What Are the Immediate Effects of an NFL Coaching Change?
I will never intend to insult a specific fan base in my professional writing. In my personal time, sure; there Iâ€™ll get into mocking the relative merits of oneâ€™s home team compared to my home team, which has won multiple titles and our fan base -- who is insanely devout no matter what.
Professional article space is not for that purpose, but it is a little bit surprising how quick some fair-weather fans and fair-weather franchises are to give up on their teamsâ€™ chances so quickly. The most frequent victim of this front office and box office pressure is the coaching staff.
Iâ€™ve crunched the numbers: in the past fifteen years, there have been 101 new head coaching hires across the 32 teams. This means that -- on average -- a head coach can expect his job to last three years. Thatâ€™s it. Three years. For some of us younger folks, we have held maybe one job for longer than two years, and the only other thing we did as long was high school or college. But consider that these NFL head coaches are mostly middle-aged career men and they need stability for their families.
In the â€œNot For Longâ€ NFL, however, no one is safe without results. Yet I have become curious about just how much a new coaching staff truly affects a franchiseâ€™s success in the immediate years after earning the positon. Most teams fire their head coaches because they werenâ€™t winning and want instant success. Fans demand immediate gratification. Does a new head coach actually provide that?
One of the worst offenders of the â€œkarmic coaching cycleâ€ are the Cleveland Browns, who have had only two coaches in the last 15 years last longer than two seasons. The worst team by far, however? The Oakland Raiders: Jon Gruden was fired in 2001, Bill Callahan in 2003, Norv Turner in 2005, Art Shell in 2006, Lane Kiffin in 2008, Tom Cable in 2010, Hue Jackson in 2011, and Dennis Allen midway through this year (2014). The epitome of stability, that Davis family.
So, did ownership and the front office really help their teams by canning their coaches?
To answer this, we turn to our trusty friend, the Net Expected Points (NEP) metric, found exclusively here on numberFire. NEP is a measure of how a player or team contributed expected points to their teamâ€™s value. By looking at the probability of each play and points that they should have scored, we can see which teams were truly proficient and which got a little lucky here or there. With this, we can see beyond the box score and within the win-loss record to judge teams, players, and even coaches. So, how do our coaching staffs match up?
The best-case scenario of a coaching hire providing instant success has to be Bill Callahanâ€™s hiring in 2002 to take over the Oakland Raiders after Grudenâ€™s firing. Callahanâ€™s 2002 Raiders have the highest Total NEP -- combined Offensive and Defensive NEP totals -- of any first year coach in NEP history (181.83), which dates back to 2000. However, if we look at Grudenâ€™s 2001 Raiders, we can see that they still had a Total NEP value of 71.47 NEP. Not bad, and this indicates that the rosterâ€™s talent may have done more to affect the team than Callahan himself.
The coaching change that actually provided the highest leap up for a franchise was actually when the Baltimore Ravens fired Brian Billick after the 2007 season and brought in John Harbaugh for the 2008 season onward. Billickâ€™s 2007 Ravens had a Total NEP value of -121.93. Harbaughâ€™s arrival brought about a titanic shift, as these AFC North punching bags rose to a Total NEP value of 178.54 in 2008, for a total swing of 300.47 Total NEP in one year. Thatâ€™s what front offices hope for when they change regimes. But do they get it?
But weâ€™re not here to look at specific names, as impressive as they are. We want to know about the big numbers behind these nixings, the deep data underneath these dismissals. To see if this is a widespread phenomenon, or just a few examples of coaching prowess shining, I accumulated the NEP data of every head coaching change since 2000 -- which is our earliest NEP data -- and averaged out teamsâ€™ NEP in Year One, Year Two, and Year Three of a coaching change. To give a baseline comparison, too, I included â€œYear Zero,â€ the last season of the previous coaching staff with the team.
The table below shows this data, represented by Adjusted NEP, Adjusted Defensive NEP, and Total NEP. Remember, negative numbers are better in Adjusted Defensive NEP. What sort of pattern do we see?
|Year||# of Coaches||Adj. NEP||Adj. D NEP||Total NEP|
Shockingly, this is a fairly linear graph that shows that from Year Zero to Year One, there is a sizable tick up in team production by NEP in all measures. Teams start out absolutely dreadfully on average in Year Zero -- which makes sense, if firing your coach is on the table -- but they see a minimum 10-point gain on average in both Adjusted NEP and Adjusted Defensive NEP. Iâ€™ll be honest, I expected closer to a minimal gain or no real gain at all, but this is fairly clear data that shows immediate results when a franchise replaces its head coach.
This table also shows us that the longer a coach stays, the better a team gets on average. There is a marked swing up in value each year by both Adjusted NEP and Total NEP, though the graph for Adjusted Defensive NEP is not as linear after Year One. The problem, however, is that most of these coaches do not get to reach multiple years before theyâ€™re let go. Looking at the number of coaches in the last fifteen years who have reached Year Three with one team shows us that the number is nearly halved from the new hires.
Teams have to have a little bit more patience with their coaches, and understand that the more time they give them to shape their vision, the better. However, it apparently is perfectly reasonable to expect immediate turnaround from a coach after theyâ€™ve been hired; the data says as much. If they donâ€™t show immediate improvement in Year One, then they can be tossed back on the heap. But if they improve the team at all, give them a shot.
Rome wasnâ€™t built in a day.