Top 500 vs Benford's law

[sawyl]

Dubious about some of the latest Top 500 results, I spent a few idle minutes knocking up a quick bit of code to parse the XML data and compare the distribution of results against those predicted by Benford's law. Sadly the span of the data in the latest list was insufficient to draw any conclusions, but when I added in the results from previous years, the curves fitted pretty well. So, who knows, maybe the results are trustworthy after all...