One of my favorite parts of Pro-Football-Reference is a blog written on March 25, 2009.
This blog measures the support a quarterback receives from his defense throughout his career. Then the numbers are crunched to figure out if the quarterback has produced more wins or fewer wins than expected throughout his career.
Briefly summed up below, it is broken into six categories, and included is the probability that an average quarterback would win the game.
Defense allows 0-10 points: A quarterback should win 93.5 percent of games.
Defense allows 11-15 points: A quarterback should win 76.1 percent of games.
Defense allows 16-20 points: A quarterback should win 52.6 percent of games.
Defense allows 21-25 points: A quarterback should win 38.8 percent of games.
Defense allows 26-33 points: A quarterback should win 17.1 percent of games.
Defense allows 34+ points: A quarterback should win 3.6 percent of games.
Like every other ranking system, measuring a quarterback's defensive support has its flaws. It does not take into account the quarterback's offensive support, such as strength of the running game, strength of receiving corps, or the strength of the offensive line.
It also doesn't count interception touchdowns, which make the defense appear to be worse than it really is.
And it could be slightly skewed to help or hurt a quarterback based on the range of numbers given.
Yet both Warner and Manning are credited for victories in which "the defense allowed 34+ points" even though Warner had to work much harder to win than Manning.
Then again, Manning's defense only allowed 27 points. Seven points came on an interception touchdown by the Patriots' Asante Samuel, which essentially is all Peyton Manning's fault.
So the system is not perfect and it has minor flaws. Then again, so does passer rating and most of the statistics existing today.
But in the long run, this should produce virtually accurate results.
On Pro-Football-Reference, quarterbacks are measured throughout their career by the support (or lack of) given by their defense in the regular season.
Through the end of the 2008 season, Peyton Manning is the most successful quarterback in NFL history, based on actual wins compared to expected wins.
In 191 games, Peyton Manning has won 124 times as a starting quarterback, against 67 losses. The Colts defense has been slightly below average since the start of Manning's career, meaning an average quarterback should have won 92.8 games, while losing 98.2. Manning's 124 wins mean he has won 31.8 more wins than average.
I updated Pro-Football-Reference's list to include the results of the 2009 season, where Manning keeps his spot at the top of the list.
Manning made all 16 starts during the 2009 season and won 4.88 more games than the average quarterback, an amazing total for a single season. Even more incredibly, Manning picked up 5.22 more wins than the average quarterback during the first 14 starts of the season. Then backup Curtis Painter played the majority of the final two games, but because Manning started the game, he is credited with both losses, dropping his wins above average total from 5.22 to 4.88.
When the overall list is updated to include the 2009 season, Manning holds the top spot with 36.1 extra wins.
Brett Favre is second with 29.7 extra wins and Tom Brady is third at 27.9.
Rounding out the top 10 are John Elway, Dan Marino, Joe Montana, Ken Stabler, Johnny Unitas, Daryle Lamonica, and Jim Kelly.
However, a quarterback who played in 250 games like Fran Tarkenton is likely going to rank higher than Otto Graham, who played in just 78 games. So I took the number of games played and divided it by the number of extra wins to find the most successful quarterback "per game" instead of throughout their whole career.
When factoring in the amount of games played, Brady is easily the most successful quarterback in NFL history, with an extra win every 5.16 games. Manning is second best, with an extra win every 5.73 games.
After viewing quarterback's results in the regular season, I was doubly curious to find out their postseason results.
Below I included a data table of the 59 quarterbacks in NFL history who made six or more starts in the playoffs, with a few notable extras thrown in at the end.
Then I simply used Pro-Football-Reference’s formula for quarterbacks and applied it to the playoffs.
One of the more difficult was figuring out when to credit a quarterback who only played part of a game for a win or a loss. Obviously, both quarterbacks can't pick up the decision, so I had to choose one. Usually, I picked the quarterback whose performance most affected the outcome of the game.
For example, I chose to not blame 49ers great Steve Young for the loss to the Green Bay Packers in the divisional round of the 1996 playoffs, simply because Young threw just five passes before an injury knocked him out of the game. When backup Elvis Grbac entered the game, the outcome was still completely in doubt.
