A thought about
OrionStats 2021 TTS that I keep meaning to post but forgetting...
Look at all the "surprising" things in this early sample:
Too many unexpected things to call out, and further down is just as crazy.
above
!
above
, who has a third the results of
!
and
only one step above the bottom!
is commonly put at the #1 character, as he's sitting here with 0.1% of overall results--a bottom 15 character.
The point isn't that Pikachu is actually bottom 15, nor is the point that these positions are so self-evidently wrong that the data is invalid.
This data sample is small, but it's not
small. It's about 10% of 2019 Phase 1. Depending on how you define and calculate it, that makes our "margin of error" about 3 times as big--not 10 times as big.
It's a very biased sample, but it's not
biased. It's almost 80% Japan, debatably the strongest region internationally. It's not like this is a bunch of events held in Argentina.
So it's a slice, and a strong slice, of our reality. It is biased but
agenda-less; there's no artificial motivation for including or excluding data. (@DasKoopa and COVID aren't sitting around a table plotting how they can make Sheik look good and Pika look bad.) You could honestly just think of it as a "What if?" ("What if the Americas didn't play Smash?")
That's the real takeaway: how asymmetric and diverse the playerbase is, and how much individual players and specific regions and contribute to the overall big picture. Even "representative" subsections of the scene will look radically different, but are no less a true facet of our reality.