Kevin Kelley made waves with his “never punt” philosophy. Now he’s got another one: explosive plays!:

Well, the Washington Post recently ran a story on Kelley’s next innovative idea that is going to buck conventional coaching, and give defensive coordinators nightmares. After using an ESPN database to study college football history, Kelley found a new trend emerge last season where teams that recorded more explosive plays of 20 yards or more won 81% of games.

Kelley also found that on typical plays where two players touched the ball (QB and RB or WR), those 20 yard plays came at about a 10% clip, but when at least three players touched the ball (on a lateral or trick play of some sort), the percentage for an explosive play almost doubled to about 20%.

This could be huge! But the statistics don’t necessarily tell us much. At best, the cause/effect is dubious.

Southern Tech had a season in the recent past where it struggled mightily. Someone did some number-crunching and decided that what the Packers really needed to do was run the ball more. In games where the offense ran the ball 60% or more of the time, they won! Less running, they lost. QED!

Except that the causality was reversed. The higher run-rate was indicative of trying to kill the clock after we’ve taken a lead. The high passing rate was the product of a team struggling to catch up against a running clock.

It wouldn’t surprise me in the least if the same were true of explosive plays. Teams with explosive plays win precisely because they’re the team that can effectively run explosive plays against the other team. Explosive plays are difficult and are more likely to require an asymmetry of talent.

Not unlike “Half Time Stats.”

When you’re watching a game, the announcers will often say something like “When Southern Tech is up by two touchdowns or more, they win 84% of the time…” which sounds impressive. It means Southern Tech can hold leads. Yay if you’re Sotech! Except that a fourteen point lead at half-time is more often than not going to be indicative of a talent asymmetry or just better play. Chances are a team ahead by that much at half time won’t just win because they can spot 14 points, but also because the way that the game has been going indicates that they are the better team. They’ll probably win the second half, too, for that reason alone.