BEST-mix.net - DDR Simfile Tournament

Chrisketchum1 wrote:So in general, bigger file size, and consistant judges.

BLueSS picks from a pool of judges he trusts and has used in the past, generally speaking. That isn't to say he isn't up for new choices but most of us don't have the time/patience to judge both brackets. Fifteen files was about as much as I could bring myself to do, honestly. =P

Chrisketchum1 wrote:I would like to start seeing some really high quality MP3 without having to worry about file size. I personally want some 600kbps songs. I want to hear the full potential of my stereo system!

Pretty sure you won't be getting 600kbps mp3s, since, you know, THEY DON'T EXIST.

Hahahahahaha

Oh yeah, and good luck finding 6 judges who are willing to judge a DDR-styled simfile competition and not compete. We had 2 this time, so I hope you can find some more!

Chrisketchum1 wrote:My only complaint is that the file size is too much of a challenge.

How much before the packs get too large though? I got pretty close to bandwidth limits on my current hosting, and would have to upload all future packs to something like MediaFire. And what's a good size bump? I could probably see 20MB being a good, new maximum, but anything above that and you really need to learn how video compression works.

Chrisketchum1 wrote:I would like to start seeing some really high quality MP3 without having to worry about file size. I personally want some 600kbps songs. I want to hear the full potential of my stereo system!

I really hope your finger accidentally hit the "6" instead of the "3" on your 10-key. As mentioned earlier, an mps above 320kbps won't exist, and neither will a song with "better definition" in any format Stepmania will play.

With an increased file size, this also means the videos could be HD without having to worry about file size. Every video in best-mix 3 I was troubled by the quality. {Even "My Boy"}

HD, in Stepmania?? What miracle build are you using? Depending on your computer, SM can REALLY lag with large video files. I don't know if you have tried it with HD files; so if you have a file with HD video that works well on your computer, I'd love it if you could upload it for us to see how well it'll work on the rest of our computers. Additionally, HD video sources for songs entered here are SUPER RARE.

I would of liked it if the 3 judges were consistant & the 3 judges are not allowed to compete.

You know what, I would too. (at least the first part) The problem as lrxevan said, is getting those who don't want to enter (and almost everyone finished a file this time, knocking that out as a possibility; so then resorting to people from the opposite brackets. I was hoping to have judged both; but my pc crashed during bracket 1, and there was no way I could have gotten them done without serious delays to the tournament.

The way it is setup with them judging the other bracket of files, they cannot harm the good files that much; even if they were malicious enough to try. The judging is public, and any inconsistent judging can be called out and challenged at that point (especially if a wildcard is brought back in).


I'm trying to further conversation with my replies; not shut down or dismiss any of the suggestions! If it sounds like I don't think it'll work, I'll try to explain why (as i tried above) or would like more support for the new idea. Please keep the feedback coming!!!

Ah crap, I edited my reply to will-i-am and jammitch!'s posts.

Basically:

1) Wildcard - I'd LOVE to bring it back, but before it always took a week (any shorter than half a week and everyone complained that it wasn't enough time to vote) and interest/forum participation DIED DOWN A LOT!

2) "Math based" bonus points - this still relies on the pr/judges' scores a lot, and since those judges vary a lot between the two brackets; a file could arbitrarily get a better bonus by being in a bracket with files that got less points compared to it. If there was only 1 good file in B1 it'd get a much higher bonus compared to the 4 good songs in B2 that had to share the bonus. Then the bonus points are more unfair due to the bracketing setup.

jammitch! wrote:Either way, nothing we do can really normalize scores between brackets.

A math based system would still be an attempt to normalize scores that shouldn't be put together due to all the differences. With the rank bonus, it is less "precise", but it doesn't try to fix something that's broken; it instead uses its own point valuation bonus.

An interesting idea; would it be preferred if the top files from each bracket were ranked together by score? This may work; but then I could see people trying to judge harder on the bracket they're not in so EVERYONE in that bracket gets lower scores, and thus a lower score for the bonus when the top from each bracket are put together.

My reasoning this time, was an estimate based on past results that the #1 files from each bracket would be close to the top 1-3 ranked files. The #2 files from each bracket likewise would be around the top 3-5 ranked files, and so on. It can happen, but is rare for a 4th place file in one to end up higher than a 1st rank file in a different bracket. The point values I used were totally arbitrary this time; just a "well, half a point is as good as any value at tie-breaking" because that's all I meant it to be for (and how it got used).

I'm not a huge fan of ties; but if this is too much work, it would be easier to just have files tie if need be like before (less work, less tournament complications, and less for people to possibly complain about if it goes wrong!).

will-i-am wrote:Wall of Text

I agree with pretty much everything you stated here sir. Oh, and the same thing feels like it kind of applied with Dj Potatoe. I had 2nd and 3rd place scores from Will-I-Am and Chmurek, but irxevan managed to just find everything he could wrong with it to drop me out of the top 3 so I couldn't survive.

