Well there the cost plays a part! It’s not independent. If I can build 1000 missiles for every 1 interceptor, the probability of interception hardly matters.

This actually was why we planned to put lasers in space: the economics of one nuclear-pumped laser reflected through Unobtanium were better than any other interceptor. And even that if the effect worked (it didn’t, they could not prove lasing and fired an engineer who blew the whistle on that), the system could be defeated by a staggered salvo.