RB Analysis Given Run Blocking Success - Through Week 4

If you are familiar with how RB's are talked about in terms of productivity, you are probably aware that Expected Points Added (EPA) is a common metric that quantifies success, as it measures how many points the RB added to their team. EPA is calculated by getting the expected points on a given play using down, distance, and field position at the start of the play and subtracting that from the points added at the end of the play.

EPA is a great metric that is helpful when looking at any position, but it does have its weaknesses. For example, EPA does not take into account how good a team's offensive line is. If a RB on a great run blocking team gets a 20 yard run on a 1st and 10 at the opponent's 40 yard line, they will get the same EPA as a RB on a horrible run blocking team that also gets a 20 yard run on a 1st and 10 at the opponent's 40 yard line. This is visibly troublesome, as a RB with a worse offensive line should get slightly more credit because it was a run with a higher level of difficulty.

This realization of the faultiness of EPA for running backs led me to analyze how well RB's run, while also taking into account their team's run blocking success. Offensive line metrics and statistics are arguably the least comprehensive among all the position groups. To evaluate how good a team's run blocking has been this season, I decided to use both ESPN's Run Block Win Rates (RBWR) and PFF's Run Blocking Grades (RBLK).

ESPN explains their RBWR metric: "In run stop win rate, a defender can earn a win by doing any of the following: beating his blocker so he's in better position to stop the runner; disrupting the pocket or running lane by pushing his blocker backward; containing the runner such that he must adjust his running lane; or recording a tackle within three yards of the line of scrimmage. If a defender earns a run stop win, his blocker earns a loss -- and vice versa."

PFF describes their RBLK grades as: "The PFF grading system isolates the role of each run blocker on every running play. Blockers can earn positives for creating movement or winning their gap, while negative grades are assigned for poor blocks, with varying levels of downgrades within the system. There are many blocks that fall into the “expected” category and will receive a neutral grade, as both the run blocker and run defender end up in either a stalemate or a situation where the play calls for both players to achieve a similar result."

I combined these two into a metric using a form of standardization that I felt was best fit for showing the true success of a team's run blocking. I then plotted each RB's team's Run Block Score by each RB's EPA per carry (to standardize RB's EPAs), and this was the result:

The first thing you probably looked at was De'Von Achane at the top of the graph. This is completely understandable as his EPA per carry blows out every other RB in the league by a wide margin. This graph should be read as being above the line is a good indicator of a RB's success. For example, Travis Etienne (bottom left in the Jacksonville Jaguars jersey) shows more success than Tony Pollard (coordinates are roughly [0.5, 0.0]). Even though Pollard has a higher EPA than Etienne, Etienne's is more impressive given that the Jaguars' Run Blocking Score is substantially lower than the Cowboys'.

Positive success is recognized by each RB's distance (on the y axis) from the diagonal line, meaning De'Andre Swift (far right) is comparable to Breece Hall (middle of the graph in the New York Jets jersey) because they are both on or right above the diagonal line. This same logic is used in reverse, as Dalvin Cook (bottom in the Vikings jersey) is comparable to Kenneth Gainwell (bottom right), even though Gainwell has a greater EPA.

I plan on updating this rather frequently during the season to account for changes in Run Block Score and RB's EPA.