But it is in your advantage! Code a simulator and do some numerical experiments. The simulations where you switch end up winning more.
Edit: from your other reply I see your beef is more with the wording.
Edit edit: although I disagree with your point even then. But it seems to lead to rather fruitless argument, so let's leave it here.
I think reasoning with the 100-door variation is likely to be more enlightening than encoding your assumptions into a simulation.