Writers of Pro Football Prospectus 2008

Most Recent FO Features

SimBla15.jpg

» OFI: Don't Make Saban Angry

Notre Dame and Baylor entered the one-loss group in what is shaping up to be an extremely tight race for playoff consideration.

28 Jan 2014

Quarterback Postseason Drive Stats

by Scott Kacsmar

In the first part of this series on quarterbacks in the postseason, I looked at some of the fallacies in how quarterbacks are judged by win-loss record along with opponent adjustments for passing stats. Today, I have an update to the playoff drive stats I first introduced three years ago at Pro-Football-Reference.

Each year this study grows and it now includes 32 quarterbacks (14 active) and 4,346 drives. I have long advocated for the use of drive stats since they adjust for the pace at which games are played. Looking at quarterbacks through this prism can provide great context, although we need a bit of a caveat since these are stats for the entire offensive unit.

Disclaimer: While passing stats are not truly individual stats either, the drive stats are even more teammate focused. There are of course drives where the quarterback never drops back and just hands the ball off every play. The entry "Russell Wilson" is another way of saying "2012-13 Seattle Seahawks" and also an abbreviation for Russell Okung, Marshawn Lynch, Golden Tate, Doug Baldwin, Zach Miller, offensive coordinator Darrell Bevell, etc. Drive stats are not adjusted for opponent.

Generally, only games started by the quarterback are included, but there are some significant backup performances included for Steve Young, Troy Aikman, John Elway, Mark Brunell, Dave Krieg, Bernie Kosar and Randall Cunningham.

Most of these stats will be familiar to users of Football Outsiders' drive stats -- my inspiration for collecting this data in the first place. Since we're specifically looking at quarterbacks, there are some differences in calculations. Yards per drive are still net yards that include penalties, but penalties during punt and field goal formations are excluded here. As always, kneeldown drives are excluded and any kneeldown is removed from the drive stats.

General Drive Stats

Our first table includes the general drive stats for yards (Yds/Dr), points (Pts/Dr) and average starting field position (LOS/Dr). "QBTO/Dr" refers to the turnovers by only the quarterback, which include interceptions and lost fumbles. "PEN +/-" is a per-drive measure of the net difference in penalty yards for the quarterback's offense. So if an offense drew an encroachment penalty and had a 10-yard holding penalty on the same drive, the net would be -5.0 for that drive. Active players are in bold and the average of the sample is included at the bottom.

