The "Corridor" of Judging - Is there an official explanation of what it is?

The "Corridor" of Judging - Is there an official explanation of what it is?

Judges in a competition are anonymous, so we don't know how individual judges actually rated a competitor. We do know the consensus of the scores when looking at the protocols but how the consensus was reached is not known.

I believe (tell me if I'm wrong) the judges must score a skater within a "Corridor" of scores.

How does that work? Is there a list of contestants given to the judges which shows the limit one can judge an individual skater?

Help!

There's this ISU Communication 1631. (revised every 2 years or so.)

http://isu.sportcentric.net/db//files/serve.php?id=1951

The “corridor” for GOEs means that if you are a judge, the mark you give for each element should be, on the average, within plus or minus one point of the average score given by the whole judging panel.

For instance, if the average judges’ score for GOE on a a skater’s triple flip is is –0.8 and you give a –2, then that counts as being “off” by 1.2 points. Then you add up all the amounts by which you are off. If there are, say, seven elements in a ladies short program and you are off by a total of more that 7 points, then you are “outside the corridor” for that skater and your judging comes under review.

For program components it works a little bit differently. It is only the total amount of deviation from the mean that counts. So if you score a skater’s Skating Skills higher than the average of the panel, but score that same skater’s Interpretation lower than the average of the panel, then those cancel out and you are OK.

But if you score every one of the program components at 1.5 points lower than the average of the panel, then you are “outside the corridor.” The idea is that in this case there is a suspicion that you are deliberately trying to hold down this skater due to bias, politics, etc.

The judges do not "have to" score within the corridor (in fact, they do not know ahead of time exactly what the corridor will turn out to be for any particular element for any particular skater.) But the overall effect of these corridor rules is to make the judges a little bit afraid to go out on a limb. They tend to give marks that are in line with what they expect all the other judges to give.

(You can check out the details on pages 5 and 6 of the ISU document in sorcerer's link.)

Thank you Sorcerer for the link and thank you MM for the point that the corridor only pertains to the GoEs.

I would gather since there is fear involved in adding plus or minus to the base value, and since judges do not know what the corridor will turn out to be that Reputation is at play. No?

(my own solution to this would be to judge an element on a scale of 0-10. No need for GoEs; no need for a Tech panel. It would be like other judged sports.) :think:

Joesitz said:

I would gather since there is fear involved in adding plus or minus to the base value, and since judges do not know what the corridor will turn out to be that Reputation is at play. No?

Click to expand...

Bingo.

Joesitz said:

I would gather since there is fear involved in adding plus or minus to the base value, and since judges do not know what the corridor will turn out to be that Reputation is at play. No?

Click to expand...

Yep. It's also why you mostly don't see much differentiation between the five marks, they are usually all within a small range. It has no relationship to the program or the performance of it. Judges almost never agree on which of the components was highest or lowest, or the relationship between all of them, except if a certain skater has a reputation for being really good or bad at something: Skating Skills, Transitions, whatever.

But then again, a skater gets a reputation for skating fast by skating fast. He gets a reputation for doing a lot of transitions by doing a lot of transitions.

The next time out, he does not skate slower or leave out his transitions. He does not abandon his choreography. So if the judges say, before he takes the ice, "this is a fast skater with good transitions and interesting choreography" -- are they wrong?

Mathman said:

But then again, a skater gets a reputation for skating fast by skating fast. He gets a reputation for doing a lot of transitions by doing a lot of transitions.

The next time out, he does not skate slower or leave out his transitions. He does not abandon his choreography. So if the judges say, before he takes the ice, "this is a fast skater with good transitions and interesting choreography" -- are they wrong?

Click to expand...

Of course not. But that's not what happens the vast majority of the time, and it's not why there's reputation scoring. And you know that.

No, Patrick Chan did not deserve any kind of hit on SS and TR for his programs at Skate Canada. However, Takahiko Kozuka didn't suddenly turn into an 8.5 SS skater from the end of one season to the beginning of the next. Reputation scoring is a reality, even if it sometimes leads to a correct result. A stopped clock is right two times a day.

