Mark Rosewater wrote an article in which he revisits his Timmy, Johnny and Spike prototypes of Magic players. He makes a good point saying that these prototypes, or psychographic profiles, are just isolations of personality traits of a certain groups of players. But I think it is an error to describe these prototypes too detailed, and even to sub-divide them into several groups. The prototypes were well understood and simple. And I think simplicity is key here. Of course a prototype doesn’t is quite black-and-white, but it’s supposed to! It’s supposed to point out certain traits in a very obvious, exaggerated way. Introducing subgroups muddles this up. Now when someone is talking about “Johnny” I have to wonder whether he is talking about “Combo Player Johnny”, “Offbeat Designer Johnny” or any of the other possible Johnnys.
For example, Mark tries to explain that “Timmy” is not just a little boy that likes to play with big creatures, although that concept has worked well so far. Instead he lumps several other types into that concept. Now Timmy is someone playing for the experience. That includes people who play to socialize with others. This doesn’t seem to fit the Timmy prototype at all. (At least it doesn’t to me.)
I would suggest taking another route: Don’t be afraid to introduce new prototypes if necessary. Matt Cavotta did it with the Vorthos, and Mark could do as well. If you need a “social player” introduce him (or her), but don’t make poor Timmy suffer. But don’t go overboard: a few basic prototypes for the most important character traits are enough!
In summary: I don’t think generalizing the prototypes was a good idea.