Interesting piece — as always. But given some other articles I’ve read — I believe by you, didn’t you just evaluate a set of 150 outliers? All of these articles are far outside the typical range of recommends for an article. So does the construction and make-up of these articles really drive popularity (or whatever you are using to define Top) or it is more likely something external to the data? Either way — I applaud the effort — but needed to ask…