NFL Passer Rating, Why Not a Rusher Rating Too?
The NFL has created a passer rating to be able to compare players across the league and across the years. Why not a rusher rating?
I am proposing some equations for evaluation and comment.
My long-term goal is to be able to compare different types of teams to be able see if there are any insights as to why one team beats another team even though statistically they are weaker. I will apply passer, rushing and receiving ratings to team offenses and defenses for a team to team comparison of the 2008 season and compare mathematical predictions with results. but that is a later article.
The NFL passer rating is based on the best performance ever by a QB and then set on a scale from 0 to 2.375. The best performance ever is set at 2.0. The maximum is set to 2.375 so they can leave some room for improvement.
For instance:
**Completion % **- The NFL record is 70.55 by Ken Anderson (Cincinnati, 1982). The NFL set this at a 2.00 with a zero is 30% then a 2.375 equals 77.5%.
**Yards/Completion **- The NFL record is 11.17% by Tommy O'Connell (Cleveland, 1957). They set this at a 2.00 and the zero to 3 yards/completion then a 2.375 equals 12.5 yards/completion.
**TD % **- The NFL record is 13.9 by Sid Luckman (Chicago, 1943). The NFL set 10% at 2.00 for some reason with 0 at 0% then a 2.375 equals 11.9%.
**Interception % **- The NFL doesn't explain how they came up with these values but they set a 2.375 to a perfect season of no interceptions and a 2.00 is set to 1.5% interceptions. This makes the 0 a 9.5% interception rate.
I propose the following.
Set the rushing rating on these criteria:
The NFL sets the minimum number of passes per game at 17. The ratio of passing plays to running plays for the regular season is 16526 passes to 14119 rushes. So if you ratio the plays we should set the minimum rushes limit at 14.5 plays say 14 plays per game. I cut the date off at 1990 for yards per carry and TD's and used 1992 for fumbles because fumble data was not gathered before that.
**Yards/Rush **- The best average was Barry Sanders of the Detroit Lions in 1997. He had a 6.1 yards per carry average for the season. If we use the same low end as passing uses 3 / 11 x 6.1 we will get 1.66 yards per carry for a zero and 6.1 yards per carry for a 2.00 and 6.9 for a 2.375.
**TD% **- The best TD% for a season was Priest Holmes of the Kansas City Chiefs in 2003. He scored 27 TD's on 320 carries for a TD% of 8.44%. Using 8.44 as the 2.00 and 0 TD's as the 0.00, we find that the 2.375 is a TD% of 10%.
**Fumble% **- The worst fumbler from the years 2008 to 1992 was Garrison Hearst of the Arizona Cardinals. He fumbled 12 times on 284 carries for a Fumble% of 4.23%. It would have been Reggie Bush in 2007 with a 4.46% but he only carried the ball an average of 13.1 times per game. If we set this as the zero and 2.375 as having no fumbles at all, the 2.00 will be 0.67% fumbles.
Now on to the equation.
Let's use Chris Johnson as an example. He had a great season with the Titans (1,228 yards on 251 carries with 9 TD's and 1 fumble and then we'll compare him to DeAngelo Williams who had an unbelievable season with the Panthers (1515 yards on 273 carries with 18 TD's and 0 fumbles)
Chris Johnson
Yards / Carry Component: Y/C equals 1,228 yards divided by 251 carries for 4.86 Y/C. Subtract 1.66 from this to get 3.23. Now divide that by 5.3 and multiply by 2.375 to get a point rating of 1.45.
Fumble % Component: Chris fumbles 1 time in 251 carries for a 0.4% fumble %. Subtract this from 4.23% (3.93%) and divide the result 4.23 (0.905). Multiply by 2.375 to get a point rating of 2.15.
TD % Component: Chris scored 9 TD's on 251 carries for a TD % of 3.59%. divide this by 10% (0.359) and multiply by 2.375 for a point rating of 0.85.
Add all three components (1.45 + 2.15 + 0.85) = 4.45. This sum is then divided by 4.5 and multiplied by 100 to get a Rusher Rating 98.9. Not bad.
DeAngelo Williams
Yards / Carry Component: Y/C equals 1,515 yards divided by 273 carries for 5.55Y/C. Subtract 1.66 from this to get 3.89. Now divide that by 5.3 and multiply by 2.375 to get a point rating of 1.75.
Fumble % Component: DeAngelo didn't fumble at allso his Fumble % is 0%. Subtract this from 4.23% (4.23%) and divide the result 4.23 (1.00). Multiply by 2.375 to get a point rating of 2.375, a perfect score.
TD % Component: DeAngelo scored 18 TD's on 273 carries for a TD % of 6.59%. Divide this by 10% (0.359) and multiply by 2.375 for a point rating of 1.57.
Add all three components (1.75 + 2.375 + 1.57) = 5.695. This sum is then divided by 4.5 and multiplied by 100 to get a Rusher Rating 126.56. The best this season.
When you compare Chris Johnson with DeAngelo Williams you can see that Chris's TD% is much lower that DeAngelo's, with the resulting difference in Rusher Rating.
Now we can start comparing teams by relating the team passer rating, team rusher rating and defense rating against the pass and the rush.