For Yuna's scores, the four I would single out for suspicion are columns 2, 4, 7, and 8 (
http://korcan50years.files.wordpress.com/2014/02/yuna-kim.png)
The way to determine which are more likely acting in collusion is to consider what they marked on each element.
For example:
In TES, Yuna opened with her biggest bv element, Row#1 (10.0), which she is capable of executing very well:
Judge Columns..........2.......4......7......8
3Lz+3T....................2.......2......3......2
Toss high and low, and Yuna averages +2 GoE (Actual Average before sov +2.38)
Her next highest bv is Row#7 (7.04x)
Judge Columns..........2.......4......7.....8
2A+2T+2Lo..............1.......1......2......1
Toss high and low = Yuna averages +1 in GoE (Actual Average before sov +1.57)
Her third highest bv is Row#6 (6.6x)
Judge Columns..........2.....4....7.....8
3Lz.........................1.....3....1.....1
Toss high and low = Yuna averages +1GoE (Actual Average before sov +1.43)
Fourth highest bv is Row#3 (5.50)
Judge Columns..............2.....4....7.....8
3S+2T.........................1.....1....1.....2
Toss high and low = Yuna averages +1GoE (Actual Average before sov +1.43)
Fifth highest bv is Row#2 (5.30)
Judge Columns..............2.....4....7.....8
3F..............................1.....1....1.....2
Toss high and low = Yuna averages +1GoE (Actual Average before sov +1.71)
***********************************
As far as PCS, let's look at two different charts, comparing the scores that counted toward the actual average with the scores given by 2,4,7,8 working in collusion:
A. Scores that counted toward the actual average, from Columns NOT in collusion w/ 2,4,7, and 8.
Judge Columns.........1..........x.........3..........x..........5.........6.........x..........x...........9..........Averages:
SS.......................9.00....... x.......9.50.......x........9.50.....9.25.......x..........x...........x...........= 9.25
T/L......................8.75........x.........x.........x........9.25.....9.00........x..........x........9.25.........= 9.06
P/E......................9.25........x.......9.75.......x........9.25.......x..........x..........x........9.50.........= 9.44
Ch/C....................9.50........x.........x.........x........9.50.....9.50........x..........x........9.75.........= 9.56
Int.......................9.25........x.........x.........x........9.50.......10........x..........x........9.75.........= 9.63
Total Average.................................................................................................................=9.39
B. Columns in Collusion. The averages in this chart are taken after tossing out their hi and lo scores. They would probably take this into consideration when devising their strategy, because they don't know how she will actually compete or how the other judges would score her, but they are betting she will compete well, and the other judges will give her high scores).
Judge Columns.........
2..........4.........
7.........8.........Averages
SS.......................
8.50.....9.00.....
9.25.....9.00......= 9.00
T/L......................
7.75.....8.75.....
9.00.....8.75......= 8.75
P/E......................
8.75.....9.25.....
9.50.....9.50......= 9.38
Ch/C....................
8.25.....9.00.....
9.25.....9.25......= 9.13
Int.......................
8.75.....9.25.....
9.75.....9.50......= 9.38
Total......................................................................................................................................= 9.13
The difference is negligible, but remember that as the Chart B strategy is being devised, they are counting on Yuna both skating well
and scoring well with the judges outside of their group. Notice that at the end of the day, although the other judges may not have scored her as well as ancipated, the cheaters' strategy was still effective. Their pre-planned scores were lower than what the other judges awarded her (the goal was to sink the score as much as possible, both in TES and PCS) One last time, side by side, blue scores from columns not in collusion, red cheating strategy, purple actual average:
SS.....=....
9.25....
9.00........
9.21
T/L....=....
9.06....
8.75........
8.96
P/E....=....
9.44....
9.38........
9.43
Ch/C..=....
9.56....
9.13........
9.39
Int.....=....
9.53....
9.38........
9.57
Finally, the individual column scores that actually counted for each component (w/columns in red scoring in collusion):
SS.......1....
x....3....
4....5....6....
7....
8....
x
T/L......1....
x....
x....
4....5....6....
7....
8....9
P/E......1....
x....3....
4....5....
x....
7....
8....9
Ch/C....1....
x....
x....
4....5....6....
7....
8....9
Int.......1....
x....
x....
4....5....6....
7....
8....9
Column 2's role seems to have the inverse of effect of column 7 on Sotnikova's sheet - it "sets the curve" to increase likelihood that the
low scores given out by the other columns in collusion will be included in Yuna's score.
***DISLCAIMER: This proves absolutely nothing lol. Even if you did this for all of the skaters, it's still proves nothing. I am also not a mathematician, and I've put this together speedily over the past hr and a half, so there may be mistakes. Feel free to check my work and/or challenge my uninformed hypothesis. I enjoy a good debate
ETA - fixed colors on the last chart