LogFAQs > #935930281

LurkerFAQs, Active DB, DB1, DB2, DB3, DB4, DB5, Database 6 ( 01.01.2020-07.18.2020 ), DB7, DB8, DB9, DB10, DB11, DB12, Clear
Topic List
Page List: 1
Topic[VGMC] Video Game Music Contest 14 nomination thread, part 2
PIayer_0
03/20/20 1:12:01 AM
#107:


Nevermind, I am dumb and can't do math at 1 in the morning. Let's try that again without me embarrassing myself this time.

Less than 24 hours to go until the locking deadline, and I think I'm gonna sit on my last three supports just in case something special comes up. Don't really see anything I wanna jump on right now anyway (amazing, I know), but I'm keeping a close eye on Weight of the World and Adventurous Inventors, as well as a bunch of other late-game prospects I've noted in my pastebin.

Let's do some more nom analysis! We are now up to 61 participants, which gives a total of 1525 supports to spend including doubles. Ignoring the locking system, you would actually need 8 votes to guarantee a spot in the top 192, since everyone could coordinate perfectly so that 190 songs are at 8/x, one at 5/x, and the rest at 0/0. In that sense, the locking system provides a way better deal if you don't like the stress and really want the insurance of a 100% guarantee.

Of course, efficient use of supports like that never happens; people hang on to their songs as the cutoff moves higher, and the ones that wait too long get burned. We can tweak the numbers to assume that only a certain portion of support votes end up useful, i.e. they contributed to a song that made top 192. In VGMC 12, we had 56 participants with a bracket of only 128, which made the cutoff 5/4 where 49% of supports were used on songs that got in. VGMC 13 gave us 69 people with a cutoff of 4/4 for top 192, where a slightly better 56% of supports were useful.

For this year, if we take a generous assumption of 60% efficient use of the 1525 supports, 5 votes just barely guarantees you a spot, so it's basically equivalent to the locking system! There's just enough to go around that reaching 5/x gets you top 183, while 4/x is still a respectable top 228. On the other hand, if only something like 45% of supports end up useful because people can't keep up (or just forget they entered the contest), you could even have 4/x be enough for top 171 and 3/x good for top 228.

I mostly wanted to do this for myself, so seeing the scenarios now I personally feel like the locks have a fair threshold; it looks like 5/x will end up being at or slightly above the final cutoff, which is a reasonable premium to pay for the peace of mind of having your song locked early. Of course, this is all assuming that we stay at 61 voters; the more people join, the higher the expected cutoff. I'm more than willing to wait it out until Sunday just in case 4/x ends up being enough, but it's always tempting to go for those locks while I still can! May throw out some doubles tomorrow, which I've been holding off on so far.

---
-Abraham Lincoln
... Copied to Clipboard!
Topic List
Page List: 1