BlastFromThePast
Members-
Posts
25 -
Joined
-
Last visited
Everything posted by BlastFromThePast
-
UIL 6A Area F 2025
BlastFromThePast replied to BandPerson101's topic in Past Contests, Events, and Festivals Archive
Interesting exercise here. If each judge's scores were normalized to the range 0.1-1.0 and added together instead of converting them to ordinals and adding those together (as currently done), these are the results: Oak Ridge and Grand Oaks swap places, as do Klein and Klein Collins (both by non-trivial amounts too). Not to take anything away from any band, but this more accurately represents the judges' true scores. Normalization preserves each judge's relative weighting per band (which ordinals do not) yet also keeps each judge's score equally weighted in the total (which is the argument put forth in favor of the ordinals calculation.) I can only think the results of other UIL competitions are similarly affected, perhaps even more significantly in some cases. Just to be clear: I don't have any horse in this race, except for mathematical fairness. I am passionate about this activity and would just prefer to see a better scoring calculation for all these hard-working students and TX UIL. Judging marching contests is subjective enough. The ordinals scoring calculation throws unintended noise into the mix. IMHO, using the total of normalized scores would yield more accurate, and thus fairer, comparisons. And it's a minor change too -- just in how the total is calculated. Nothing about judging itself would be affected. I'll be repeating this exercise on the State results. -
UIL 6A Area F 2025
BlastFromThePast replied to BandPerson101's topic in Past Contests, Events, and Festivals Archive
Congrats to all bands, but don't get me started on ordinal points. It's a mathematically challenged system. -
The UIL rubrics have changed a lot in just the past few years, and have been split into separate branches: Open Class and Military Class. So, UIL Open Class is much more aligned with BOA rubrics now. Perhaps that will eventually play out with TX bands taking home the Eagle more often. In general, TX bands do seem to be getting more competitive: The BOA San Antonio Super Regional and UIL State contests are very high level, arguably the highest nationwide as a whole. There just seem to be the few perennial exceptional bands (Avon, Carmel, Broken Arrow) that manage to stay on top at GN, though TX bands did break through in 2007, 2013, and 2019 (seems TX is due in 2025 ) As for "limitations of practice" -- I cannot imagine H.S. bands practicing any more than they already do! How would that even be possible and still get other things done such as, oh, class work ? What more do those consistent GN top placement bands actually do that don't have these limitations? Or are there other factors? But I suppose this is a discussion for another place.
-
The ordinals system is mathematically, er, "challenged". It introduces an element of randomness (you could say spiciness) into the overall rankings. Thus there can be position shuffling in the overall placements that have little to do with the judges' actual scores. Without seeing the score sheets, I can't say that happened in this particular contest. But being a "fuzzy" subjective contest in the first place, IMHO it'd be better if the scoring system wouldn't add this additional spice. Then again, maybe it just adds to the excitement.
-
Agree that scores from different contests can't be directly merged and compared like that. Yet, I also agree it is useful. If you look at any given band's BOA scores throughout the season, they nearly always increase from contest to contest (e.g., Regional -> Super Regional -> GN). Placement changes from contest to contest happen because a band improves their score over time more than other bands, rather than bands' scores declining. So it seems reasonable to assume that most, if not all, of the SA finals scores represent a minimum of what they would have scored at GN (barring a major disaster of course). What's more unpredictable is how much higher these bands would have scored at GN.
-
2022 6A UIL State
BlastFromThePast replied to BandFan59's topic in Past Contests, Events, and Festivals Archive
Yep. Mapping each judge's score to an ordinal before adding them together introduces statistically random "noise" into the total score for each band. Perhaps this is what makes UIL results more surprising/unpredictable than other contests -- the element of randomness. Maybe this approach is meant to counter some of the subjectiveness.
