And this is why the word "objective" should only be used by people who aren't blithering idiots.

Seriously, statistics are only as useful as the foundational premises, and even a cursory glance should make obvious that this list is useless. It ignores the extremely important issues of replication and consistency. Let me assure you -- these are not just issues that could make the list even more accurate. They're issues that, in removing them, actually make the list next to worthless. This contrast alone should have given you pause. Add to that the fact that the "Pythagorean wins" simultaneously rewards lowered expectations while punishing higher ones, especially when games are governed by football scoring....

Or, to put it another way -- I wonder where Chizik would have ranked in February 2010? Probably approximately where Malzahn ranks on this list. The almost-incidental mention of the likely course of Malzahn's numbers over time stands in stark contrast with his attention-grabbing and inane leaps to broad conclusions in the rest of the post.

Sorry, dude. I'm not attacking you for reposting this, or for not analyzing it closely before doing so. But common sense alone should have sent up alarm bells after reading the Reddit post (which is chock-full of subjective and unsustainable observations even over and above the numbers.) The mistake many people make is being dazzled by counter-intuitive results without grasping that being counter-intuitive isn't, in itself, validation. Indeed, it's a reason to invite much harsher scrutiny. In this case, the poster pretty much took an inapplicable algorithm and applied it improperly while deriving unsustainable conclusions from insufficient data points.