Playoff Drive Results
Quarterback Drives Games Yds/Dr Pts/Dr Punts/Dr QBTO/Dr PEN +/- LOS/Dr
Colin Kaepernick 60 6 41.48 (1) 2.75 (1) .300 (1) .100 (12) 1.07 (1) 27.20 (31)
Peyton Manning 229 22 36.31 (2) 2.14 (10) .323 (4) .105 (14) 0.42 (9) 27.34 (30)
Aaron Rodgers 94 9 35.86 (3) 2.48 (2) .404 (20) .085 (7) 0.82 (3) 27.37 (29)
Drew Brees 130 11 35.54 (4) 2.35 (6) .315 (3) .069 (4) -0.22 (21) 27.51 (28)
Steve Young 141 15 34.63 (5) 2.46 (3) .333 (8) .128 (25) 0.38 (11) 31.44 (17)
Russell Wilson 42 4 34.14 (6) 2.33 (7) .333 (7) .048 (1) -0.74 (31) 34.81 (1)
Kurt Warner 144 13 33.65 (7) 2.35 (5) .354 (9) .118 (21) -0.69 (30) 32.69 (6)
Troy Aikman 164 16 32.81 (8) 2.44 (4) .329 (6) .116 (20) -0.31 (24) 33.12 (5)
Joe Montana 248 23 32.41 (9) 2.15 (9) .395 (16) .089 (9) -0.27 (23) 31.62 (15)
Warren Moon 107 10 32.37 (10) 1.87 (18) .308 (2) .168 (31) -0.79 (32) 28.01 (27)
Tom Brady 282 26 32.13 (11) 2.12 (12) .408 (21) .089 (8) -0.02 (17) 30.80 (20)
John Elway 239 22 32.03 (12) 2.13 (11) .385 (14) .109 (16) 0.63 (5) 31.79 (14)
Neil O'Donnell 78 7 30.92 (13) 1.85 (19) .410 (23) .115 (19) -0.38 (26) 33.33 (2)
Jim Kelly 195 17 30.89 (14) 2.08 (13) .328 (5) .159 (30) 0.58 (6) 33.18 (4)
Ben Roethlisberger 154 14 30.82 (15) 2.23 (8) .364 (12) .123 (22) 0.19 (15) 31.82 (13)
Matt Ryan 53 5 30.72 (16) 1.74 (26) .358 (11) .189 (32) 0.96 (2) 25.72 (32)
Quarterback Drives Games Yds/Dr Pts/Dr Punts/Dr QBTO/Dr PEN +/- LOS/Dr
Bernie Kosar 93 8 30.60 (17) 1.98 (15) .387 (15) .108 (15) 0.44 (8) 29.14 (25)
Jeff Garcia 64 6 29.98 (18) 1.89 (17) .438 (25) .125 (24) 0.25 (14) 29.92 (24)
Brett Favre 274 24 29.97 (19) 2.03 (14) .358 (10) .128 (26) -0.04 (18) 33.22 (3)
Eli Manning 119 11 29.88 (20) 1.80 (23) .445 (26) .076 (6) -0.22 (22) 30.12 (23)
Mark Sanchez 65 6 29.75 (21) 1.85 (20) .523 (31) .062 (2) -0.32 (25) 32.28 (8)
Philip Rivers 96 9 29.74 (22) 1.83 (22) .469 (30) .104 (13) -0.64 (28) 31.25 (18)
Jake Delhomme 93 8 29.61 (23) 1.78 (25) .452 (28) .129 (27) -0.15 (20) 30.35 (22)
Dan Marino 205 18 28.64 (24) 1.79 (24) .380 (13) .141 (29) 0.04 (16) 30.56 (21)
Matt Hasselbeck 133 11 28.22 (25) 1.95 (16) .451 (27) .068 (3) 0.33 (12) 32.08 (10)
Tony Romo 45 4 28.09 (26) 1.49 (30) .400 (18) .089 (10) -0.64 (29) 28.64 (26)
Donovan McNabb 187 16 27.48 (27) 1.68 (27) .401 (19) .134 (28) 0.27 (13) 31.99 (11)
Joe Flacco 152 13 27.29 (28) 1.84 (21) .461 (29) .072 (5) 0.70 (4) 31.92 (12)
Steve McNair 117 10 26.97 (29) 1.54 (29) .410 (22) .111 (17) 0.57 (7) 32.34 (7)
Mark Brunell 113 11 26.42 (30) 1.56 (28) .398 (17) .124 (23) -0.52 (27) 31.16 (19)
Randall Cunningham 125 10 25.64 (31) 1.46 (31) .432 (24) .112 (18) -0.06 (19) 32.09 (9)
Dave Krieg 105 11 23.36 (32) 1.31 (32) .543 (32) .095 (11) 0.41 (10) 31.57 (16)
AVG 135.8 12.4 30.89 1.98 .394 .109 .06 30.82

Just think what Colin Kaepernick's stats looked like before his worst playoff game in the NFC Championship. There are six main stats here, and Kaepernick (four) and Russell Wilson (two) lead the way in all of them. Of course they only have 10 playoff games between them.

I have tried to only include players with at least six postseason starts, but Tony Romo (four) has been there since I first started doing this research because I assumed he would be making more appearances. That just has not happened. Wilson will get his fifth start on Sunday and Matt Ryan is the only other quarterback with five starts. I do not have a crystal ball for Romo, but we can safely assume Ryan and Wilson will start more playoff games in the future.

Kaepernick is likely entering a similar position to Aaron Rodgers after his 2010 Super Bowl run. Rodgers' playoff stats were off the charts, but have regressed following each of his last four games. Kaepernick even getting the biggest advantage from penalties (+1.07) probably has a few Carolina fans sneering right now. Fortunately, the average in penalties is almost zero, so in the rare event penalties are called in the playoffs, they have balanced out decently on offense versus defense in this study.

Last week I pondered where Denver's 60.5 yards per drive against New England would rank for the postseason. A full collection is on the to-do list so we can marvel at oddities like the Kelly Holcomb game or the time Philadelphia scored 58 points on Detroit. However, for the time being any references here to the best and worst playoff games are only based on the games in this study, but the 2013 AFC Championship is the second time a team has averaged over 60 yards per drive. Here are the top 10 studied games for points per drive and yards per drive:

