The NFC and AFC championship games are this weekend. The quarterbacks in these games -- Tom Brady, Patrick Mahomes, Drew Brees, and Jared Goff -- are the last signal callers playing and clearly must be the best quarterbacks in the game. Right?
Although these quarterbacks are the only quarterbacks with something to play for anymore, the NFL marketplace did not rank them among the best before the season started. According to Spotrac, here were the top quarterbacks in average salary in 2018.
- Aaron Rodgers (Green Bay Packers): $33,500,000
- Matt Ryan (Atlanta Falcons): $30,000,000
- Kirk Cousins (Minnesota Vikings): $28,000,000
- Jimmy Garoppolo (San Francisco 49ers): $27,500,000
- Matthew Stafford (Detroit Lions): $27,000,000
- Derek Carr (Oakland Raiders): $25,000,000
Not only are these quarterbacks not playing this weekend, none of these quarterbacks even made the playoffs. How can the quarterbacks the market decided were the best fail so miserably on the field?
The appendix to Sports Economics provides one answer. NFL Quarterbacks are -- relative to basketball players and hitters in baseball -- remarkably inconsistent. Only 31.1% of a quarterback's current completion percentage is explained by what that same NFL quarterback did the previous season. For passing yards per attempt, only 22.1% of current performance is explained by last year. And for touchdowns per attempt, only 10.1% of a quarterback's current mark is linked to last year's performance.
When we turn to interceptions we see predicting the future is almost impossible. Only 0.6% (i.e. less than 1%) of a quarterback's current interceptions per attempt are explained by what the quarterback did last year. This means that you can't know how many interceptions a quarterback will throw before the season starts.
Salaries are a statement about a player's future performance. The salary paid to Aaron Rodgers in 2018 was determined before the 2018 season started. This salary was set so high because the Packers believed he would perform well. But as the data indicates, the performance of quarterbacks is immensely difficult to predict. In sum, it simply isn't possible for teams to know with certainty which quarterbacks will perform well.
Why is a quarterback's performance so difficult to predict? A quarterback's passes depend on his offensive line, his receivers, the quality of his running game, the plays that are called by his coaches, and the quality of the opposing defense. In sum, quarterbacks do not act by themselves and the quality of everyone else on the field -- and the coaches off the field -- dramatically impact the outcomes we assign to quarterbacks.
Given all this, as chapter two notes, payroll in the NFL explains less than 3% of wins in the NFL. Yes, predicting outcomes in the NFL is very difficult.
So, maybe it isn't surprising the highest paid quarterbacks are watching the playoffs this weekend. When it comes to player pay in football, teams definitely do not have a crystal ball. And the inability to predict the future will mean we will see cases where the market tells one story while the outcomes on the field tell a very different tale.