Warren Spector doesn't update his blog often, which is why I was surprised to discover he had actually written about Hideo Kojima's GDC 2009 keynote a while back. He says:

In describing his creative process, Kojima talked about identifying a problem (e.g., Get a Character Over That Wall) and then coming up with a bunch of ways the problem could be solved. Eventually, he settles on the coolest solution and executes that solution. I was dumbstruck that he goes to the trouble of thinking up all those answers but then limits the player to only one. In other words, the concept of choice belongs to developers, in Kojima's world, not to players!

I was at the keynote as well, and this is a wild misrepresentation of what Kojima said. The metaphor of "getting over walls", which Kojima used as a visual aid to his talk, was an illustration of his development process, not his game design philosophy. The talk was strictly about how he and his team approach production challenges. Kojima didn't even mention his personal theories of player agency, let alone explain them.

Spector's willingness to misread Kojima this way concerns me, because it is indicative of the way Kojima is often misread by Western game designers. It makes me wonder whether the people who pick on him for his supposed crimes against interactivity have ever spent a decent amount of time with his games. Spector goes on to say:

My thinking is, if you're only going to offer players one way to solve a problem, well, for starters, maybe you really want to make a movie... But, if you're going to go to the trouble of thinking up a bunch of ways to "get over the wall," as he put it, why not attach some consequences to different wall-climbing approaches and let players in on the fun?

Why not indeed? Kojima must have asked himself the same question, since there are about about a dozen ways to tackle any given problem in Metal Gear, with Snake Eater and Portable Ops offering the player especially rich possibilities. These two games are on par with the dizzying emergent complexity found in Thief and Hitman, which puts them among the best stealth games ever made in my opinion.

I am seriously beginning to think that very few of Kojima's critics have actually played his games in any significant capacity. (And by "significant capacity" I don't meaning having played MGS1 11 years ago when everyone else did. I meaning having played and finished at least a handful of the other dozen or so games he's made over the course of his career.) Kojima's got lots of problems, but choice-driven emergent dynamics isn't one of them. If criticism of Kojima's work were a little more informed we might be having useful discussions about his virtues and vices as a game designer instead of taking cheap shots.