What if sacks were a stat when Andy Robustelli played?
My analysis suggests he may have been top ten all-time
by Jason Pauley
Andy Robustelli was probably one of the best QB sackers in history. The unfortunate thing is that he was already past his physical prime when 1960 came around, which is where unofficial sack data now begins. Thanks to the research by John Turney and Nick Webster we now have new sack data starting at 1960. But in 1960 Robustelli was already 35 years old. He played from 1951–1964 until age 39. He had three double-digit sacks seasons in the 5-year window where we now have sack data. Here are his career sack totals:
In the short period of his career where we do have sack data, Robustelli averaged 0.77 sacks per game. This is on par with all-time QB sackers like Lawrence Taylor (0.77), Aaron Donald (0.78), and J.J. Watt (0.79), except that they benefit from their prime years being included in the data.
Can you imagine what a guy probably did earlier in his career when you see that he was getting 15.5, 12.0, 14.5, and 6.5 sacks from age 36–39 in a 14 game season? It’s safe to assume that he was even better in his earlier years. For context, I created and curve of sack data by the number of years in the league from the top 20 players in all-time sacks to visualize where the peak years are which might help us think about what could have been for Andy Robustelli. Note that this is a sample size of 20, so an odd jump in the data, like in year seven, is most likely due to the small sample size rather than signaling some causal relationship between sack data and a player’s 7th year.
Looking at this chart through the lens of Robustelli’s career, you can probably imagine what a terror he was considering we only have his data from seasons 10–14. Another way to continue to drive home my point that Robustelli was dominating during his probable decline is by looking at the all-time sack leaders from age 35 and up. No one in the history of the NFL has had more QB sacks from 35+ than Robustelli had. Here are the top 5:
If you look at the top 5 all-time leaders in sacks and isolate their 35+ years, 4 of the 5 had a much lower sacks-per-game rate than Robustelli. The only one who had a better rate was Kevin Green, and it’s probably not crazy to assume that Greene would have dipped below Robustelli’s 0.77 if he player until the same age as Robustelli (Green played to 37).
Where would Robustelli have ended up in all-time sacks? I attempted to answer this by backing into his earlier years using the curve I showed earlier. A couple of notes about his career before I back into the data. It appears that Robustelli missed only one game in his career, playing 175 out of a possible 176 of his team’s games, and starting in 169 of them. He was a Pro Bowler in 7 of his 14 seasons and an All-Pro in six of them. He was either an All-Pro or Pro Bowler in all but three of his nine seasons (1st, 2nd, and 4th) prior to 1960. This sustained success doesn’t give me any concern about backing into his numbers using that curve. Another important point is that he started his career about age 26, about three years later than most players. This late start was due to his service in the military following high school. Given the fact that he was 26 in his rookie year, I believe that my estimates will be conservative. Generally speaking, a 26-year-old player is much more mature physically and mentally than a 22-year-old player, so I wouldn’t be surprised if his production in his early years didn’t start as low as my production curve chart would suggest.
Here is the same production curve chart I showed earlier, with Robustelli’s data included. The variation in Robustelli’s data vs the total is the result of a sample size of one being compared to a sample size of 20. Many of the players within the sample size of 20 would also have similar variability in their data. If you look at Robustelli’s average sacks-per game of 0.77 from years 10–14, his rate was 22% higher than the average of the top 20 all-time sack leaders during the same years (10–14). But now, having both sets of data (the group and Robustelli) I have the information needed to back into his previous years.
Because of the variability in Robustelli’s final five years, I had trouble determining a starting point to work backward from. So I thought it might make more sense to break this information into segments. I used segments that had logical starting and stopping points based on the shape of the chart. I decided to lump the production curve into year 1, year 2, year 3, peak years (years 4–9), and decline years (years 10–14). Perhaps the decline years could have been split into early and late decline, but the 10–14 year time frame works well because that allows me to use Robustelli’s average for that entire segment.
Here is the production curve segmented into larger groups comparing the top 20 groups actual to Robustelli’s projection. To project Robustelli, I back into the data using the increase or decline at each segment using his 0.77 from years 10–14 as the starting point. For example, the peak years (years 4–9) have a sack rate that is 27% higher than the decline years (years 10–14). By applying that 27% increase to Robustelli, that takes his 0.77 from the decline years to 0.98 sacks per game during his peak years. The same logic applies to each segment.
Now that I calculated the rates, the fun part is applying those per-game rates to Robustelli’s number of games played each season and rounding to the nearest 0.5 to get the season and career sack totals. Because I have grouped the years, the season-to-season consistency for Robustelli’s peak years is unrealistic, but I’m not going to predict variability in that data, and ultimately the goal is to determine a best guess in his career sack totals. The final projection puts Robustelli at 151.0 sacks and a per-game rate of 0.86 sacks per game. He would be 6th all-time in career sacks and 2nd all-time in sacks per game.
Here are the top 10 in both categories along with Robustelli’s hypothetical numbers:
Obviously, we don’t know where his sack numbers really are. There are probably people smarter than me who would have a projection using a better methodology, but the differences probably wouldn’t change his numbers dramatically. You can see that is not a reach to suggest that Robustelli would be in the top 10 all-time in sacks and top 5 all-time in sacks per game. Someday, great researchers like Turney and Webster may have data from the 1950s and we will know Robustelli’s actual stats. And perhaps there will be guys before Robustelli, like Len Ford, who will shoot to the top 10 all-time. But for now, we can have fun forecasting backward. Andy Robustelli is a Hall of Famer, a 6-time first-team All-Pro, 4-time second-team All-Pro, and a 2-time champion, he is not lacking recognition, but it’s still helpful to put another number to his greatness and to see where he stacks up against other great players in NFL history.
More of my posts and analysis using the new pre-1982 sack data:
Adjusting all-time player sack rankings for era using my Sack Index
From Gino Marchetti to Bruce Smith: The evolution of the NFL sack record
