Chip Kelly Wants Big Guys, But Do They Matter?

facebooktwitterreddit

Jan 17, 2013; Philadelphia, PA, USA; Philadelphia Eagles new head coach Chip Kelly addresses the media during a press conference at the Philadelphia Eagles NovaCare Complex. Mandatory Credit: Howard Smith-USA TODAY Sports

Philadelphia Eagles head coach Chip Kelly has said many times that he wants tall and long football players because, according to his experience, height provides a substantial advantage on the football field.  In other words, “big guys beat up little guys”.  However, as he transitions to the NFL and modifies his philosophies accordingly, will height afford him the same advantages in Philadelphia as they did in Oregon?

In order to answer this question, I compiled the 2012 rosters for all 32 NFL teams and used linear regression to compare team wins to team height, where team height is defined as the number of players 6’3” or taller.  The test: to determine whether team wins increase as team height increases.  Since Kelly’s desire for height applies especially to defense, I decided to look at height on defense first.

Surprisingly, I found there was no statistical relationship between defensive height and wins.  None.  At all.  When comparing variables using linear regression, you ideally would like to see the correlation coefficient (r2), a number between zero and one, to be as close to one as possible.  For defensive team height and wins, the r2 was .0002, or pretty much as close to zero as possible.  I highlighted the “trend” line in red, otherwise it would be too difficult to see:

Graph 1

How can this be?  Does height on defense really have no bearing on team success?  Well, it seems so.  In addition to comparing the number of defensive players above 6’3” to wins, I also compared defensive players above 6’ to wins, then average defensive height to wins, and then height on offense to wins.  All confirmed what you see above: there was no relationship between height on defense, or height on offense, and NFL success during the 2012 season.  Since I have faith in Chip Kelly, I was really disappointed by these results.  After all, I would love nothing more for his ideas to translate to championships in the NFL.  So I forced myself to dig deeper and eventually I did find some correlations.

I decided to abandon the idea that height on defense or offense matters, and focused more on overall team height.  I performed dozens of regressions using individual positions and position groups (like offensive lines, defensive lines, etc.), varying the players included by “games played” and “game started”, and eventually discovered that height does matter, to a certain extent, but again… not in ways you might think.

It turns out, when comparing team height to wins, where team height includes players who started at least one game during the season (so excluding practice squadders, non-starting special teamers, consistent bench players, etc.), team success increases as team height decreases.  Yes… decreases:

Graph 2

Graph 3

The correlation coefficient (r2) for starters who were 6’ tall and above was a moderately weak (but very significant) .28, and gets weaker when comparing even taller players (r2 = .14).  To me, this was worse than finding no relationship at all… the likelihood of team success actually decreases with team height! Really? I tried to find the source of this so again looked at individual positions and position groups.  I discovered that the source of this relationship primarily comes from the skill positions, WR and RB:

Graph 4

As the number of six-foot WR and RB starters decreases, the likelihood of team success increases.  In order to confirm this finding, I looked at the team wins for the top 30 running backs and top 30 wide receivers, relative to their height.

Graph 5

On average, the top 60 RBs and WRs with 10+ wins were one inch smaller than the RBs and WRs with 5 or less wins.  Convincing?  I’m not sure… after all, it’s just an inch.  But these results are still counterintuitive.  How can Chip Kelly, this really smart innovative coach who has the eyes of the NFL upon him, a coach who wants big guys, be so wrong?

Well, by looking at the results above a little closer, we might notice that the relationship with team success may be less a function of team height and more a function of the number of players and, to a certain extent, player injury.  Ideally, in a universe with no injury and optimal performance, a team will have 22 combined starters on offense and defense throughout an entire season.  In Graph 2, the team that started the most players above 6-feet tall (40), had only two wins (Jacksonville), while the team that started 23 players above 6-feet tall had eleven wins (San Francisco).  Since we don’t live in an ideal universe, this could possibly be explained by injury and/or poor performance.  As more players were injured or performed poorly, teams may be forced to start additional players, a drop in consistency and talent results that could correlate with a decrease in team wins.  To control for team height and this “injury effect”, I compared just the height of players who started all 16 games with team wins:

Graph 6

Graph 7

Finally, this is the direction I expected to see.  The relationships between height and wins are very weak (r2 = .09 and .05, respectively), but still statistically significant.  And if we take height out of the equation altogether and focus solely on the players who started all 16 games, the relationship becomes a little stronger (r2 = .15):

Graph 8

So after all of this (confusing) analysis, what can we conclude?  First, to some extent, height matters in the NFL, but not as much as we (or I) want to believe and only when that height is healthy.  Second, adding height to each position’s depth is less important than adding talent to each position (and of course, more health).  This second point is evident within the RB and WR group (Graphs 4 and 5), where team success increases as WR and RB height decreases, a relationship most likely indicative of deeper talent and a higher number of players used due to injury at those positions of skill.

In my opinion then, Chip Kelly’s preference for bigger players offers no real advantage in and of itself.  Where the advantage lies is with assembling a team that consists of “the right” big guys: men of talent and considerable health, a health that will hopefully be maximized with the help of a Sports Science Coordinator, which, coincidentally, the Eagles have.  Phew… That Chipper guy is smart.