> To ensure that I had the most useful and well-reported data, I limited my research to “successful” startups.

So, in other words, the author didn't consider the failed % of startups based on founder? As in, at all, if the article is anything to go by?