Analyzing NFL coaching changes: does experience really help?

A few weeks back, I wrote a post about how Madden players might just turn out to be decent NFL coaches (and also the surprising lack of time today's coaches spend in game situations). Seeing as it's unlikely that NFL teams are going to be hiring gamers any time soon, I thought I'd take a look at how those teams make their decisions on who their head coaches should be, and how backgrounds influence NFL coaches' success. This hopefully will provide some insight into what kind of head coaching candidates teams should consider and how this year's new hires (Chan Gailey in Buffalo, Pete Carroll in Seattle and Mike Shanahan in Washington) might do.

Now, coaching success can't be considered in a vacuum; the personnel coaches inherit have a lot to do with their success. Thus, I looked at how coaches do both in immediate turnarounds (so, the next season) and over the long term (their complete records), and I also looked at their backgrounds. I selected the last five years (2005-2009) as a sample that would be reasonably large and readers would be familiar with, but this analysis could be easily done for any time period. Data is from Pro Football Reference, and is only looking at the regular season. Here's the complete spreadsheet; I'll break down the highlights of it further down.



First, let's consider the coaches with the top 10 first season turnaround percentages (calculated by taking their winning percentage from their first season in charge and subtracting the winning percentage of their predecessor in his final season). Here's that list.



One of the interesting things about that list is that only one of the new coaches (Wade Phillips) had previous head coaching experience. Three coaches had been defensive coordinators, three had been offensive coordinators and three were NFL assistants, but not coordinators. This would seem to fly in the face of the idea that it's best to hire someone who's been there and done that as an NFL head coach before. Let's see if that holds up over time by looking at the top 10 new coaches by overall winning percentage and how they compared to their predecessors.



As we can see, there are several holdovers from the previous sheet, but there are some new faces. What's more interesting to me is that the trend I discussed earlier has continued, though; only two of the new coaches have previous head coaching experience (Wade Phillips and Norv Turner), and those two are probably the most criticized members of this group for their lack of postseason success despite tremendously talented teams. Of the other eight coaches, two were NFL offensive coordinators and two were NFL defensive coordinators. The remaining four were NFL assistants. One other interesting note from this list is that Turner actually had a negative percentage change in his first season (-.188), as he took over Marty Schottenheimer's 14-2 Chargers team and led them to a measly 11-5 record. I'd suggest that Turner actually is quite a talented coach, despite public perception, but the fact remains that he had a tremendous amount of starting talent to work with.

Another element to examine is the top 10 coaches from an overall change perspective (how their overall record compared to their predecessor's overall record). Here's that list.



This list features some of the same names we've seen on the previous lists, and some new ones. One thing that interests me about it is that it includes two guys typically panned as horrible NFL coaches; Lane Kiffin and Romeo Crennel. Yes, neither was particularly good, as they had overall winning percentages of .250 and .375 respectively, but they were significant improvements over their predecessors (Art Shell and Terry Robiskie). However, depending on how you consider coaches, Crennel could be omitted; Robiskie was only an interim coach who guided the 2004 Browns for the five games of the season, so you can argue that Crennel's real predecessor is Butch Davis, who actually had a superior career winning percentage in Cleveland to Crennel (he was 24-34 with the Browns from 2001 to 2004 for a .414 percentage). Davis went 5-11 and 3-7 (.313 and .300) in his last two seasons, though, so it's clear that whatever you think of Crennel as an overall coach, he didn't have a lot of talent to work with at the start.

More interesting still, though, is that the trend noted above is most pronounced here. Only one of these coaches, Phillips, has previous head coaching experience at the NFL level. Four were NFL defensive coordinators, two were NFL offensive coordinators, Kiffin was an offensive coordinator in college (with previous experience as an NFL assistant), and the other three were NFL assistants. Thus, it looks like hiring guys who have previously been head coaches might not be the best move. Let's investigate this further by looking at the worst coaches in my spreadsheet, initially by first season change.



The reverse trend appears to be taking place here. Four of these new coaches were previous NFL head coaches, and another one (Bobby Petrino) was previously a head coach in college. Of the remaining coaches, two were NFL assistants, one was a defensive coordinator and two were offensive coordinators. Thus, hiring up-and-coming guys isn't a sure guarantee of success either, but hiring former head coaches doesn't seem to work too well. It's also interesting that Raheem Morris comes in at the top of the list; he took over Jon Gruden's Tampa Bay team that went 9-7 in 2008 and led it to a 3-13 record this year. Despite that, he seems to have received much less criticism than many other coaches this year, including Tom Cable, Jim Zorn, Dick Jauron and Eric Mangini, all of whom were more successful in their first season than Morris. That doesn't mean Morris is necessarily a worse coach then those candidates, but it is curious that he hasn't received a lot of criticism for taking a team that almost made the playoffs and turning them into one of the NFL's worst franchises.

Now, let's look at the worst coaches by overall change. Here's the top 10:



Here we see many of the same offenders from the last list and many of those traditionally considered some of the worst NFL coaches, including Cam Cameron, Art Shell and Dick Jauron. The head coaching trend also continues here, as five of these coaches were previously NFL head coaches and one (Petrino) was previously a college head coach. Two of the others were NFL offensive coordinators and two were NFL assistants.

What's also interesting is that five of these hires were internal. Most of the NFL's hires have been external (from other franchises); only nine of the 38 coaches hired over that period were internal, so it's quite significant that five of those nine showed up on this list of the worst coaches out there. Here's a look at all nine coaches hired internally, sorted by winning percentage:



As you can see from that list, only two of the internal hires had an overall record over .500 with their team. Those hires are Jim Caldwell and Mike Singletary. Of course, Caldwell took over from the retiring Tony Dungy, who had a career winning percentage of .777 with the Colts and was 12-4 in his final season; not exactly the toughest situation to be thrust into. He did find a lot of success this year, taking the Colts all the way to the Super Bowl, but he had a terrific staff already in place and a quarterback who's perhaps more involved in offensive playcalls than any other player in the league. He also managed to find a way to give history the finger, angering fans and members of his team in the process, and there's a good argument to be made that Caldwell was outcoached in the Super Bowl and may have cost his team the game. Thus, Singletary's really the only impressive internal hire, as he took over a horrible team and made them into a potential playoff contender.

So, what overall lessons can we take from this study? First off, it looks like teams are generally better off going with external hires. Secondly, it looks like those candidates who haven't previously been head coaches are more likely to be successful. The first point bodes well for Shanahan, Carroll and Gailey, as none of them came from the organization they're now coaching. The second doesn't, though, as all three were previously head coaches. Shanahan was quite successful as an NFL head coach, though, while Gailey was only 18-14 and Carroll was 33-31 (but successful as a college head coach at USC). We'll see how they do, but history isn't in their favour.

Comments? Questions? Other trends you've noticed from my data? Let me know in the comments, or by e-mail, Twitter or Facebook.