Judges really are well trained, very knowledgable people (well most of them).
Even in the 6.0 system, if a judge's mark was way out of line they had to meet with the referee after the event and account for their mark. This is part of the system.
Judges don't just see skaters for the first time in the competition. They attend the practices to familiarize themselves with the programs and the content. You will see them sitting in one area of the seats usually near the VIP section of the stands at practices.
This is standard procedure. The skaters and coaches know it. Sometimes you put a big jump out in practice that isn't ready for the program yet just to let the judges know you have it or almost have it and that you are in that 'league'.
Judges also meet with skaters and coaches to give them advice and pointers for improved scores. They may even go as far as suggesting music changes etc. This happens in monitoring sessions at their rinks, at meetings, and even less formally in the hotels after the competion is over. Coaches can arrange these meetings and there is more formal structured monitoring throughout the year for national and potential national level skaters. Before and during the actual duration of the competition however, judges will not speak to skaters or their coaches, even in the hotel.
For every skater at every competition, a program content sheet must be handed in at registration outlining the elements of the skater's programs, right down to the exact time they occur (ie. 3A at 39 seconds). Nothing in the program comes as a surprise to any of the technical panel or the judges.
This is all normal and part of skating.

Mathman said:

But then again, a skater gets a reputation for skating fast by skating fast. He gets a reputation for doing a lot of transitions by doing a lot of transitions.

The next time out, he does not skate slower or leave out his transitions. He does not abandon his choreography. So if the judges say, before he takes the ice, "this is a fast skater with good transitions and interesting choreography" -- are they wrong?

Click to expand...

But is all this in line with the policy of the CoP? Isn't the CoP supposed to be judged on what one does in a given competition? If not then all we need is an Award Ceremony on Body-of-Work.

The nationalities of the judges are justified because there are nine judges, but the reputations of skaters should not be brought into the judging area. The CoP itself disallows it.

trains. Very good post on what judges are supposed to do. I don't think any of us question that. Some of us just wonder if it actually happens and whether their are extraneous thought also entering the picture.

Reputation being a factor is undeniable but to quite a limited extent IMO.
If it was all reputation no new faces will be able to come into the scene till the existing ones deteriorate because of age.
On the contrary for instance Lepisto made a good international debut simply with her flawlessness in her SP at Skate Canada 08? IIRC. She didn't need reputation to do that. She very much impressed the judges.
Do judges nowadays overmark Laura because of her reputation?
To an extent, probably yes, in the sense that the judges are now always unconsciously waiting to be impressed by Laura with her fluid transparency of skating.

So it's rather the "expectation" from each judge's personal memory of being impressed by that skater, than the commonly shared "reputation" of that skater.
And once you are impressed with something, you tend to keep wanting to be impressed the same way.
Very much human habit, I think.

therein lyes the problem .
the judge according to reputation-and what the judges think they are seeing and/or remembering seeing instead of actual skate.
so once you are know for ur's you are nailed for ur's despite cleaning them up.
the judges still see you doing them in their mind instead of seeing clean up-especially if not the skater the federation wants up there.
so they aren't seeing the actual skate. the judges are seeing how they think you are skating in their mind.

The corridor is very simple. Before a competition, the judges all meet out in the corridor and decide whom that are going to boost, and whom they are going to hold down.

Joesitz said:

But is all this in line with the policy of the CoP? Isn't the CoP supposed to be judged on what one does in a given competition? If not then all we need is an Award Ceremony on Body-of-Work.

The nationalities of the judges are justified because there are nine judges, but the reputations of skaters should not be brought into the judging area. The CoP itself disallows it.

Click to expand...

I think you are expecting more of the CoP than it can deliver. Yes, everything is supposed to to be fair, just, unbiased and objective. All judging systems are supposed to have those virtues.

Mathman said:

But the overall effect of these corridor rules is to make the judges a little bit afraid to go out on a limb. They tend to give marks that are in line with what they expect all the other judges to give.

Click to expand...

I wonder if this effect impacted Rachael Flatt's scores at Skate America. Theoretically speaking, is it possible that some judges wanted to give her higher scores but were afraid to because it would be out of line with her previous scores and the expected scores of other judges? I am not asking if this really happened, but rather whether a scenario like that is theoretically possible.

krenseby said:

I wonder if this effect impacted Rachael Flatt's scores at Skate America. Theoretically speaking, is it possible that some judges wanted to give her higher scores but were afraid to because it would be out of line with her previous scores and the expected scores of other judges? I am not asking if this really happened, but rather whether a scenario like that is theoretically possible.

Click to expand...

Quite possible, I would think. However, the corridor for PCSs is pretty wide. 1.50 for each component. To get an "anomaly" you would have to be off by 1.5 points for each component, or a whopping 7 points off the average for the total PCS.

So a judge would be pretty safe if he thought, well, Rachael usually gets 6.75, but that was a better than usual, so I will give her a 7.25.