Antar said: Keeping in mind that we'll be talking Glicko rating and not Elo, the cutoff will likely be 1630. But I will need to verify some things and might end up going 2 std.devs to 1760.





Correct.



UU beta changes every month until it becomes official, after which it's back to once every three. Click to expand...

Well, certainly 1 stdv isn't "1337", but it seems to be a good benchmark of a battler who at least has some competence, although one does have to filter out troll alts or people who handicap themselves in other ways (like to face easier opponents to test a team) whose presence would increase the relatively rankings of those who play seriously. That seems to be the only source of inflation for ratings, since the normal distribution is based on one's relative ranking (unless it is something that was standardized some time ago, then fixed, and then the abilities of the new sample population changed relative to the standardized population, much like the Flynn Effect of IQ testing, or the decrease in SAT scores, mostly on the Verbal section, in the 70-80s that caused the need for recentering in 1995.)Also, is there a metric to show how some Pokemon's moveset is unpredictable. Something like information entropy? A concrete example is that Rotom-W's moveset is very predictable: Hydro Pump, Volt Switch, Will-O-Wisp, and either Trick, Rest, or Pain Split causing the most "entropy" for Rotom-W. (Mega) Lucario and Gengar are much more unpredictable. However, Scald/Surf/Hydro Pump on Greninja should contribute little to entropy because they are mostly used as a coverage move, and whatever one chooses is based on personal preference for power over reliability over potential utility.