The Optimal Blog Post Length to Maximize Viewership

I don't understand how any of this is useful.

- he samples only his own writing

- he has a far bigger sample of "medium-length" articles (how big are his samples of 100 word posts? 2000 word posts?)

- the title seems intentionally misleading, making it seem like it was actually a scientific analysis, and that a positive result was actually found

- content isn't controlled for. How are retweets a useful metric for measuring optimum length? Wouldn't that have far more to do with content?

- doesn't take into account the presentation of information

- doesn't take into account his audience

- etc, etc.