUMPS..........BV.......2.......4.......7......8...........GoE/Actual/GoE.........1.......3........5.......6.......9
3Lz+3T.........10.10......2.......2.......3......2...........1.58/1.60/1.68..........2.......3........2.......3.......2
2A+2T+2Lo....7.04x.....1.......1.......2......1...........0.63/0.79/1.00..........2.......3........2.......2.......1
3Lz...............6.60x.....1.......3.......1......1...........1.05/1.00/1.12..........2.......2........1.......2.......1
3S+2T...........5.50......1.......1.......1......2...........0.88/1.00/1.12..........1.......2........2.......2.......1
3F................5.30......1.......1.......1......2............0.88/1.20/1.54..........2.......3........2.......2.......2
3S................4.62x.....1.......1.......1......1...........0.70/0.90/1.26..........3.......2........1.......1.......2
2A................3.63x.....1.......1.......2......2...........0.75/0.79/0.80..........2.......1........1.......2.......2

SPINS..........BV..........2........4.......7.....8.............GoE/Actual/GoE............1.......3........5.......6.......9
FCCoSp4........3.50..........2........1......2......2.............0.88/0.93/1.00.............1.......3........2.......2.......2
CCoSp4..........3.50..........2........1......2......1.............0.75/0.71/0.70.............2.......1........1.......2.......1
LSp3.............2.40...........1........1......1......1.............0.50/0.64/0.80.............1.......2........2.......2.......1

STEP SEQ......BV..........2........4.......7.....8............GoE/Actual/GoE........1.......3........5.......6.......9
StSq3............3.30..........2........2......2......3............1.13/1.14/1.00.........2.......3........2.......3.......2
ChSq1...........2.00...........1........2......3......2............1.40/1.50/1.54.........2.......2........2.......3.......2

W/out tossing high and low, or making adjustments for proper calls - this is purely what each block thought of Yuna's element execution:

Cheating Block: 11.13(GoE) + 57.49(BV)= 68.62 TEs
Other Columns: 13.56(GoE) + 57.49(BV)= 71.05 TES

Adelina's Scores: Suspicious vs. Not-so-much

TES

JUMPS...............BV..... .....7.......5.......2......1...........GoE/Actual/GoE..........6.......3.......4.......8.......9
3Lz+3T............10.10............3.......2.......2.......2...........1.58/1.00/0.42..........2......-1.......1.......0.......1
3F+2T+2Lo.......9.24x...........-1.....-1......-1.....-2...........-.88/-.90/-.98.........-1......-2......-1.....-2......-1
2A+3T.............8.14x............3.......3.......3.......2...........1.93/1.80/1.68..........3........2.......3.......2.......2
3F....................5.30............3.......2.......3.......2...........1.75/1.50/1.26..........3.......1........2.......1.......2
3Lo...................5.10............3.......3.......3.......2...........1.93/1.60/1.40..........2.......2.......2.......2........2
3S...................4.62x............3.......2.......2.......2...........1.58/1.20/0.98..........1.......2.......1.......1........2
2A...................3.63x............3.......2.......3.......2...........1.25/1.07/0.90..........2.......2.......2.......2........1

SPINS..........BV..........7.......5.......2......1............GoE/Actual/GoE............6.......3.......4......8.......9
CCoSp4.......3.50.........3.......3.......3......2............1.36/1.21/1.10.............3.......2.......2......2.......2
FCSp4........3.20..........3.......3.......3......3.............1.50/1.36/1.20............3.......2.......3......2.......2
LSp4..........2.70...........3.......2.......2......2.............1.13/1.07/1.10............2.......2.......3......2.......2

STEP SEQ......BV..........7.......5.......2......1.... .......GoE/Actual/GoE........6.......3........4.......8.......9
StSq4..........3.90..........3.......3.......3......2...........1.93/1.70/1.40.........3.......1.......2......2.......2
ChSq1....... ..2.00..........2.......2.......3......2............1.56/1.50/1.54.........3.......2.......2......2.......2

Same for Adelina:
Cheating Block: 16.62(GoE) + 61.43(BV)= 78.05
Other Columns: 12.00(GoE) + 61.43(BV)= 73.43



Breakdown for PCS scores:

Adelina's Actual TES: 74.41
And if you apply the the 1.6 factor to each component, for each judge:
Judges........7...........5..........2............1.........................................6..........3..........4.........8..........9
TES........76.80......76.40......76.40........77.20.....................................74.40......72.8......70.80.....70.80.....72.00

Yuna's Actual TES:74.50
Yuna's scores w/1.6 factor applied:
Judges........2...........4..........7............8.........................................1..........3..........5.........6..........9
TES........67.20......72.40......74.80........73.60.....................................73.20......78.00.....75.20.....76.40.....76.80

Average PCS for each group:

Cheating Block:
Adelina = 76.70
Yuna = 72.00 (suspiciously even/rounded number, isn't it?)

Other Columns:
Adelina = 72.16
Yuna = 75.92



The results - not tossing out high or low, and not adjusting anything for wrong calls - this is just to see what each block thought of each skater:

Cheating Block:
Adelina = 78.05(TES) + 76.70(PCS) = 154.75
Yuna = 68.62(TES) + 72.00(PCS) = 140.62

Other Columns:
Adelina = 73.43(TES) + 72.16(PCS) = 145.59
Yuna = 71.05(TES) + 75.92(PCS) = 146.97

Without the influence from the cheating block, it was a much closer competition. Adelina bested Kim in TES, but Kim edged her out by virtue of PCS.

So let's see by show of hands (or keys? lol) -- which judging block do you all agree with?




From another perspective:

Adelina
Cheating Block: 78.05(TES) + 76.70(PCS) = 154.75
Other Columns:73.43(TES) + 72.16(PCS) = 145.59

Yuna
Cheating Block: 68.62(TES) + 72.00(PCS) = 140.62
Other Columns: 71.05(TES) + 75.92(PCS) = 146.97

********ETA I made a mess at the bottom portion, fixing now....
**ETA2 - fixed
**removed question for mathphiles, too exhausted to care about the answer lol

 

[Nadya]

2. Helene CUCUPHAT of France
- Responsible for 3 of the most important win in Sotnikova's career.
- Marked Adelina at Junior worlds 2012 (Where Yulia was 1st, Adelina 3rd)
- Marked Adelina at European Championships which saw her PCS rose to the 7th highest of all time despite coming 2nd.
- Marking Adelina again at Sochi.

3. Zana KULIK of Estonia
- Responsible for Adelina's most critical win up to date, 2011 Junior world championships.
- Judged here at Sochi which makes a nice Eastern bloc judging if skating history has proved to rear its ugly head again here in Sochi.

How can any one judge be RESPONSIBLE for any of the judging outcomes made by a panel of nine? Did you just use that word because it felt nice? Do you think the other eight judges on these panels were out having fish and chips while Cucuphat and Kulik were busy being, I dunno, RESPONSIBLE for something?

 

[Nadya]

More like they were scared of their lives. Knowing Putin, I am not surprised. Just kidding, lol. As for Plushenko, had he not withdrawn, he would definitely medal. If he had not fell, then it might have be gold.

And seeing how the other men skated, that would not even have been all that controversial.