The Stats Room: Week 5 QB Projections

First, here are last week's results, which included in-season projected pass attempts.

As usual, I over projected a few guys (Brian Hoyer, Case Keenum and Josh McCown), but the results remain encouraging. With my projections setting a baseline, I've decided to test/add two more possible inputs, the team's projected points score and if the opponents defensive pass-run mix.

As for the home team's expected points ((OU-Spread)/2), I've wanted to use it as a proxy for opposing defense, weather and playing surface. Until now, I haven't found an effective and simply add it.

I do now. For each point a team scores over the historic league average (20.8 points), the quarterback gets an additional 0.5 points to his projection. If his team is expected to score less than the average, his score takes a hit.

It was a small additional improvement but really helped to adjust for potential blowouts.

The other addition I tested was the passing-running split the defensive team allowed. While the spread and O/U would be a good proxy for the overall defense, I was wondering if the defense's nature, good against the pass versus the run, would make a difference. Basically, if a team the quarterback was up against had porous secondary but stout run defense, would his performance be better than expected?

The short answer is no, the pass-rush ratio doesn't matter one bit. Projecting any defensive metric beyond the spread is not possible. The pass-run split barely correlates from the season's first half to second half. And with so much turnover from one season to the next in coaches and personnel, very little intra-season correlation exists.

Well, I'm finally out of ideas for quarterbacks. Please let me know if you have any ideas but I'm ready to move on. Now for quarterback ranks for Week 5.

Name Zman RotoWire Diff

•Carson Palmer (10.9 pts diff): This comes down to a difference in pass attempts (29 vs. 38) and the yards and touchdowns with them. My system, compared to the RotoWire projections, is weighing the 2016 season more than the current one at this point.

•Russell Wilson (7.9 pts diff): The difference is for two reasons. First, the Vegas line is all over the place for this game, when there is one, so Wilson might be getting a bump from it. Second, my projections just like Wilson a little more across the board with both rushing and passing.

•Alex Smith (6.9 pts diff): My projections like him for two reasons. First, the standard projections have him running only once for eight yards while I have him taking off four times and getting 20 yards. Besides the rushing yards, I have him for few more yards passing and the increase in touchdowns because Vegas has it as a close (-1), reasonably high-scoring game.

I'm going to take a break from projection articles the next few weeks and examine the idea of red-zone running backs and wide receivers. As always, let me know if you have any questions.