I should point out that the factual error in his review of Labyrinth wasn't the only thing that bothered me, I think he got some basic assumptions about the film wrong. He thinks that because anything could happen within the Labyrinth, nothing has any meaning.

The point of the Labyrinth is that it's supposed to be unsolvable, and the random nature of the film represents that. Sarah doesn't win by solving puzzles correctly, because this is obviously impossible - she wins by making friends within the Labyrinth, and realising that the Labyrinth has no real power over her anyway. The Labyrinth doesn't follow the rules set out by previous films, but that's the point.

It's like saying that one game where you can take back your moves is invalid because it doesn't follow the rules of chess, where you can't take back your moves. You can't criticise something for not following the precedent set by something completely and utterly different.

Anyway, the whole review put me off Ebert as a critic, and so does this article about computer games.