Previous post:

Next post:

Predictions in Review: AFC West

by Chase Stuart on February 13, 2014

in Stat Recap

During the 2013 offseason, I wrote 32 articles under the RPO 2013 tag. For some of the teams, the article functioned as a team preview, but in most cases, it was just my way of researching something I found interesting about each team. I thought it would be fun — and perhaps educational — to go back and review each of those articles. Today, we’ll begin with the AFC West.

Are the Chiefs better than your average worst team?, May 16, 2013

The point of this article was to examine a Chiefs team that was inconsistent on paper. Kansas City was a bottom-2 team in the NFL in 2012 based on record, SRS, and the efficiency models of both Advanced NFL Stats and Football Outsiders. On the other hand, Kansas City sent six players to the Pro Bowl in 2012, and went 17-15 over the prior two seasons.  Here was the threshold question in the article:

So, when projecting the 2013 Chiefs, how much “bonus” credit do we give them for having a bunch of Pro Bowlers or for being a pretty good team (based solely on record) the prior two years? And what about the fact that they added Andy Reid, Alex Smith, Anthony Fasano, and Donnie Avery (and Eric Fisher, Travis Kelce, and Knile Davis) on offense and Sean Smith, Dunta Robinson, and Mike DeVito on defense?

I questioned the legitimacy of some of the value of the six Pro Bowlers, but noted that having a good record the prior two seasons may net the team an extra win or two. Here was my conclusion:

Most bad teams experience a lot of turnover, and the Chiefs are no different. As bad as they were in 2012, a new coach and a new quarterback can solve a lot of problems. According to Vegas, the Chiefs are projected to be about 2.3 points worse than average, which jives with the 6.5 win total projected for the team. Kansas City has also hired Chris Ault as a consultant, and the father of the Pistol offense could mean the Chiefs are actually fun to watch this year. I’m cautiously optimistic about the Chiefs, but they seem unlikely to resemble last year’s version in style or production. That has to be considered a very good thing.

We all know what happened: Kansas City started the season 9-0, and finished 11-4 before resting 20 starters in the season finale against San Diego. The additions of Reid and Smith turned out to be outstanding, and defensive coordinator Bob Sutton did an excellent job with the defense. But the biggest change between the Chiefs in 2012 and 2013 was in the turnover department. After finishing with 24 more giveaways than takeaway in 2012, the Chiefs were +18 in the TO department in 2013. That’s the biggest improvement in one season since 1970, and the fourth largest improvement ever:

turnover diff

What can we learn: It’s tempting to think the conclusion would be that a new coach and new quarterback can make all the difference. I’m not sure if that’s correct, though. The Chiefs faced the 2nd easiest schedule in the league in 2013, and that doesn’t even consider the randomly high number of backup quarterbacks the team faced. Smith actually averaged the same NY/A as Matt Cassel did in 2013, but the turnover improvement drove much of Kansas City’s success.

The Chargers Hired Mike McCoy — What Does That Mean?, July 8, 2013

The Chargers hiring of Mike McCoy was typical of the NFL trend: fire your head coach, hire the hot-shot offensive coordinator who happened to have coached a great quarterback. In 2012, McCoy’s Broncos ranked 2nd in points, 4th in yards, and 1st in ANY/A in 2012, which certainly made him a qualified candidate for promotion. But the real question I had was whether hiring another team’s offensive coordinator tended to improve the new team’s offense.  As it turned out, the answer was no. On average, the median rank of teams that hired offensive coordinators to be their new head coach in Points-Yards-ANY/A was 22-24-21 in the year before he was hired, and then only 21-20-23 in Year N. Here was how I ended the article:

Simple regression to the mean principles would have predicted that these offenses would show some improvement. As it turns out, most of these teams produced virtually identical numbers. Joe Philbin wasn’t able to bring any Green Bay magic to Miami. Mike Mularkey looked great with Matt Ryan and terrible with Blaine Gabbert (perhaps not so coincidentally, Dirk Koetter exhibited a similar trait). Of the 16 offensive coordinator hires from 2000 to 2012, only four were able to improve their team’s rank in both points and yards (Whisenhunt, Linehan, Turner, and Mularkey).

Those results aren’t damning on Mccoy, of course: he’s his own person, not an amalgamation of his predecessors. McCoy’s success will depend on many things outside of his control, but also on his ability to be a head coach. As part of my Super Bowl preview, I wondered if the success John Harbaugh was having might prompt teams to promote more special teams coaches. I ended that post by saying: “Being a head coach is more about managing a team, creating a vision for the roster, hiring talented people, and being a leader more than it is about play-calling.”

For McCoy, success as a head coach will mean more than just turning the offense around. It will be about finding the right coaches and players to surround his roster. But if anyone expects his presence alone to solve the problems that have haunted Philip Rivers, this post might cause you to rethink that notion.

