This is a very interesting article, however, IMHO it has several major problems:<p>1. The author doesn't understand, or is too lazy to recognize, that "correlation does not imply causation". It's hard to read sentences like <i>it seems like the length of the title has a minor influence on the quality of the question</i> and not shudder.<p>2. There's no interpretation. At all. It's just queries, and doesn't provide any real insight into how to write better questions, get better answers, or get more upvotes. No hypothesis or model.<p>3. It's missing meaningful statistical measures, such as standard deviations, error bars, etc. For example, is this even statistically significant: <i>The average title length of the top questions was about 5% shorter than that of standard titles (47 characters vs. 50)</i> ?<p>4. Biased samples.<p>So my takeaway is that it was fun and interesting to read, but not very meaningful.