Adjusting all-time player sack rankings for era using my Sack Index

by Jason Pauley

As I continue to play with my new favorite stat (pre-1982 sacks), I’m starting to think about era and how that impacts players’ stat totals (like most of us consider with QB stats). We know that the league played a 12 game season up to 1961, a 14 game season from 1961–1977, and a 16 game season since 1978. So, it makes sense to look at sacks per game. But the next step for me was to find out how each player’s sacks per game compare to the era that they played in.

When I look at sacks per game by year, there isn’t a strong steady trend over time, but there are distinct periods where sacks are well above / below historical average. For example, I looked at rolling 5-year averages.

Here are some of the highest and lowest 5-year league averages for sacks per game:

High sacks per game

• 1982–1986, 2.79
• 1962–1966, 2.74

Low sacks per game

• 1970–1974, 2.24
• 2006–2010, 2.16

Sack Index

Knowing that there are changes in the game that would lead to varying sack levels over different eras, I wanted to create a sack index which would compare each player’s career sacks per game to the league’s sacks per game during the period of the player’s career.

Using the all-time sack leader as an example:

Bruce Smith with 200 sacks, averaged 0.72 sacks per game. He played from 1985–2003, so I calculated the league’s sacks per game from ’85-’03, which was 2.42. To normalize Bruce Smith to his era, I created a sack index (essentially a %) by dividing his per-game average by the league average during his career. 0.72/2.42 for an index of 29.6%.

For some players, the index doesn’t have a significant impact on their ranking, but for others who were helped or hurt by the era they played in, the sack index provides useful insight into how that player ranks historically.

Here is an example the sack index can change a players’ ranking significantly as a result of the era they played in:

DeMarcus Ware and Mark Gastineau have virtually the same sacks per game with 0.78. Gastineau is ranked 4th in sacks per game on this top 50 list, and Ware is ranked 5th (slightly lower than Gastineau when you carry out the decimals). However, even though they have the same sacks per game of 0.78, over Ware’s career the league average was 2.26, and Mark Gastineau played in an era with 2.56 sacks per game, 13% higher than Ware’s era. When adjusting the two, DeMarcus Ware ends up with a higher sack index.

D.Ware: 0.78 sacks per game / 2.26 league avg = Sack Index of 34.4%

M.Gastineau: 0.78 sacks per game / 2.56 league average = Sack Index of 30.7%

DeMarcus Ware moves up from 5th all-time to 3rd all-time, and Mark Gastineau moves from 4th all-time to 11th all-time.

Players with significant moves up in the rankings when using sack index are: Cedrick Hardman, John Abraham, and Terrell Suggs

Players with significant moves down in the rankings when using the sack index are: Dexter Manley, Rickey Jackson, Mark Gastineau, and Lawrence Taylor.

Here is the rest of the analysis. I’m starting with the top 50 in all-time sacks (volume). Note that per-game rankings and sack index rankings are rankings within the top 50 list, there are likely players who didn’t have the volume to make the list with better efficiency numbers than some on this list. A good baseline target of greatness for the sack index is 30%. There are 13 players above the 30% threshold.

TL;DR Sack Index is an efficiency metric that uses each player’s sacks per game and divides their per-game stats by the leagues per game stats in the years in which they played. This normalized everything to era (style of play, passing frequency, OL blocking changes, etc) and normalizes for games played. The index uses unofficial sacks prior to 1982 which are now being reported by pro-football-reference.