The Great Quarterback Fallacy: Part I

As football fans, we often lionize the performance of great quarterbacks as the focal point of the offense and the face of the franchise. We see them carrying their team to victory with their heroics, winning big games and championships. Their greatness is measured mostly in wins, particularly post-season wins, which they are credited for delivering. The key to achieving “greatness” for a quarterback is winning a Super Bowl, although there are some exceptions to the rule. Joe Flacco is not seen as a great quarterback, despite winning a Super Bowl with Baltimore several years ago. Deshaun Watson may be seen as a great quarterback today despite an average winning percentage and only one post-season victory on his resume. The ‘great’ quarterbacks today, or top tier anyway, are Tom Brady, Patrick Mahomes, Aaron Rodgers, Deshaun Watson, and Russell Wilson. They are also, with the slight exception of Tom Brady, the most expensive.

They get paid top dollar because they deliver victories. They win Super Bowls. And everybody knows having a top-tier quarterback like one of the above is essential to winning a Super Bowl. Chances of winning a championship without one are minuscule. Even less than the 3% chance a team has in any given year, all things being equal.

But there is a funny thing about great quarterbacks delivering wins and Super Bowls, carrying their team to victory through sheer force of will and great performance: it’s not true.

Measuring Quarterback Performance

One of the most commonly used quarterback performance statistics is passer rating. It was invented in 1973, and updated in 1979, and incorporates completion percentage, yards per completion, touchdowns and interceptions according to this weighted formula:

A. Completion Percentage: Completions divided by pass attempts, subtract 0.3 and divide the result by 0.2.

B. Passing Yards per Attempt: Passing yards divided by pass attempts, subtract 3 and divide the result by 4.

C. Touchdown Passes: Touchdown passes divided by pass attempts, divided by .05.

D. Interceptions: .095 less interceptions divided by pass attempts. Divide that product by .04.

The range of each of these elements must be within 0 - 2.375. If outside that range, the maximum or minimum number is used.

Final Formula: Add A+B+C+D, multiply by 100 and divide by 6.

The formula is perhaps more complicated than it would seem necessary, but the reason and rationale is to properly weight each of the above key elements, while also limiting how much any one element can contribute to the total.

There are other statistics and derived statistics to measure quarterback rating, and each have their pros and cons, but passer rating has stood the test of time in measuring quarterback performance within a given era. Changes to the rules of the game to favor passing over the years has resulted in the average passer rating going from around 66 in the 1970s to 88 in 2017. Passer rating has remained highly correlated to winning, and the average passer rating corresponds more or less with a .500 winning percentage.

Source: Pro Football Reference

The chart above represents the correlation between winning and passer rating, on a single-game basis, from 2015 to 2020, including postseason games. As you can see from the trendline, the correlation between passer rating and winning percentage is pretty strong, and for a quarterback to really be helping his team win in a meaningful way, his passer rating should be at least 100, and anything below 80 is a pretty clear liability.

For example, a quarterback who went 22/32 for 240 yards, 1 TD, and 2 INTs would end the day with a 75.0 passer rating - clearly not a good day at the office and a negative factor to team success. On the other hand, a quarterback that goes 25/35 for 292 yards, 3 TDs and no INTs would end the day with a passer rating of 125 - clearly making a significant contribution on offense, and performing at a high level.

And so it stands to reason that any great quarterback should be performing above a 100 passer rating if he is to be considered a significant factor in delivering the win for his team - and particularly in big games like a playoff game.

The GOAT ?

Tom Brady is the undisputed greatest quarterback of all time - the GOAT. He was before winning the Super Bowl this year, and the time before too, most likely. In the six seasons between 2015-2020, he played in four Super Bowls, and won three of them- the most concentrated period of Super Bowl success in his very successful career.

His regular season record over that period was 70-22 - a .761 winning percentage. His record in the playoffs was even better- 13-3 - an .813 winning percentage. Looking at the chart above, Brady should have been consistently delivering passer ratings of 110 or better to achieve that kind of winning percentage.

But he didn’t. Not even close. Below is the chart of Tom Brady’s passer ratings and corresponding win rate over the past six seasons:

Source: Pro Football Reference

As you can see, he had a passer rating of 110 or better in only about a third of his games. And another third of his games his passer rating was under 90 - which is a negative factor in team success. The remaining third his passer rating was between 90 - 110, which according to the chart above correlates to only a modest positive in winning percentage.

This begs the question: if great quarterbacks carry their team to victory, and Tom Brady is the greatest of them all, and this was the most concentrated period of success in his career, how can he win so much when his performance was poor or below average in 1 out of every 3 games?

The fact that Brady’s team won almost half the time when he had a poor outing - sub-80 passer rating - suggests the team has other ways of winning without a strong quarterback performance. Brady’s team also had higher winning percentages at just about every passer rating range than the league overall. Both of these facts point strongly toward Brady’s team, outside of Brady’s own performance, being much better than league average.

In fact, Brady’s team had relatively modest needs when it came to passer ratings: any 90 or above passer rating- basically any above average performance- would deliver an astounding 91% win rate. Only once a year, on average, did the rest of the team come up short when Brady’s passer rating was over 90. The rest of the league didn’t achieve that level of success even when their quarterback’s passer rating was over 120.

On the other hand, given the sharp fall-off in win rate when Brady’s passer rating fell under 90, suggests Brady’s below average or worse performance was a significant factor in those losses, and even then his team was able to overcome his sub-par performance at a much higher rate than league average.

