# Estimating Crime

Let's say I have a dataset containing crime statistics for a given city block, how would I go about estimating the probability of a crime occurring to me while I am on that block?

-
I don't think you can. You can use the statistics to estimate the probability of a crime occurring while you are on the block (look up Poisson Distribution). But estimating the probability that a crime will occur to you (I assume you mean that you are the victim of the crime; not that you will think of a crime) would require you also to know the number of potential victims in the block at the time. – user22805 Jul 18 '12 at 8:00
Construct some hypothesis and test the probability of the event . May be these things can be of some use for your hypothesis proving – Iyengar Jul 18 '12 at 8:43
Also, there appears to be a hidden assumption that the probability of being a victim of a crime is spread uniformly across all humans in the area, which is so far from any reasonable reality as to render the whole exercise useless. If most of the crimes in the area are house burglaries, then just visiting someone who lives there carries no risk of becoming a victim of that. Similarly if it's shoplifting and you're just living there (rather than running a business), or domestic assault and you're just walking through. Or, for that mattter, if it's all tax fraud and you're not the government. – Henning Makholm Jul 18 '12 at 11:34