Why is it hyperbolic?

On a per game basis Allen and Hurts are basically tied in TDs and TOs and Hurts’ is the literal favorite.

That’s all the article is pointing out.

Well yes, but also no.

A QB can play good (or at least produce a positive EPA) and their team can still lose.

So while losing more games will make a QBs sample size larger, that doesn’t automatically mean he will be lower on this chart.

Obviously these are extreme examples, but NE only has two more losses than LAC.