Highest Points per Drive Average in Playoff Game (Starter + Min. 7 Drives)
Rk Quarterback Team Season Opp. Round Result Drives LOS/Dr Yds/Dr Pts/Dr
1 Steve Young SF 1994 CHI NFC-DIV W 44-15 7 44.00 44.29 5.29
2 Peyton Manning IND 2003 DEN AFC-WC W 41-10 8 36.38 57.38 5.13
3 Joe Montana SF 1989 DEN SB W 55-10 11 45.64 37.18 5.00
4 Peyton Manning IND 2004 DEN AFC-WC W 49-24 10 34.80 53.80 4.90
5 Peyton Manning IND 2003 at KC AFC-DIV W 38-31 8 26.75 54.88 4.75
6T Jim Kelly BUF 1990 LARD AFC-C W 51-3 9 33.78 51.00 4.56
6T Aaron Rodgers GB 2010 at ATL NFC-DIV W 48-21 9 34.67 50.56 4.56
8T Drew Brees NO 2011 DET NFC-WC W 45-28 10 27.80 65.80 4.50
8T Kurt Warner ARI 2009 GB NFC-WC W 51-45 OT 10 35.20 54.90 4.50
10 Steve Young SF 1994 SD SB W 49-26 11 35.55 43.45 4.45
Highest Yards per Drive Average in Playoff Game (Starter + Min. 7 Drives)
Rk Quarterback Team Season Opp. Round Result Drives LOS/Dr Pts/Dr Yds/Dr
1 Drew Brees NO 2011 DET NFC-WC W 45-28 10 27.80 4.50 65.80
2 Peyton Manning DEN 2013 NE AFC-C W 26-16 8 24.13 3.25 60.50
3 Bernie Kosar CLE 1987 IND AFC-DIV W 38-21 8 30.25 3.88 58.13
4 Tom Brady NE 2007 JAC AFC-DIV W 31-20 7 29.86 4.43 57.71
5 Peyton Manning IND 2003 DEN AFC-WC W 41-10 8 36.38 5.13 57.38
6 Kurt Warner ARI 2009 GB NFC-WC W 51-45 OT 10 35.20 4.50 54.90
7 Peyton Manning IND 2003 at KC AFC-DIV W 38-31 8 26.75 4.75 54.88
8 Peyton Manning IND 2004 DEN AFC-WC W 49-24 10 34.80 4.90 53.80
9 John Elway DEN 1998 MIA AFC-DIV W 38-3 8 22.63 3.88 53.63
10 Joe Montana SF 1984 MIA SB W 38-16 10 30.20 3.80 53.20

Peyton Manning makes the most appearances of any quarterback while Joe Montana and Steve Young present the only Super Bowl performances. It's unlikely Manning will add another historic game to his resume on Sunday given Seattle's defense, but in Super Bowl XLIV against the Saints, he did lead the Colts to 52.0 yards per drive -- the second-highest average by a losing team in the study and the highest in any non-Wild Card game.

Speaking of the Super Bowl, Wilson has just two turnovers on 42 playoff drives, including a Hail Mary interception in his playoff loss in Atlanta last year. That is Ryan's only playoff win and no quarterback has turned the ball over more frequently than Ryan in the playoffs (10 times on 53 drives).

Ryan has plenty of time to write his legacy. A retired Dave Krieg continues to bring up the rear in drive stats for points, yards and most punts. His offense also went three-and-out more frequently than anyone.

Three-and-Out Playoff Drives
Quarterback Drives 3Outs 3Outs/Dr Avg. Lead
Troy Aikman 164 25 .152 (1) 1.40 (14)
Tony Romo 45 7 .156 (2) 5.29 (3)
Ben Roethlisberger 154 27 .175 (3) 2.52 (10)
John Elway 239 42 .176 (4) 0.10 (20)
Drew Brees 130 23 .177 (5) 2.26 (11)
Kurt Warner 144 26 .181 (6) 5.12 (4)
Warren Moon 107 20 .187 (7) 0.10 (19)
Brett Favre 274 52 .190 (8) 4.13 (5)
Russell Wilson 42 8 .190 (9) 4.13 (6)
Neil O'Donnell 78 15 .192 (10) 7.67 (2)
Steve Young 141 28 .199 (11) 0.50 (18)
Jim Kelly 195 39 .200 (12) 2.74 (9)
Peyton Manning 229 46 .201 (13) -0.09 (21)
Randall Cunningham 125 26 .208 (14) -6.46 (31)
Aaron Rodgers 94 20 .213 (15) 7.80 (1)
Joe Montana 248 53 .214 (16) 1.42 (13)
Quarterback Drives 3Outs 3Outs/Dr Avg. Lead
Colin Kaepernick 60 13 .217 (17) -0.69 (27)
Donovan McNabb 187 41 .219 (18) -0.12 (22)
Steve McNair 117 26 .222 (19) 0.81 (16)
Dan Marino 205 46 .224 (20) -6.17 (29)
Eli Manning 119 27 .227 (21) 1.56 (12)
Philip Rivers 96 22 .229 (22) 1.36 (15)
Jake Delhomme 93 22 .237 (23) -0.41 (24)
Bernie Kosar 93 22 .237 (23) -0.50 (25)
Tom Brady 282 69 .245 (25) 3.33 (7)
Matt Ryan 53 13 .245 (26) -8.23 (32)
Mark Brunell 113 29 .257 (27) -0.62 (26)
Mark Sanchez 65 17 .262 (28) 0.65 (17)
Jeff Garcia 64 17 .266 (29) -5.47 (28)
Matt Hasselbeck 133 36 .271 (30) -0.22 (23)
Joe Flacco 152 42 .276 (31) 2.90 (8)
Dave Krieg 105 32 .305 (32) -6.31 (30)

