How do pass and rush rates influence offensive efficiency? Does establishing the run help the pass thrive? Can offensive efficiency be due in part to the rate at which teams run and pass the ball. Does a higher pass rate lead to a better passing efficiency? Through graphical analysis, we can see how correlated rush and pass rates are with offensive efficiency (run, pass, overall) through the first 5 weeks of the NFL season?

1. load libraries desired

library(tidyverse)
library(ggimage)
library(nflverse)

2. load data (nothing in the parenthesis means the current year)

data <- load_pbp()

3. create data frames for rushes, passes and epa (overall, rush, pass)

total_rushes <- data %>%
  filter(rush ==1) %>%
  group_by(posteam) %>%
  filter(posteam != 'N/A') %>%
  summarise(rushes = n())
total_passes <- data %>%
  filter(pass == 1) %>%
  group_by(posteam) %>%
  filter(posteam != 'N/A') %>%
  summarise(passes = n())
team_epa <- data %>%
  group_by(posteam) %>%
  filter(posteam != 'N/A') %>%
  summarise(avg_epa = mean(epa))
team_rush_epa <- data %>%
  filter(rush == 1) %>%
  group_by(posteam) %>%
  filter(posteam != 'N/A') %>%
  summarise(rush_epa = mean(epa)) 
team_pass_epa <- data %>%
  filter(pass == 1) %>%
  group_by(posteam) %>%
  filter(posteam != 'N/A') %>%
  summarise(pass_epa = mean(epa))

4. join all the data frames into one new data frame “data”

data <- total_passes %>%
  left_join(total_rushes, by = 'posteam') %>%
  left_join(team_epa, by = 'posteam') %>%
  left_join(team_rush_epa, by = 'posteam') %>%
  left_join(team_pass_epa, by = 'posteam')

5. mutate columns to get rush rates and pass rates

data <- data %>%
  mutate(plays = rushes + passes,
         rush_rate = rushes/plays,
         pass_rate = passes/plays)

6. plotting time!

The teams that run the ball definitely seem to thrive off it. As some of the best teams in terms of rush epa run the ball a lot like the Browns and Falcons. It can also be seen that teams like the Bills and Buccaneers (who both run the ball infrequently) are some of the worst teams in terms of rush epa/play.

There seems to be no correlation between pass rates and pass epa/play, rush rates and epa/play, pass rates and epa/play, and rush rate and pass epa/play.

There is however, a very negative correlation between pass rate and rush epa/play. As some of the teams who run the ball the least (Bills and Buccaneers) are some of the worst teams as running the ball.

Conclusion

Is “Establishing the Run” a true statement? Through the first few weeks this season there doesn’t seem to be much of a correlation between rush rate and pass inefficiency. What is true, however, is that teams who run the ball a lot are playing to their strengths.

Unfortunately, there is no evidence to conclude that play type rates are linked to offensive efficiency. Aside from rush efficiency, it seems that the most efficient offenses find a lot of success regardless of the rate they run or pass the ball. Besides the Chiefs and Bills, most of the teams in the league all cluster in the (-.1, .1) epa range, due to this, there’s almost no correlation from play type rates as to why the offenses perform the way they do.