Well, turning around the Chargers offense is exactly what happened. San Diego ranked 5th in yards and 12th in points, but in reality, may have been the second best offense in the NFL. San Diego ranked 2nd in both ANY/A and in Points per Drive, in both cases behind the Broncos. The Chargers even edged the Broncos and lead the league in yards per drive. The Chargers finished in the bottom half of the league in points, yards, and ANY/A in 2011, so it’s hard not to be amazed at the incredible turnaround in San Diego. Whether it was McCoy, Rivers, or offensive coordinator Ken Whisenhunt, the Chargers got exactly what they wanted when they hired McCoy.

What can we learn: The takeaway from the article remains the same: a great offensive coordinator is not a human talisman that will fix any offense. Two other teams hired offensive coordinators as head coaches, with differing results. Bruce Arians helped improve the Cardinals offense (with a tip of the cap to Carson Palmer), but remember that he took over Whisenhunt’s offense. What can we learn from that? Who knows. And Rob Chudzinski’s tenure in Cleveland was not successful by any definition. San Diego also hit a home run with Keenan Allen and D.J. Fluker in the draft, and at least some of the turnaround was removing the taint of Norv Turner.

Is Wes Welker the Most PPR Receiver Ever?, August 1, 2013

This was a fun exercise to determine how much more valuable Wes Welker had been in fantasy leagues that award one point per reception than in standard leagues. As it turned out, Welker was twice as valuable in PPR leagues as non-PPR leagues, which made him the most PPR-dependent fantasy wide receiver ever. That wasn’t a surprising result, of course, but quantifying what we know is as important part of football analytics.

The irony, of course, was that in 2013 Welker set a career high in touchdowns. In fact, Welker actually ranked 20th in standard scoring leagues and ranked 21st in PPR leagues.

What can we learn: Predicting even obvious things is impossible?

How much will Terrelle Pryor help Darren McFadden?, September 1, 2013

In this article, I looked at the 18 examples from 1990 to 2012 where a mobile quarterback split time with a non-mobile quarterback. As it turns out, there was scant evidence to support that theory that a running quarterback helps a running back. The 2013 Raiders provide another example to analyze, as Terrelle Pryor started 9 games, while stationary Matt McGloin started six games (with Matt Flynn starting the other game).

Pryor was outstanding on the ground in 2013, so he certainly meets the definition of a running quarterback. Oakland wound up using three running backs this season: Darren McFadden, Rashad Jennings, and Marcel Reece. In fifteen of the games, one quarterback took the vast majority of the snaps. In the Eagles game, Pryor started the game but McGloin entered in the 4th quarter and threw 15 garbage time passes. I’m going to exclude the final 8 minutes of that game, which turns it into a Pryor only game, and that means eliminating the 3 carries for 40 yards that Jennings had while playing alongside McGloin and trailing 49-13.

The table below shows the results of all three running backs in both Pryor games (which came against Indianapolis, Jacksonville, both Denver games, San Diego in week 5, Kansas City in week 6, Pittsburgh, Philadelphia, and New York) and games with either McGloin or Flynn.1

Here’s how to read the table below. In Pryor games, Jennings rushed 46 times for 192 yards, producing a 4.17 YPC average and 21.3 yards per game. In games with either McGloin or Flynn, Jennings rushed 114 times for 501 yards and averaged 4.39 yards per carry and 71.6 yards per game. That means he averaged 0.22 fewer yards per carry with Pryor around, and 50 fewer yards per game.

Runner
Rsh (TP)
Yards (TP)
YPC (TP)
YPG (TP)
Rsh (MM)
Yards (MM)
YPC (MM)
YPG (MM)
YPC Diff
YPG Diff
Jennings461924.1721.31145014.3971.6-0.22-50.2
McFadden953173.3435.219623.268.90.0726.4
Reece12574.756.3341614.74230.01-16.7

McFadden averaged more yards per game with Pryor around, but that was a function of his starting role. This is just one extra data point, but at least in 2013, the Raiders running backs weren’t more effective with Pryor as quarterback.

What can we learn?: It’s tempting to believe a running quarterback will help out the team’s running back. And I’m not sure the data tells us that’s not the case. The data just shows that the evidence doesn’t prove that running quarterbacks do help out running backs. In any event, it seems pretty clear that mobile quarterbacks don’t have a big impact on the fortunes of their team’s running backs. There will always be isolated examples, but if there are as many examples on the other side, then it’s not a real trend.

  1. Based on Football Outsiders rush defense rankings, the opponents in Pryor games had an average rank of 16.2 compared to 18.6 in the non-Pryor games, which I think is small enough to ignore. []