"Glide-Pathing" or, IOW, "Position Slotting"

[ShortAndFast]

But he wasn't asking about the corps' scores getting better, he was asking whether corps' performances get better in every show - which if so would justify an increase in scores every show throughout the season, which as you pointed out is what happens.

Generally I think there's more variance in corps performances than in the linear trend of scores. I think judges are better than they used to be about responding to the night-to-night performance changes of the corps, but scores definitely move up predictably as we get closer to finals.

[MikeRapp]

Generally I think there's more variance in corps performances than in the linear trend of scores. I think judges are better than they used to be about responding to the night-to-night performance changes of the corps, but scores definitely move up predictably as we get closer to finals.

Maybe. But you won't convince me that virtually every corps improves their performance every night, and almost in lock step with the corps directly above and below them from the previous show.

"Scoring", in DCI, isn't really scoring NEARLY as much as it is relative ranking.

[ShortAndFast]

Maybe. But you won't convince me that virtually every corps improves their performance every night, and almost in lock step with the corps directly above and below them from the previous show.

"Scoring", in DCI, isn't really scoring NEARLY as much as it is relative ranking.

I totally agree. DCI judging does a good job with ranking, but it's a foregone conclusion that the champion is going to score a 98.1 +/- 1.0 regardless of how they compare to the previous year's champion.

[soccerguy315]

Pretty much, especially with the top corps. You could prolly count on one hand then times a corps score went backwards from one night to another. It doesn't happen often.

That's basically how it is

but there is a box just for how good it is already!!!!!!

[Pete Freedman]

Pretty much, especially with the top corps. You could prolly count on one hand then times a corps score went backwards from one night to another. It doesn't happen often.

For each corps, maybe. So far this year at CorpsFans.com I can see that while Bluecoats have not backslid so far, BD has done so once, Crown once, Cavaliers twice, Santa Clara twice and the Cadets four times. This is from one show to the next, not necessarily on consecutive nights.

Edit: CorpsFans.com not CorpsFans.org. Link added.

Edited July 31, 2016 by Pete Freedman

[Kamarag]

I totally agree. DCI judging does a good job with ranking, but it's a foregone conclusion that the champion is going to score a 98.1 +/- 1.0 regardless of how they compare to the previous year's champion.

Because the scoring system doesn't compare one year to the next. It never has, and it likely never will.

[Pete Freedman]

I would distinguish between glide-pathing and slotting.

The definition of slotting varies, but it almost always implies a fix of some kind as to who wins. Often it refers to fixed groupings of corps in the placement order that corps cannot escape from. Sometimes the writer/speaker refers to an inherent grouping held over from the previous seasons, and sometimes it is seen as a bias that becomes fixed in the judges minds after the first few shows. Either way, according to "slotting theory" only the Bluecoats, Blue Devils and Carolina Crown can possibly win this year because the judges fundamentally bias them into a group.

What also varies in the opinions about slotting is whether the judges are deliberately forcing the corps into these groups for some scandalous reason or are merely victims of bias (the more common view, I think).

Personally, I think it's all bunk. The appearance of slotting is what happens when some corps are actually better than others and stay better throughout the season because, well, they are better. People expect randomness to appear evenly distributed, but it isn't

Not that biases don't happen; I'm sure they do. But the evidence people use doesn't actually support it.

Back to the topic at hand, glide-pathing. I think they use a Rank, Rate, Calibrate system, with the calibrate part not mentioned publicly.

First: Decide who is better. Rank.

Second: Decide by how much. Rate.

Third: Adjust all the resulting scores together to fit roughly the expected curve. Calibrate.

In order for this to happen the chief judge must be giving some kind of indication of the expected range of scores. Either that or each judge must be following the scores and must do that themselves. Or the judges in fact talk to each other constantly behind the scenes about the recent numbers (I would) and self-calibrate as they go.

But I don't believe corps really get .5 better each day. More like .25 at best.

Here's a thought experiment: Take several judges who have only been judging the past four years or fewer. Now show them HD videos of the Colts show from five years ago and Blue Devils from that same year. Let them judge them as well as they can from the video. Then tell them that the Colts show was from Semis, but BD's show was from early in the season just after they broke 80. So Colt's score was in fact higher than BD's. It's easy to trot out whatever gobbledy-gook you want to use to explain the false equivalence, but it would be a lot harder to fool judges who are blind to all that and just evaluating what they see in front of them. (For the record, I think BD would cream the Colts. How could a judge seriously score the Colt's demand as equivalent to BD's?)

I don't really care though. They say you can't compare scores from different shows. Maybe now we know why - they've got their thumb on the scale, presumably to motivate the kids with how much their corps is improving day by day. I'm not opposed. If you can't be accurate at that scale, at least be motivating. After all, the system is designed to max out the student. That's their actual goal.