A Comparison of NFL Coaches


Bill Belichick just won his fifth Super Bowl championship, further solidifying his standing as the best coach of the modern era. Belichick has been with New England for seventeen years and, during that time, he has won a record 26 playoff games, a truly incredible feat. There is simply no comparison of any coach to Belichick in terms of winning. That being said, the NFL does have some very good coaches and a number of up-and-comers who will eventually be great.


As a Bengals fan, I’m extremely jealous of the New England Patriots. Coach Marvin Lewis has not had nearly that much success in Cincinnati, even though he’s been with the team almost as long as Belichick has been with New England. When Lewis started in Cincinnati, he had an immediate impact, starting with two 8 win seasons in 2003 and 2004, then leading them to their first playoff berth in 15 years in 2005. There was a lot of reason for hope and excitement in Cincinnati with Coach Lewis at the helm. Yet, after 14 years in the role, he has yet to bring home the Lombardi Trophy. For about five years now, I’ve been arguing that Marvin Lewis simply doesn’t have what it takes to take the team to the next level. In my opinion, he is, at best, a mediocre coach, capable of making it to the playoffs on a regular basis but not capable of going much further than that. Unfortunately, Marvin Lewis is seen as a sort of god in Cincinnati, having pulled the Bengals out of the dark ages (1990-2005) and building a half-decent team, so my argument has largely fallen on deaf ears. That being the case, I decided to take a look at the data to see what it has to say.

Visualization
Instead of just focusing on Marvin Lewis, I decided to do a comparison of all current NFL coaches, looking at their win percentages, playoff success, etc. I collected a variety of data on each coach’s performance from Wikipedia and USA Today and visualized the data in Tableau. Below is a screenshot. You can click on the image to see and interact with the full visualization.


So, before we jump in, let me first help orient you to what’s here. The visualization is broken into a few different sections:
  1. Win % vs. Years (Left) – Scatter plot comparing each coach’s win percentage (with current team) to their time with that team.
  2. Career Win % (Middle) – Ranking of coaches by their career win percentages.
  3. Most Tenured Coaches – Breakdown of the 5 coaches who have been with their teams the longest.
    • Wins and Playoff Results (Top Right) – Line charts for each coach showing number of wins and results in the playoffs.
    • Wins Chart (Middle Right) – Chart showing each coach’s average wins per season, total playoff wins, and total championships.
    • Player Arrests – Raw number of player arrests (and the per season rate) for each coach’s tenure.
If you view the full visualization, you’ll find that it is fully interactive—hovering over different elements on the page will provide additional information. For instance, you can hover over each X in the arrests section to see the player name, date, and reason for the arrest. Again, you can find the full visualization here.

Analysis
While the visualization has a lot of really good information about NFL coaches in general, I want to focus my analysis on Marvin Lewis. When you look at the scatter plot, it is immediately clear that something is wrong. He is an obvious outlier, having spent many years with his team, despite a relatively low win percentage. Obviously, this alone is great reason for concern, but so much more can be gleaned from this data. Here are just a handful of additional observations:
  • With the exception of Lewis, every coach with a tenure of more than 3 years has won at least one playoff game.
  • With the exception of Lewis, every coach with a tenure of more than 7 years has won at least one Super Bowl.
  • Of the 27 coaches with at least 1 year as a head coach, Lewis ranks a meager 20th in career win percentage (53.3%).
  • With the exception of Lewis, every one of the top 5 most tenured coaches (Lewis is the 2nd most tenured) has won at least 6 playoff games.
  • Lewis has the lowest number of wins per season of the most tenured coaches.
  • Perhaps, most importantly, Lewis has taken the Bengals to the playoffs in 7 of his 14 seasons, but has lost every single one.
And that’s just wins and losses. Perhaps the most distressing statistic of his time as the Bengals coach is the sheer lack of control he has held over his players’ behavior. This has been observed both on the field—for example, their complete meltdown in last year’s AFC wild card game vs. the Steelers, a game that was all but won—and off the field, as demonstrated by the number of player arrests. Lewis has overseen 35 separate arrests in his tenure, a rate of 2.5 arrests per season (all other top 5 tenure coaches have had 1.6 or less). And many of these arrests were of repeat offenders. For example, Chris Henry, who died in a tragic accident in 2009, was arrested 6 times between 2005 and 2008. And, yet, he continued to play on the team with seemingly few repercussions. Granted, this could have been an owner or GM decision, but a strong coach with control over his team would have had enough power and leadership to either remove him from the squad or to make such an example of him as to prevent the continued offenses. Can you imagine such a thing happening on a Bill Belichick team?

So, there’s my case. The data clearly shows that Marvin Lewis is, at best, a mediocre coach, good enough to earn a winning season and playoff berth from time to time, but not good enough to lead a disciplined team to a playoff victory, much less a Super Bowl championship. After 14 years and no playoff wins, I’d think that Cincinnati management would have realized this by now. As a passionate fan of my team, I certainly hope they realize it soon.

Ken Flerlage, February 6, 2017

No comments:

Powered by Blogger.