Evaluating QBs, Beamer's legacy and Ricky Dobbs 

November, 20, 2009

Let's jump into the mailbag.

From John in New Orleans: Just wanted to throw something out that's been bugging me: Passer Rating/Pass Efficiency. Do you think people put too much stock in that stat? Or use the stat as an overall "passer rating" too often? It seems to me that the formula puts way too much weight on yards-per-completion and TDs-per-completion (and probably not enough weight on interceptions). Have you ever looked at the formula, or more specifically, ratings for individual games?

It doesn't seem like a good measuring stick. I'm looking at two quarterbacks right now. One quarterback has 2,543 yards, 21 TDs and 11 INTs and another has 3,815 yards, 28 TDs and 5 INTS, but the first is ahead 167.6-158.9 in rating. Anyway, not trying to knock any individual QBs by any means, just curious if you think people value the stat too much.

This is a ripe topic. John also submitted an interesting article by Rob Daniels that supports his point well:

David Cutcliffe called it the finest quarterback performance he had ever seen. Why not? Duke's Thaddeus Lewis did go 40-for-50 for 459 yards, five touchdowns and no interceptions at N.C. State. (October 10.) What's not to like about that? Surely the achievement would be heralded when the NCAA calculated passing efficiency and released national leaders the following day. Oddly enough, Lewis' day was only good enough for the 86th-best performance in passing efficiency. This year. According to the NCAA's formula, a passer would rather be Michigan State's Keith Nichol against Wisconsin on Sept. 26 than Lewis at Duke on Oct. 10. Nichol went 7-for-12 with two interceptions and two touchdowns but 195 sacred yards and wound up with a rating of 216.5 -- some 26 points higher than Lewis' 190.11.

If you were the coach and you could choose one set of passing statistics for your team on a particular day, would you take 7-for-12 with two interceptions and 195 yards? Or would you prefer 40-for-50 for 459 yards, five scores and no picks?

Take a moment to ponder that one. OK. Time's up. Paul Johnson might prefer the first option. Virtually everyone else would side with Lewis.

I think this is why stats can be pretty misleading and you can't read too much into them. Some of it's sensible. Some of it you have to take with a grain of salt. I use the NCAA stats a lot in my picks analysis to try to back up some ideas. Thing is, all stats are flawed to a degree.

In my mind, the best barometer for a quarterback -- in addition to the team's win-loss record -- is TD-INT ratio because it shows decision-making and accuracy, although some interceptions can definitely reflect receivers' running the wrong routes. I talked to Tennessee offensive coordinator Jim Chaney, Drew Brees' old coach at Purdue, about the merits of the passing efficiency stat. He buys the stat, saying he thinks it has a lot of validity, and that the pass efficiency defense formula is also relevant because that takes into account passing yards per attempt and reflects a combination of accuracy and the ability to push the ball down the field.

"I've never had any luck converting third downs if I can't complete a pass," Chaney said. "TD-INT ratio is great, and I want to see for a young quarterback at least a 2-1 because I think that shows more than just decision-making. It also shows his timing. It's not just about is a guy going to the right guy, but also, is he doing it on time? Is he late? How quick is he when he makes that decision?"