[pycal]

Just a guess, but wouldn't this approach used in isolation result in a high potential for getting stuck at local optima? It seems to me that the success of this experiment is the result of interaction with players in different MMR brackets.

[Bartweiss]

Yes, it can. It's fairly common to keep a pool of learners which aren't identical for exactly this reason - either playing against prior snapshots, forking training, or lightly randomizing some components of the learner.