Say strategies X and Y are roughly equal in power, so there's no strong reason to switch from one to the other. Now, say strategy X is popular in one region, and strategy Y is popular in another. Perhaps because the strategies were devised by players in those respective regions, and spread to being popular... perhaps by chance... regardless, this can be self-sustaining, as those will be the strategies new players will be exposed to, and will then learn. Now, in the region where X is prevalent, strategies that are strong against X but weak against Y will be good, but vice-versa will be very weak. In the other region, the opposite applies.
And really, this is an oversimplification, as the relative powers of X and Y also depend on what the opponent is playing - if the opponent is playing the strong-against-X strategy, then choosing strategy X may not be the best idea. Think of it like evolution - there's a lot of random chance and irrelevant variables, but strategies will be selected that are strong in the environment - but the environment is itself made of other strategies, which are themselves evolving.
To keep the evolution analogy going a little further - if you restarted the world at, let's say, the extinction of the dinosaurs, have the random mutations occur differently, but otherwise leave evolution to run its course, and fast forward 65 million years... chances are you'll find a different world, with its selections of stronger and weaker creatures, all interacting in their usual natural ways... but there won't be humans there. And indeed if you were to transplant some humans there, chances are they wouldn't survive nearly as well as they do in our current world. And vice-versa - the dominant life form of alta-Earth likely wouldn't do as well in our own world. Now, strategies aren't nearly as environment-dependent as lifeforms, but they still are to an extent.
Or, tl;dr: the metagame.
While no one overhear you quickly tell me not cow cow.
but how about watch phone?