http://fivethirtyeight.com/features/...s-eli-manning/
I found the following interesting:
Quote:
Here’s the problem: This way of thinking about quarterbacks forces them to compete against themselves. Sure, the Patriots have often been favored to win their postseason games. But a lot of that is because Brady is their quarterback. How might the Pats have expected to do with a replacement-level QB instead?
They might not have been totally hopeless. Brady has usually had a little bit more talent surrounding him than Peyton Manning has. (Matt Cassel, who rates as somewhere between average and replacement-level, led New England to an 11-5 record when Brady was hurt in 2008.) Bill Belichick would probably have snuck them into the playoffs a few times. But they’d also have been playing good opponents. Our method projects them to a 12-14 or 13-13 postseason record rather than Brady’s 18-8.
I calculate these estimates based on a quarterback’s adjusted net yards per attempt (ANY/A), a metric that accounts for yardage, attempts, touchdowns, interceptions and sacks — basically it’s a better version of the NFL’s passer rating. A replacement-level quarterback typically posts an ANY/A at about 80 percent of the league average, so a QB gets credit for any performance above and beyond that.1 I then translate this into points added or subtracted in the regular season2 and translate points into a team’s Elo rating to evaluate the impact the QB had on his team overall.3
It’s notoriously difficult, of course, to distinguish the performance of a quarterback from that of his teammates, but this method produces some reasonable-seeming results. This year’s Green Bay Packers project as a slightly below-average team with a replacement-level guy subbed in for Aaron Rodgers , for instance. Instead of having been 59 percent favorites in their Sunday game against the Cowboys, as they were based on Elo ratings, they’d have been roughly 2-to-1 underdogs.
The principle is simply that the better the quarterback, the more his team would be harmed by removing him. In the case of Peyton Manning’s teams, I estimate that pulling Manning would hurt them by about a touchdown (7 points) per game. That’s enough to demote them to a projected 8-16 record in the 24 postseason games Manning has played.
|
This article projects that, with a replacement level QB, The Patriots postseason record would have been 12-14 or 13-13 without Tom Brady (they are taking out the Bledsoe game).
It projects that Manning's teams would have been 8-16 without Peyton Manning.
Pretty much has always been my point. Manning has a history of not playing like God in the playoffs, but he has always had a hell of a lot more on his shoulders than Tom Brady. And the whole idea of the "Tom doesn't have enough weapons!!!" is criminally stupid.