Given two pitchers, one of whom is great and the other of whom is not great, obviously the great pitcher will have the better odds of throwing a perfect game. That is so obvious that I can't believe I'm even typing this out.
Here's the thing, though: baseball's got a lot more not-great pitchers than it does great pitchers. So while you can try to predict who's going to throw the next perfect game, most of the time you're going to be wrong. Nobody would've predicted Philip Humber, even against the Mariners. Nobody would've predicted Dallas Braden. Plenty of people would've predicted Roy Halladay, but nobody, not nobody would've predicted Armando Galarraga, and I do count that one. Hell, Galarraga generated one more out than all the other guys.
If you were to sit and try to predict the next perfect game a few days ago, though, Matt Cain would've or should've been near the top of the list. Especially if he were starting at home. What's a more obvious measure of perfect-game likelihood than on-base percentage against? By that very simple measure, dating back to 2009, Matt Cain has allowed a .265 OBP when pitching in San Francisco. That's the fourth-lowest mark in all of baseball, behind Jered Weaver, Ted Lilly, and Justin Verlander. Weaver already threw a no-hitter at home this season. I'm going to go ahead and conveniently ignore Lilly. Verlander flirts with a no-hitter or perfection just about every time he takes the mound.
There's something about Matt Cain that allows him to surrender fewer baserunners than you'd expect, particularly within AT&T Park. That quality gave him some of the highest perfect-game odds in the league. They still weren't good odds, but they were higher odds than most. Wednesday, Matt Cain threw a perfect game within AT&T Park.
Sometimes these things come completely out of nowhere. Sometimes they do not.