There are a total of three upsets half way through (which is 2 out of 32 matches). It feels rather low, for any kind of tournament. I’m more used to the 25% range. By upset I just mean versus seeding. Furthermore, none of these upsets are between similarly seeded entries–the inverse might sound much more unusual–songs that seeded one place better consistently beat their opponents (which happened…once?), and songs within 10 places of each other always put down their opponents (which happened a couple more times). The initial explanation I came up with the low percentage was that round one acted like a filter to eliminate any high(er) ranked songs that probably shouldn’t be, and proping songs that slipped through the nomination process. The other likely (and major) factor was that the voter pool might be relatively the same as the nomination ballet pool. Given all round 2 polls come in at < 150 votes, that is probably true.
There’s no real way to guess who will vote on the games initially, but the nomination system might be too precise, “too good” in a way, that makes doing a bracket less meaningful. With so many entrants, the only meaningful way to play is to come up with some model and plug the various songs in. So the safe bet this time happens to be the only realistic bet at winning, which is all-in based on the seeding. And when the seeding is way too close to the voter base, it is…no fun.
Here are the upsets:
- (223) Hybrid Rainbow over (95) Euforia (BBR2P4 or FH)
- (190) Katayoku no Tori over (62) Light Before We Land (BCR2P2 or FV)
I would even further suggest that Hybrid Rainbow’s upset is predictable because of its prior performance.
This leads to the last point. I think the tourney should post seeding on the matches. This helps people who have no game, to have a metagame, so to speak, as we shave down the number of songs and invariably people’s favorites will lose, leaving them disinterested. And then we can collectively scratch our heads as to why which song is seeded higher or lower is beating whichever song that is better or worse. It adds some unpredictability, basically.
For the prize pile, I have a couple boxes behind my desk of things I probably should give away, so it’ll be from there that I draw some music-related things. I hope Shut Up And Explode go far, LOL.