It means you tried to find the cause that makes the effect most likely, rather than the most likely cause given the effect. The two are only equivalent if what

It means you tried to find the cause that makes the effect most likely, rather than the most likely cause given the effect. The two are only equivalent if what you know a priori is that all causes are equally likely. In this case they aren't, so a MAP estimate would still give a very low probability to an actual collision, whereas ML doesn't care about the probability of a collision itself -- merely the probability of what it causes. :)