This reasoning has flaws as well. We do not automatically assume that all events start in the "off" position and are switched on as soon as one thing says so.
There are no flaws, and none of the contions are assumed to be anything. Its a precise explaination of logic.
Hes showing that if something only happens when something else does, it doesn't mean you can assume that it ALWAYS happens when something else does.
Thats why in close combat models only make close combat attacks, but close combat attacks aren't always made in close combat.
For example, I bet you would be pretty mad if your friend would tell you that a game did not get called off, so you go to the field and find it covered with snow and see a sign saying that the game was canceled due to snow. The game was not called off due to rain. Stating that the game was not called off IS a real statement with meaning that is not simply beaten by a statement to the contrary. Why should a positive statement take precedence over a negative one, anyways?
But thats that point, in the example presenting you don't ask if the game is off or not. Its demonstrating that asking if its raining or not isn't proof that the game is on or not. The whole point hes making is that you can't say, "Is it raining? No, then game on!" because it might be snowing.