Hacker News new | ask | show | jobs
by rpicard 4988 days ago
Great point! Dividing by expected would have been a much nicer solution. I think the method I used still works though, just not as nicely. Am I wrong?
1 comments

Z-Score is your friend for this kind of data normalization.

http://en.wikipedia.org/wiki/Standard_score

Thanks. It looks like I need to do a little self-study in statistics.