However, it's all objective, and I got a couple votes in the Bonus Polls, so I'm not too upset.

I thought you said your file sucked.

It was just good enough for my "advancing to the finals" breakdown, but I didn't expect it to move on since taste in what makes a good step chart varies, especially with boss songs. I gave it a score that tied with MY BOY for 4th, btw, with 23.5 points (which somehow still hasn't been edited in the scoring sheet). Chmurek gave you the 2nd, Evan gave you 12th, and the PR's gave you 6th. I can see where you're coming from, but the final results mirror the PR's scores in this case and the public reviewers either liked it or disliked it, which was also the case with the judges.

But I still support Wilcard Rounds in the future, of course.

For the record, in most cases I had no idea who did which files when I judged them. jammitch! I knew, and there might have been one or two more. But I didn't lower Spork!'s score unfairly or anything. I was pretty honest and sometimes my opinions are weird.

The reality is everyone gets ripped off (or feels they were ripped off) in contests sometimes. In multiple cases I've submitted entries with bad steps to DDREvolved contests and they place higher than the files I submit with good steps because they have a video or better graphics or the judges just disagree about what makes a file good.

In the end the results aren't the most important thing. The community we have is small and dwindling and really these "contests" are just an excuse to get us together and force us to put out a pack and talk about them. I doubt my next contest file will even place. And TKO, one of my best files, was dj potatoe'd by a judge during some BEST-mix contest so I know how it feels.

As for your file specifically I'm never impressed when someone makes a very hard step file just for the sake of it. There was this air of inevitability regarding your file in some kind of cynical, sarcastic way before the contest even started. Like, "10s can never place, even when they're good." That isn't really true. I have nothing against 10s, I just didn't think it was very good and I didn't think the difficulty was justified. I'm don't think every note in a song, especially one that fast necessarily needs a note to follow it. I don't know that every held note needs a freeze arrow (because that isn't usually even what freeze arrows are for). I don't know that every sixteenth note needs a corresponding note. In most DDR charts the music isn't even followed all that closely. So there you have it.

Like I said, I'm not upset now, a couple hours after seeing the results I was just fine. I knew it wasn't going to succeed, but seeing some really positive reviews from some folks just really pushed my hopes up.

At least this one was taken a hell of a lot better than my last one, and it's a HELL of a lot better than my first contest sim.

BLueSS wrote:

The -2 late penalty would've knocked out any of the finalists

I would also make it very explicit that chances of making the finals late are very low

,
So in other words keep it at -2? or slightly less (1.5?)

No, what I meant was that when you write your blurb about the possibility of late entries (and I don't think the late entry thing should be repeated every contest, just when you think there should be more entries), you should write out explicitly something like "this penalty will make it very hard for any late entry to reach the finals; unless you think your entry will get perfect scores from everyone, the main value of entering late is to garner feedback for your file." You could cite statistical examples from this mix if you wanted to, but in any case I think it's a good idea to state something like what I just typed so that people don't enter late and then complain about the late penalty bumping them out of the finals.

It's up to you really to decide how much people should be penalized for entering late. One could argue it's not necessary at all since you already have that rule that if any late entry makes the finals, the top person that didn't make the finals will go as well (and so on for more than one late entry making the finals); on the other hand it's still not fair that late people get more time to work on their entries. As for the magnitude of the penalty if you have one, well, I'm not really sure what would be best. -2 seems fine but -1.5 would probably give people a fighting chance, -1 is a chance for sure. I dunno.
--
Regarding the wildcard/results:
You could have an automatic wildcard; out of the entries that didn't make it, the one with the highest score in the PR goes on. But you could also replace PR with "judging" and that would also be a reasonable possibility. You could also have both.

Regarding bonus points:
Yeah, the main problem is the difference in judges and in public reviewers between the rounds. Also, will, there's nothing that distinguishes the situation where two entries have a bonus difference of .5 (or, for that matter, any number between 0 and 1) and where they have a bonus difference of .01, because the public can only vote in integers. Either way, if entries x and y both have 5 votes from the public, it is the entry with the larger bonus that gets the higher final placing. Hence there would be no difference (at least not for Release Me and My Boy) in the system you proposed, if I read it correctly. One would still place higher than the other in a fashion that looks kind of arbitrary. You could change this by rounding the bonuses using whatever system feels appropriate. If you rounded all of jammitch's bonuses to the nearest integer, well, Come Back to Me ends up having no advantage over My Boy. You have to decide how much more an entry needs to place over another entry in order to have an advantage over that other entry in the final round. That itself is kind of arbitrary, but I guess it would still be fairer. But the whole argument does leave out bracketing. Bracketing itself leads to scoring being more arbitrary no matter which way you split it, though.

Edit: If something in this post looks wrong, it's because I'm really tired. I just edited it to insert the word "don't" in the last part of my first paragraph.