Digits and voting
I thought this analysis of the numbers in the Iranian "election" was interesting so I wanted to do it on my own, except that I used the data from the county level instead of the province level (mainly because that was the first set of data I could find). With this set of data, you do not get the same oddities in the ones and tens digits that they got in the province level data. Plus, there are a lot more data points, so any systemic tampering should stand out even more. My graphs are below. You'll notice that each digit gets about 10% of the vote or what you would expect.
Update: Ah... they mention the difference between province and county analysis here.

