Does DVOA weight second half performance more than first? How about when teams winning big go into prevent defenses? The reason I ask is that, of the Steelers' games I've seen, they seem pretty consistent in their play; but the Patriots seem wildly all over the place. For example, the Patriots played poor defense in the first halves of the Cowboys and Colts games, but really clamped down in the second half. Does DVOA adjust for half-time adjustments?

Something tells me even if Pitt is having success in the first half, they may get shut down in the second (or, it could certainly go the other way if the Patriots wear down).

"For example, the Patriots played poor defense in the first halves of the Cowboys and Colts games"

So the patriots played poor defense against two of the best offenses in the league....

How much does sample size impact the stat about the Falcons being the only team in the NFL not to turn the ball over in the red zone? I'm assuming there's a direct correlation between scoring the second fewest pts in the league and number of trips to the red zone.

"The Steelers have the best run defense in the NFL but are only average against the pass."

This doesn't seem to mesh with the DVOA page. [I did find it amusing that the Madden EA simulation apparently has an irrational love of Feeley, giving him 2 TD's, zero interceptions. The only thing missing from the "write-up" was a call to trade McNabb.]

4: The article refers only to the Steelers red zone pass defense, not overall pass defense (which is ranked 1st overall).