I never know how much credence to give the
statistics compiled on buttonmen.com -- especially those created by the "adoptions" that encouraged players to generate data on every possible matchup -- but how good is that data?
Even beforehand, there are lots of buttons that I think are "better" than their statistics indicate -- I suppose there must also be buttons that are worse than their statistics, but those don't jump out at me.
It's most notable with buttons with atrocious records and unusual recipes -- Max, Tess, Socrates, maybe Synthia. I'm not going to say any of these are GOOD in some absolute sense (well, maybe Max is good, at least against many opponents) -- just that they are better than their stats.
I used to wonder if one possible source of misleadingly low statistics is that buttons play more often against others in the same set, and the runt of the litter on Krypton might still be Superman on earth. [am I using that analogy correctly? I don't really know the Superman mythos.] I don't actually have any
examples of this, though!
I still think the real isue is that some buttons are just subtle to play --and if you suddenly get one of them in a randomly chosen matchup, you're not likely to make optimal decisions.
There were
several examples listed with Max (also
this) in this blog a few months ago; maybe the thing to do next is to pick other strange buttons -- especially buttons with misleadingly low statistics -- and try to see what's going on with them, maybe through examples. I'd be interested to hear if anyone else has can suggest particular buttons they think are better than their stats.