In Super Bowl V, Baltimore great Johnny Unitas was knocked out of the game in the second quarter with an injury. Earl Morrall entered and the Colts ended up winning on field goal with five seconds left. In this situation, I gave Morrall the victory. Had the Colts lost, I probably would have given the loss to Unitas, because he played poorly before his injury.
In a similar situation two years earlier, I gave Earl Morrall the loss in Super Bowl III when he was benched with the Colts trailing 13-0. Unitas entered and didn't play much better but since the result of the game had practically been decided, he was credited with neither a win or a loss.
Below is the complete table of all quarterbacks in the postseason. I included their name, which team(s) they have taken to the postseason, their overall playoff record, their projected record according to average defensive support, whether they were above or below average and by how many games, and how many championships they have won, with the number of championship games they have played in included in parentheses.
Min, Was, TB
Car, Nyg, Ten
Sea, KC, Det
SF, Phi, TB
Analysis of the Results: The Winners and the Losers:
41 of the 59 quarterbacks won more games than average, while 18 underperformed in the postseason. Just 23 of the 59 overperformed or underperformed by more than a single game.
According to my calculations, the 10 quarterbacks who have won the most games above average in their postseason career are: Tom Brady (4.96), John Elway (4.38), Kurt Warner (4.10), Troy Aikman (3.25), Ben Roethlisberger (3.22), Joe Montana (3.12), Brett Favre (2.80), Steve Young (2.58), Dan Marino (2.39), and Terry Bradshaw (2.20).
When factoring the amount of games played, the most successful postseason quarterback in NFL history is two-time Super Bowl champion Ben Roethlisberger, with an extra win every 3.11 games.
Roethlisberger has mastered the art of winning close games in the postseason, such as 20-17 versus the Jets in his first playoff game, 21-18 against the powerful Indianapolis Colts in 2005, and 27-23 against Arizona in last year’s Super Bowl.
However, it should be noted that Roethlisberger could easily move up or down the list in the next 10 seasons or so. After all, had this list been made five years ago, Brady would be hands down the most successful quarterback in postseason history with a 9-0 mark and 4.0 wins above average.
Since 2005, Brady is 5-4 in the postseason and has dropped from four games above average to 3.63.
Roethlisberger just edges out future Hall of Fame quarterback Kurt Warner, who has an extra win every 3.17 games. Warner is the only quarterback in postseason history to twice win a game in which his defense allowed 37 or more points and he is the winner of the highest scoring game in postseason history.
An interesting note is that if Roethlisberger had failed to lead a last second touchdown drive against Warner's Cardinals in Super Bowl XLII, Warner would be the best postseason quarterback in NFL history.
Rounding out the top 10 are: Tom Brady (3.63), Drew Brees (3.77), Troy Aikman (4.62), Steve Young (5.04), John Elway (5.22), Dan Fouts (5.56), Joe Theismann (6.02), and Bernie Kosar (6.36).
Brady’s ranking came as no surprise to me.
He is the only quarterback to appear in the top five for wins per game in both the regular season and the postseason. Four times Brady engineered a last second game-winning drive in a postseason game, and if not for epic collapses by his defense in both the 2006 AFC championship game and Super Bowl XLII, Brady would occupy the top spot on both the regular season and the postseason lists.
Drew Brees’ fourth place ranking is the result of a quarterback who has played extremely well in just six games. I’m excited to see where he will rank on the list after several more postseason games.
For anybody who says that Aikman’s postseason success came as a result of a strong defense, they’re talking about the wrong Cowboys quarterback. That would be the legendary Roger Staubach, who won just .80 games above average and benefited greatly from a Doomsday Defense which allowed 10 points or fewer in seven playoff games.
Steve Young (5.04) ranked significantly higher than his predecessor Joe Montana (7.37). Montana benefited from a very strong defense which put him into position to win at a much higher rate than Young. In fact, Young’s defense has performed worse in the postseason than every single Hall of Fame quarterback except for Dan Marino and Dan Fouts.
