In signal detection terms, the cost of missing a real conspiracy is higher than that of false-alarming to a suspected one. This calls for setting our bias toward the trigger-happy rather than the gun-shy end of the scale, adapting us to try to get wind of possible conspiracies even on tenuous evidence.
