Wednesday, July 23, 2014

Is Bryce Brown Expendable?

Travel Deals

Bryce Brown fumbled, again.  This time, not only did it cost the Philadelphia Eagles a possession, it cost them points.  He took the ball from Jacksonville’s 26 yard line, ran 23 yards, then lost the ball at the three yard line.  The ball subsequently travelled through the end zone for a touchback.  Points lost.  And it could have been worse.  The Jaguars drove down the field only to miss a 47 yard field goal (queue Chip Kelly: It’s not how about the turnovers, but how we respond to them).  Yes, it’s only the preseason, but last night’s fumble continues an alarming trend for the second year running back.  Bryce Brown simply cannot take care of the football.

A few weeks ago, I used a production formula to illustrate how there is a higher correlation between tight end production and team wins than running back production and team wins.  Thanks to your feedback, I’m taking this a step further.  I’ve also included in my analysis production for quarterbacks, wide receivers, and for a few defensive positions: defensive end, linebackers, and corner backs.

To calculate production for RBs, TEs, and WRs, I multiplied carries and receptions by rushing yards and receiving yards, then by touchdown points.  For more on why I did this, you can read the first article.  In a similar manner, I multiplied attempted passes, rushes, yards and touchdown points for quarterbacks.  For defensive players, I multiplied together key defensive stats: tackles, sacks, interceptions, forced fumbles, and passes defensed.  Since my results were normally distributed, I "normalized" them in order to come up with production values that were more or less on the same scale (values between 0 and 1); the higher the production value, the more productive the position.  And I do mean position, not players; production values represent averages for the players at each team’s position.  Here are the r-squared values for each position, and all positions combined (I could probably write ten more stories on these results, by the way).

According to my results, the positions with the highest relationship to wins are quarterback and wide receiver (passing league!).  Linebacker and defensive end positions are weak in comparison.  But I think this is in part (or mostly) due to the fact that many defensive positions don’t show well on the stat line (how do we quantify disruption at the line of scrimmage?).  When the production values for all positions are combined, the relationship to wins is very strong (.5667).  The interactive graph below summarizes my results in more detail.  You can filter by team, position, and season.