And so there is only one conclusion to draw from these data:

Tom Brady, the GOAT, the best of the best, in the most concentrated period of Super Bowl success, didn’t carry his team. His. Team. Carried. Him.

What About the Big Games - What About the Playoffs ?

Okay, so Tom Brady was largely carried by his team into the playoffs, but surely he rose to the challenge once the stakes were raised - the win or go home playoff atmosphere - and carried his team to victory.

Once again, that couldn’t be further from the truth.

Tom Brady is one of the very few quarterbacks to have played in a enough playoff games to have a significant sample size. He’s started a total of 45 in his career - one of his many career records not likely to be broken. Overall he is 34-11 in the postseason, a .755 winning percentage. Below is his winning percentage broken down by passer rating, along with the percentage of games for each passer rating.

Source: Pro Football Reference

Compared to the two earlier charts, the above chart shows relatively little correlation between Tom Brady’s passer rating and winning percentage. And more so than ever, his team’s winning percentage is exceptionally high - even when his passer rating was poor!

Now compare Brady’s chart with that of all playoff quarterbacks going back to 2012:

Source: Pro Football Reference

If you compare the percentage of games at each passer rating range, you can see Tom Brady doesn’t vary much from the average playoff quarterback. Both Brady and the average playoff quarterback had a high percentage of poor passer rating games, and nearly half were below average overall. Only Brady’s team was able to win well over twice as often, compared to the average playoff team, when their quarterback had a poor (sub-80) passer rating. Once again this demonstrates the strength of Brady’s team, and their ability to carry him in poor performances. Moreover, when Brady was able to do his part, and be at least an above average performer, his team hardly ever let him down. In fact, his team lost only 3 playoff games when Brady’s passer rating was over 90 - an .875 win rate. The rest of the playoff teams needed their quarterback to have a passer rating over 120 to achieve that rate of success.

In fact, given Tom Brady’s passer rating performance in playoff games, and the average team win rates at each level of performance, Tom Brady may never have made it to the Super Bowl in his 21-year career. The fact that he played in 10 Super Bowls, and won 7 of them, speaks overwhelmingly of the strength of the teams he played for - teams that carried him through a dozen poor (sub-80) performances and only once fell short in the 18 games he had a passer rating over 100.

Not every quarterback is as fortunate as Tom Brady has been to have a team able to carry their quarterback through bad performances at such a high rate, while almost never proving to be the weak link when their quarterback’s passer rating is over 100.

The Kirk Cousins Comparison

To put the value of team in even more stark relief, consider the Kirk Cousins comparison.

From 2015-2020, the same period for the Tom Brady chart above, Kirk Cousins performed remarkably similar to Tom Brady. Overall, Cousins had a passer rating of 100.5 over that period, while Brady’s was 100.2. Both were credited with ten 4th quarter comebacks, and Cousins had 15 game-winning drives, compared to 11 for Brady, according to Pro Football Reference. But the win-loss results couldn’t have been much more dissimilar.

Source: Pro Football Reference

Cousins had:

• a lower percentage of poor and below average passer rating games than Brady;
• a notably higher percentage of 110+ passer rating games;
• more game-winning drives and an equal number of 4th quarter comebacks;
• slightly higher overall passer rating including a slightly higher passer rating in the playoffs;

But Cousins’ winning percentage was only .516, going 49-44-2 over the six year span, and 1-2 (.333) in the playoffs. Brady went 70-22 (.761) over that span, and 13-3 (.813) in the playoffs, including 4 Super Bowl appearances, 3 championships, 2 Super Bowl MVPs, and one NFL MVP award.

But the stat that really stands out above all the rest, is the comparison between Brady and Cousins when they’ve had a game with a passer rating under 80 - basically a bad game. Brady’s team still won nearly half (45%) of those games. Cousins’ team has yet to win even one of those games - 0% winning percentage.

Brady cemented his GOAT status during this span, while Cousins seems to have secured a different goat status - scapegoat - despite quite similar performances over the same period.

Why such a stark contrast despite similar performance? The answer is simple: Team.

Tom Brady was surrounded by a team that could carry him when he performed poorly, and seldom failed him when he performed well.

Kirk Cousins’ team, in sharp contrast, has never been able to carry him through a poor

performance. Not even once. In fact, in any game Cousins had a passer rating below 100, his team managed only a .220 winning percentage. That winning percentage is even lower than league average for quarterbacks with a sub-80 passer rating. For Tom Brady, that winning percentage was .519 - over twice as high. League average was .350. But even in high-end performances - passer ratings of 110+ - Cousins’ team wasn’t much better than league average in achieving victories, while Tom Brady’s team was nearly invincible - .944 winning percentage.

Be that as it may, the high correlation between Cousins’ passer rating and team winning percentage, along with the extremely poor results when Cousins’ passer rating is below 100- or even 110 for that matter (still only a .288 win rate), may help explain the frustration with Cousins. His team needed him to perform at a very high level to have at least an even chance of winning. Brady’s team didn’t really need that - especially in the playoffs.

Bottom line, both Cousins and Brady had similar performance levels since 2015, but Brady had a team that could carry him. Cousins didn’t. One quarterback is the GOAT. The other a scapegoat. But in the end, it was the team that made the difference, not the quarterback.

In Part II, we’ll look at the relationship between quarterbacks and offensive line pass protection.