Right now when the 32 voters are ranked from Fluffer at #1 to Cynic at #32 you rank #20 (which is a tie with you and @Southern_Dan) .
If you look at just the 14 people that have voted on every poll (of which you are one), then you rank at #11 (again tied with Dan)
You voted with the winning tier (on a popular basis) 46% of the time, and voted with the winning tier (on a weighted basis) 50% of the time. Additionally when you aren’t voting with the winning tier your vote is typically lower than that tier.
Don’t everyone ping me for something like this, I’m just providing it to Sociable here because I put a lot of the leg work into formulas and automating stuff in my excel sheet because it was a slow day at work, and figured I’d throw it out since it’s fresh data.
At the end of this, in addition to the “universal” stats or whatever I was thinking about giving all the frequent voters a breakdown similar to the above. If you have ideas for what it should include let me know
I haven’t tried working on the “UMfriend” idea yet, so not sure how feasible it is.
This is an issue with 2nd Self and 40’s Theme right now. 2nd is a popular vote tie between B and C, but is weighted to a B. 40’s is a popular vote tie between A and B, but is weighted to a B. Funny enough 2nd Self, even though it was a B/C tie actually ranks higher on a weighted basis (and weighted modified by consensus basis) than 40’s which was an A/B tie.
An alternative could be to give it a + or - rating. In your example for 40s, you could give it an A- or B+. Since it’s weighted lower, give it the lower of the two; B+.
This probably has unintended consequences that I’m not thinking of.
Or just list it as “A-/B+” and leave it at that for your rankings.
I don’t think removing it from popular vote analyses is the right answer though.
I thought about the plus/minus grades but it’d add more complications to the consensus formula and what I’ve started working on for the UMfriend stuff.
When it comes to the songs ranked by weight I’m already going to have markers for the crossover between S/A and A/B etc so it’ll be somewhat intuitive what the minuses and pluses are, but bringing those into a definitive number adds a lot more complexity to what I was already planning. It’s certainly matmatically possible, but it would cloud most other analysis I was planning.
