Evaluating QBs by their playoff performances.
Quarterbacks are often evaluated by their record and/or their playoff record, rather than their actual playoff performances. There are five quarterbacks in NFL history who have started 20 or more playoff games (they are all fairly recent players, since the NFL playoffs have expanded since the days of one NFL championship game). If we judge these five players by their statistical performance--not by wins, not by championships--here is what they look like.
Brett Favre (24 games)
60.8%, 5,855 yards, 44 TDs, 30 INTs, 86.3 rating, 7.4 yards per attempt
Tom Brady (23 games)
62.9%, 5,629 yards, 41 TDs, 20 INTs, 89.1 rating, 6.76 yards per attempt
Joe Montana (23 games)
62.7%, 5,772 yards, 45 TDs, 21 INTs, 95.6 rating, 7.86 yards per attempt
John Elway (22 games)
54.5%, 4,964 yards, 27 TDs, 21 INTs 79.7 rating, 7.63 yards per attempt
Peyton Manning (20 games)
63.2%, 5,679 yards, 32 TDs, 21 INTs, 88.4 rating, 7.46 yards per attempt
Of this bunch (and by the way, Bart Starr in his 10 games and Kurt Warner in his 13 games were each way better), Joe Montana still looks better than anybody, with higher rating and yards per attempt. Favre, Brady, and Manning are rather comparable (though Brady's yards per attempt is low, Manning's TD total is low, and Favre's INTs per game is a little high). Elway is far behind.
But let's look at a different number: the team's points allowed per game in the playoff games these QBs started.
Favre: 22.54 points allowed
Brady: 19.52 points allowed
Montana: 18.57 points allowed
Elway: 23.86 points allowed
Manning: 21.75 points allowed
Isn't it interesting--and probably expected--that those two QBs revered for their playoff records--Tom Brady and Joe Montana--each played on teams that allowed fewer than 20 points per game in their playoff starts? Certainly both players have admirable playoff numbers, but the points per game data suggests that, for example, Brady had a three-point advantage over Favre when it came to actually winning these games (I'm going to assume each of these QBs had turnovers that led to points for the opponent--for now I'll consider that all a wash).
Now for the sheer hell of it, the points per game for each of these teams scored in these QBs' starts:
Again, special teams and defense play a role in getting a team points, and for now let's consider that a wash too. But if we're going to use imperfect team success measurements for a QB's playoff legacy, points per game makes more sense than wins and losses, simply because a QB has more direct impact and control over his team's points. If we look at it this way, Montana and Brady are barely better than Favre, about a point better than Elway, and about two points better than Manning.
I'm not proposing anything radical, and I don't claim that this is anything other than simple (you can take out a calculator and go to pro-football-reference.com just like I can). But it's something else to consider when we'll be hearing a lot about QBs' playoff performances and legacies this weekend.
Sunk Costs and the NFL (The New Yorker).