I think a lot of Michael Byrd's work. He's a great Twitter follow and as rational as one can be as a Michigan Man. However his piece on the Mark Richt total-debacle game is questionable.
His evidence, using the spread:
Game | Spread | Georgia's losing margin | Delta |
2015 Alabama | -2 | 28 | 30 |
2014 Georgia Tech | -10.5 | 6 | 16.5 |
2014 Florida | -10. |