Fouts has always been considered a poor postseason quarterback, due to his subpar 3-4 record and zero Super Bowl appearances. Yet not many seem to realize that Fouts’ defense allowed 27 or more points in five of his games. In fact, Fouts was winning quarterback in a classic 41-38 overtime duel against the Miami Dolphins in a 1981 wild-card game.
I was surprised to see Bernie Kosar on the list. Kosar sports a disappointing 3-4 playoff record. However, he has not played a single postseason game in which his defense allowed fewer than 20 points.
Three notable postseason heroes failed to impress me.
Jim Plunkett has always been named as a snub from the Hall of Fame based on his two Super Bowl rings, but in reality, he won less than half a game more than he should have.
Otto Graham, with his 10 championship appearances and seven titles, won just .48 games above average, and twice Graham failed to win when his defense allowed just 17 points.
Johnny Unitas, who will forever be remembered for his incredible drives at the end of the 1958 NFL championship game, won just .41 extra games in his postseason career. He also played poorly before an injury knocked him out of Super Bowl V and would have been credited with a loss if Earl Morrall hadn't rallied the Colts to a last-second victory.
And there's one final guy I have to address. Peyton Manning.
Twist the statistics all you want, but there's just about no way to defend it. The Colts' superstar is simply average in the postseason or maybe just a little bit above, given his extra half win over 18 games. He may be a four-time regular season MVP, but in the most important month of the year, he has performed at a lesser rate than Jake Plummer, Craig Morton, or even Rex Grossman, the very quarterback he 'outdueled' in Super Bowl XLI.
In case you were wondering, the five worst postseason quarterbacks based on their career totals are: Billy Kilmer (-1.76), Jack Kemp (-1.53), Daryle Lamonica (-1.36), Drew Bledsoe (-1.31), and Dave Krieg (-1.04).
In terms of per game standards, the five quarterbacks are the same, but in the following order: Jack Kemp (-3.92), Billy Kilmer (-3.97), Drew Bledsoe (-4.58), Daryle Lamonica (-6.62), and Dave Krieg (-8.65).
Just five of the Hall of Fame quarterbacks listed underperformed in the postseason: Bob Griese, Fran Tarkenton, Sammy Baugh, Warren Moon, and Joe Namath.
As for Namath and his guaranteed victory in Super Bowl III, it’s really not that hard to win when your defense allows just seven points. It’s a 95 percent + guarantee, and while Namath does deserve credit for leading his team to a 16-7 victory, imagine how Namath would be viewed today if his defense had allowed 17 points.
Regarding Dolphins’ great Dan Marino, yes, he never won a Super Bowl.
He also played with an absolutely dreadful defense in the postseason. He should have lost more than 70 percent of his games, yet somehow managed to win an extra 2.4 games in his career. And even more ironically, Marino led his team to the same amount of points in his Super Bowl appearance (16) as Namath did.
Eagles' quarterback Donovan McNabb might be the prime example of a quarterback who wins when his defense plays well and loses when his defense plays poorly. McNabb is 9-1 when his defense allows fewer than 20 points, and 0-6 when they allow 20 or more. As it is, McNabb has won more games than any other quarterback in NFL history who has actually underperformed in January football.
The aforementioned Jake Plummer also fits the bill. He won when his defense allowed seven and 13 points, and lost when they allowed 34, 41, 41, and 49. You could substitute almost any quarterback in Plummer's place and he would win the same two games and lose the other four.
What I Learned:
It's simple. How much support a quarterback receives from his defense is much, much more important than people have realized.
It can easily make or break a quarterback's career, as well as his reputation. It shouldn't, but it has.
Who knew that Dan Marino (8-10 postseason record) actually won more games above average than Terry Bradshaw (four Super Bowl rings)?
How many people are aware that a guy like Trent Dilfer, who has a 5-1 postseason record, actually underperformed in the postseason?
Or that former Eagles and Vikings great Randall Cunningham, with his dreadful 3-7 playoff record, was basically right on track with an average quarterback?
So the next time you pass judgment on a quarterback's postseason play, make sure that you have looked at more than just his numbers and his win-loss record.
Check his defense.