> How many of the 3000[0] "terrorists" that Obama ordered blown up (accruing about 400 civilian casualties in the process) had actual plans and setups in place to make the trip to America and kill someone?

This is a relevant question, but not precise enough. What I would be interested instead, is, what's the total mean and variance of american lifes saved in all those operations? And more importantly, what other hypothetical possible combinations of these variables (civilians killed, combatants killed, unuidentified casualties, mean and variance of americans saved) we could have if different decisions have been made - i.e. what's the whole problem space?

Of course, this data is impossible to get and hard to estimate, but this is much closer to the real dilemma at hand.