After examining Star Fleet historical files, Kirk realizes that a particular kind of attack on the Romulans that he is considering is very difficult, succeeding only 5% of the time. Because of the challenge, he asks Spock to build a simulation to predict whether an attack successfully achieves its objectives or not. Spock runs the simulation many times under different random historical conditions, which are then validated against historical real-world data. Spock's simulation is fairly accurate -- in only 3% of the cases does it indicate failure when there was an historical success and in only 4% of cases does it indicate success when there was an historical failure. Kirk plugs in the current situation he is considering into Spock's simulation and runs the simulation which predicts success against the Romulan scourge. But he can't cheat on this one, and so he is unwilling to commit the Enterprise to the battle unless the probability of an actual success is at least 70%. Should the Enterprise attack? Why or why not? I've seen answers all over the place for this questions from 0.55 (55%) to .65 to .77 to .92, and as low as .05. I know im supposed to bayes' theorem