Can you elaborate more? I would have thought your example isn't comparable, in that we death is inevitable, but being robbed isn't (as much, depending on circumstances)
Imagine you have a giant 10000-sided dice. You throw that dice once every day searching for a magic number (e.g. 1337). "Hitting 1337" is an experiment with a binary outcome (yes/no).
On day 500 you hit the magic number. Did you have a 1 in 500 chance of hitting it? No, you had 1/10000 chance of hitting! Even if you throw it 10000 times, there's no 100% chance of hitting 1337 since the dice is still 10000-sided even after you hit any number (this is called "no replacement") so you can have, for example, 1338 coming twice and 1337 none. There is never a 100% probability (but it approaches 1 rapidly near 10000).
On the other hand, you can calculate the probability of "hitting 1337 at least once in N throws", which is actually the CDF of a binomial distribution[1], but you need the initial probability of a single event!
Bringing back the robbery theme, living a day of your life is just repeating the "being robbed today" experiment (throwing the dice) once a day. Being robbed on day N of your life just means you repeated the experiment N days and N-1 times the outcome was "no" and then a single "yes". This does not mean that the CDF was 1 at N attempts, it just means that it was greater than 0... and this is just the probability of "being robbed after N days", i.e., the CDF of "being robbed today", not the probability of being robbed itself.
Also: you can't evaluate probability a posteriori unless the events are repeatable under controlled conditions, in which case you repeat the experiment lots of times and derive the probability from the outcomes. Burglaries are not repeatable under controlled conditions!
