Best VGM Song contest 2006: Discussion

I made those next round pairings lists assuming the standard method would be used, but I am in favor of mixing up the process. I like your idea because there's still a ranking factor applied. It's unpredictability with an order to it.

Originally Posted by Aardork

Maybe it would be better if we made a list of all songs by the amount of votes they received (so that the song with the most votes would go against the song with the least, in the next round)? That way, it would be more likely that the good songs advance, as opposed to some worse songs moving to quarter-finals just because they were in a weak pairing. Also, that would mix the songs between brackets, which could be more interesting.

As long as it is going to help us with disasters like this one:

Quote:

Pairing XXIII
041 Ace Combat 4 Shattered Skies — Megalith -Agnus Dei-
169 Jets 'n' Guns — Shop Music

it's fine.

Seriously, though, first thing I want to say is that we definitely should mix the songs between brackets, that means two songs from the same bracket shouldn't meet each other in the next round. Second, the idea that the the song with most votes go against the song with the least is pretty much a two edged sword: On the one hand side, it would be a real dilemma to see pairings like the one above, on the other hand, we would absolutely block insider tip tracks from proceeding, since only the big names will come through. But I think that your ideas should be realized.

Actually that's a nice idea, but there's a huge problem - The voter participation. Pairing I naturally gets more attention than Pairing VIII, that's just how it is. Therefore you should compute the percentage share.

Okay, that's a good point. Well then, we can just take the percentage of votes, instead of the number.

Double Post:

Originally Posted by wvlfpvp

Like, a re-seeding? I dunno. I mean, it's an interesting idea, but this is a tournament, and as such, those initial seedings determine what songs go up against each other.

Yes, but I think it would keep things more fun, and make the final outcome more accurate. Of course, it would not be possible in a real tournament (sports), as the participants would object, but I see no reason why we couldn't do it in this situation.

Originally Posted by Aardork

Okay, that's a good point.

Nah, it's not. If you have a look at the results, you'll notice that the first pairings have like two (pairing B) or three (pairing C) more votes than the other pairings. Eventually, I'd say that everything's pretty balanced.

Well, I have not looked at all the results, but, for instance, Pairing I (C) has eight more less votes than Pairing I (A). It's not a huge difference, but it would change some potential future pairings. I mean, wouldn't it simply be more accurate to just take the percentage?

Oh, I thought he was talking about the difference of votes between the first pairings and the other pairings in ONE bracket. My bad, then.

If this is done and by percentage, it gives a big boost to middle of the road songs that have gone up against the not so well received. Case in point, 096 Sonic Advance — Egg Rocket Zone currently has the largest percentage at 80. Point total takes into account that a member may have abstained out of lack of interest in either.

Well, I actually meant it that way, Mush.

Pairing I (A) - 91 Votes
Pairing XV (A) - 83 Votes
Pairing XVI (A) - 78 Votes
Pairing XIII (A) - 71 Votes

Aardork, the question is, do you want to mix EVERYTHING or do you want to let the tracks stay in their associated bracket?

Hmm...I like the idea of taking the % into account and creating the matchups based on that. That way we'd avoid things like The Legendary Theme vs. The Final Fight and such in the 2nd round. Good idea.

Not sure how I feel about mixing songs between brackets. The other way to do it would be to keep the songs in the same brackets, but still use the %s to make the matchups.

I don't really like the idea of a reseeding, since I like knowing the paths tracks will have to take through the contest, but I wouldn't express outrage were it to happen. It'd defeat the purpose of seeding the tracks in the first place, though, but the numbers don't mean terribly much anyway.

If they are mixed up, though, I'd say randomly would be the best way to do it, in that it would be the easiest. Any plan to reseed based on how well the tracks performed in their matchings will NOT accurately reflect the track's "strength" and end up being mostly pointless beacuse the track's opponent has a big say in how the voting turns out. Percentage would lead to tracks like Egg Rocket Zone and Ocllo being top seeds because they drew weak opponents most people didn't like, and tracks like Nightless City Guara Bobelo ending up on the bottom because they had stiff competition. I think number of votes would actually be a little more "accurate," if you will, because the matchups between strong tracks tended to draw in more voters, although that system would be stilted by voter participation and would still end up placing things like Ocllo too high.

no offense isdapi

And for those who want a reseeding because there are tough matchups between great tracks that both should make it past the second round, I guarantee you that any system of reseeding that doesn't involve someone manually matching up every great track with a crappy track will result in at least one terrible, hair-pulling match, and even then it might not possible. That's just how these things go.

I'm normally all for surprises and shake-ups. Problem I see with rewarding the ones with the most votes are the tough match-ups with two really good tracks and about a 1-3 vote difference. Then those would end up going against the ones that got weak opponents and got a ton of points and two of the better tracks in the bracket would fight each other, while maybe two of the worst tracks got to fight it out and one still reaches the bracket quarter finals.

Case in point: Guara won by 3 votes in a tough match and 44 is a low overall number for a winner (I would think...low percentage anyway). So it would go against something that had a lot of points from beating on a weaker opponent. The scenarios vary too much to read too much into the vote counts...Nothing's guaranteed.

Maybe a better solution would be to ride the brackets as is until the bracket finals (when all the weak tracks are gone anyway), have people nominate their favorite tracks that have been eliminated along the way (like the honorable mention disc from 2004), take the top 16 nominations and create an extra round of the main 16 that have already advanced vs the "honorable mention" top 16. So technically, you'd be dealing with the top 32 tracks in the contest and all the weak ones will be gone already. That would also help in the event 1-2 brackets aren't good, while another one might have 6-7 tracks worthy of making the bracket finals.

Random brainstorm...I just remember in 2004, I wasn't only one saying how much more I liked the honorable mention disc than the "Top 16" disc.

Just a friendly reminder that it's been 3 days since Bracket E went up.

*is extra-eager for Bracket F*

Not everybody is as fast as you. I'll still need one day for Bracket D and if we don't want to kill this contest, we shouldn't hurry that much. That's at least the opinion of one of those slow voters. =(

Opening a new bracket doesn't necessarily mean closing an old one, though. D could be left open for another day.

But yeah, I'm just really impatient for these last few brackets (since they contain most of the songs I know / my noms).

Originally Posted by David Deluxe

Not everybody is as fast as you. I'll still need one day for Bracket D and if we don't want to kill this contest, we shouldn't hurry that much. That's at least the opinion of one of those slow voters. =(

That's right, but we also do not want this contest to protract, right? I mean, I always thought that it would be a good idea to keep this all as short as possible so that people wouldn't lose interest. Just my two cents, tho.

Closed the polls of bracket D (Pairing XI and Pairing XV are both almost tied, so I will give those two one more day). I'll start the next bracket in a few hours (whenever an admin creates the subforum).

I'm still thinking about what to do with the further rounds. I will probably create a poll to see whether most people prefer the normal seeding, or the method that I proposed.

edit: well, the forum hasn't yet been made, and I will go to sleep now, so you will have to wait a bit longer, music fans!

I've gotten a lot of requests for the arranged version of Global Network, so I'll link it here.

http://www.slightlydark.com/secretsq...%20Network.mp3

It's from a Dojin called Bird Eyes View / Acoustic Asturias, and I think the performing group is called Asturias. Since it's a Dojin, I can't update it at GMR, but I have put an entry containing all the info I have on it into the GFA database, so it'll be there when we launch. ^_~