Football Outsiders managing editor Bill Barnwell isn't the type of guy to take a comment at face value.
While he was referring specifically to the Bills' failed attempts to replace him with Rob Johnson, J.P. Losman and Trent Edwards, Kelly was promulgating a common belief that warm-weather quarterbacks can't perform in the cold.
A couple weeks ago, Edwards called Kelly "a little bit naïve."
Barnwell decided to find out whether that's true. His research shows Edwards was right.
In an educational column for ESPN Insider, Barnwell runs game data from games back to 1993 and finds QBs who grew up in warm-weather states fare better in the cold than quarterbacks raised in chillier environs.
Barnwell defined the quarterbacks geographically by where they played in high school, but the home state had to be clear-cut one way or the other. Barnwell then sorted their stats in outdoor games where the kickoff temperature was no higher than 35 degrees.
The numbers then were broken out in three charts that pitted cold-weather QBs versus warm-weather QBs: in all cold games (cold-weather QBs had minor advantage), in cold road games (warm-weather QBs had noted advantage) and in road games that didn't qualify as cold (warm-weather QBs had an advantage, but not as much as they did in cold games).
The third chart, used for control, was the clincher for Barnwell.
"The warm-weather quarterbacks are better than the cold-weather quarterbacks specifically in cold-weather games," Barnwell writes. "Not only is Kelly's theory incorrect, the truth is the exact opposite of what Kelly suggested."