Read this, which was linked from Kotaku today. The thrust of the article is that a game isn't a series of interesting decisions, and games like Guitar Hero prove that some wave of new games moves away from this definition into something new.
I mean, I don't even know why I should have to explain this, but maybe it's difficult to understand if you're not a working designer. Rock Band/Guitar Hero/Beatmania, etc., all rely on a combination of pattern memorization, quick reflexes, and a good sense of timing - no decisions there - but the good rhythm games, and when you get to harder levels where the "expert play" comes in, you have to be constantly making decisions about fingering, which hand to use at a particular time, and when to use your Star Power/Overdrive to maximize your score or save your friends.
These are all decisions - the fact that you're functionally repeating a linear pattern of notes doesn't make those decisions go away - they're the "stuff" in between the decision points that Soren Johnson was referring to in his quote in the article.
The author of the article states that this may be an appropriate definition for a strategy game, but isn't appropriate for other genres. I disagree. FPS's are almost entirely about ammo management, proper weapon selection, and use of the environment - all those are decision points supplanted by gameplay that requires ridiculous reflexes and hand-eye coordination.
Brain Training is a serioes of short, controlled decision points (what does 2+1 = ?), Guitar Hero is described above, and the Sims *is* a straightforward strategy game, with a suburbanite layer, full stop.
I could see there being an argument that the granularity of the decision points makes them into "gameplay" rather than "decision making," but I think that's improper - the speed and magnitude of the decisions you're making doesn't transform that process into something different.