Original Comment by: Charmaka
An interesting thing to consider here is the Uber-licensed. It's already been mentioned, and yes, it's Star Wars. I can't off the top of my head think of any other license which has anything like as many titles associated with it. Some are great, some are abysmal.
The interesting this is what happens when you take a scatter graph where "how good is it?" is one axis and "how closely is it tied to one of the movies" is on the other. The result? Very good correlation. There are a few oddities, like Lego Star Wars (great, yet entirely movie-based), but by and large, the best games associated with the Star Wars license have nothing to do with the movies the license revolves around. KotOR, the Jedi Knight series, Battlegrounds (seems to be doing reasonably well), EaW (ditto).
I agree that trying to match release dates certainly doesn't help, but I think there's another straightforward problem which is that, shockingly enough, games aren't movies. Games that try to be movies are doomed to failure, as are movies that try to be games.
Conclusions? In the first instance, the way forward is to simply allow developers on movie tie-ins to do their own thing in a way which ties in with the license to a satisfactory degree. This will at least allow decent games to be made. However, there's a better option. Take a movie that wants a game to go with it. Make the movie tell a huge, over-arching story but reference say a key battle or past event or prominent individual whose story is never actually told in the film. Then tell that story in the game, with constant references to the film story but without ever properly explaining it. If both are good, then you get true cross-fertilisation - movie-goers will need to play the game to understand that particular detail, and gamers will have to watch the movie to see where it fits in to the overall story. Everybody wins, and game developers are actually allowed to tell a story in a way which is suitable to the medium. And there was much rejoicing.