Suppose someone suddenly gets sick, pulls off the road, and witnesses a three-car pile-up that he would have been involved in otherwise.  Would he betray confirmation bias if he concluded that his guardian angel had assisted him?


It is not unusual to pull off the road because of sudden illness, it is not unusual for three-car collisions to take place, and there would be nothing at all surprising in both occurrences happening together.  So such an incident would not constitute proof of angelic intervention.

On the other hand, if one already has good reason to believe in guardian angels on other grounds, there is no fallacy in believing that this may have been an instance of angelic intervention.

So if you were thinking you had proof of angelic intervention, think more clearly.  But if you thought you couldn’t believe in it, you can stop worrying.