Sunday, May 20, 2012

IDOLMETRICS: SONG CREATIVITY/SINGING MOMENTUM, WEEK BY WEEK

As you've most likely read from my previous mathematics (or some would say Idolmetrics) studies, these calculations are based on song creativity and how well the contestant sings their song. If they sing a song that's been done a ton on Idol, they get docked points because their score is then compared to the average of that song. If they do a new song, the contestant receives a straight up score (that explains why Phil Phillips routinely ranks high; he needs to put less effort to produce the same amount of impact as a Joshua Ledet or Jessica Sanchez, given how often they choose songs that have been done many times). The numbers are scaled and added to each other each week to give a composite of the contestant's momentum below.








The spreadsheet data  above only accounts for the contestant's singing ability. Now, below we account for the voting patterns, where if a contestant is in the bottom two despite having superior scores, they get docked off points accordingly (below is therefore the realistic representation of how much momentum the contestants truly have):



























As you can see, the results are quite accurate; I predicted Erika Van Pelt and Colton Dixon were a bit overhyped (Erika in particular was plagued with very low creativity) and I was able to calculate Phil Phillips' momentum without accounting for some fudge factor to account for Southern voting (I kid, I kid). But this is a realistic composite of how voting has gone. Nigel has gone on record saying voting has fluctuated, so obviously this isn't true, but it's good food for thought anyway. He said Elise led the votes when she did 'Whole Lotta Love"; I predicted she was fifth in pure singing and sixth in votes at the time, so clearly I was way off. He also said Skylar and Colton were leading the votes: at first I thought Colton was, but he didn't have enough moments, as I mentioned in another Idolmetrics post. Skylar however showed up well in some of my pure singing algorithms.

I'll be putting up another post focusing in depth about how the contestants' scores are calculated soon.

No comments:

Post a Comment