Yeah, the OP has serious blinders here. Like the student who threw alarmist BS about how the data could be used to track visits to stripper clubs...this seems to be an example of cherry-picking the data. So, I'm not good at discerning deviation from thousands of small dots...what's the numerical measure of the correlation here? What are the significance of those blue lines when they show a trend over empty black space?

And what does the analysis/trendlines look like on drivers who can be discerned as not being Muslim?

I'm all for personal data protection, and the TLC erred in not anonymizing the data. But this weak analysis and alarmist bullshit by a PhD serves little but to give governments more excuse to hide public data.

edit: Linking to someone more well versed than me, regarding the cherry-picking of data in the "Riding with the Stars" article http://blogs.law.harvard.edu/infolaw/2014/11/21/the-antidote...