Note: Drives where an offense ran three plays, gained zero first downs and kicked a field goal because they started in ideal field position count as a three-and-out drive here.

Ryan's three-and-out drives came with an average deficit of 8.23 points, so that's definitely an issue. Compare that to Rodgers, who may only be average at avoiding three-and-out drives -- Green Bay started with three in a row this January against San Francisco -- but has led by an average of 7.80 points when it's happened. Leading teams are less likely to play aggressive and will try to literally run out the clock.

Not surprisingly, of the 12 quarterbacks with a deficit for their average three-and-out drive, only three have a winning record in the playoffs. The five with a deficit of more than five points are a combined 18-32 (.360).

The 2011 Ravens went three-and-out eight times against Houston in a 20-13 AFC Divisional win. That's the most three-and-out drives by one team in any of the games studied here. Baltimore averaged just 16.43 yards per drive, but that's not even the worst performance of moving the ball in a playoff win. The 2005 Redskins behind Mark Brunell beat Tampa Bay on the road despite only averaging 10.9 yards per drive on 10 drives. Only two drives gained more than nine yards. That game was especially brutal for me to watch given my furnace failed that weekend and I was without heat. The Redskins were offense-less behind Joe Gibbs 2.0, but Tampa Bay had Chris Simms at quarterback, so there's usually a good explanation for unimpressive victories.

Punts are the most common way for a drive to end, but here's a handy reference chart that summarizes how each quarterback's drives have ended in the playoffs (11 options). "Miss" is a missed field goal. "BOTFG" is a botched field goal that means a bad snap or hold never allowed the kick to happen. "Clock" is a drive that ended because the clock expired at the end of a half. "FUM-TM" and "FUM-QB" are lost fumbles split up by teammates (TM) and the quarterback himself. "Downs" always means a failed fourth-down conversion.

Playoff Drive Results
Quarterback Drives Punts FG Miss BOTFG Clock INT FUM-TM FUM-QB Safety Downs TD
Tom Brady 282 115 40 6 0 9 22 6 3 2 11 68
Brett Favre 274 98 34 13 0 6 30 9 5 1 13 65
Joe Montana 248 98 25 14 0 4 21 12 1 0 7 66
John Elway 239 92 28 12 0 7 21 5 5 2 6 61
Peyton Manning 229 74 41 7 0 10 22 8 2 1 12 52
Dan Marino 205 78 19 9 0 3 24 11 5 1 11 44
Jim Kelly 195 64 29 8 0 5 28 5 3 0 7 46
Donovan McNabb 187 75 26 5 0 9 17 6 8 1 6 34
Troy Aikman 164 54 22 5 0 2 17 5 2 0 9 48
Ben Roethlisberger 154 56 19 2 0 5 17 6 2 2 4 41
Joe Flacco 152 70 19 1 0 4 8 7 3 1 7 32
Kurt Warner 144 51 15 8 1 3 14 3 3 0 4 42
Steve Young 141 47 14 3 0 7 13 4 5 0 4 44
Matt Hasselbeck 133 60 19 3 0 3 9 2 0 0 8 29
Drew Brees 130 41 17 4 0 9 6 5 3 1 8 36
Randall Cunningham 125 54 16 4 0 8 9 5 5 0 5 19
Quarterback Drives Punts FG Miss BOTFG Clock INT FUM-TM FUM-QB Safety Downs TD
Eli Manning 119 53 18 6 0 4 8 1 1 1 4 23
Steve McNair 117 48 16 8 0 2 11 7 2 0 4 19
Mark Brunell 113 45 17 2 0 5 11 3 3 1 7 19
Warren Moon 107 33 16 7 1 3 14 4 4 0 3 22
Dave Krieg 105 57 9 3 0 4 9 3 1 0 3 16
Philip Rivers 96 45 10 7 0 2 9 1 1 0 0 21
Aaron Rodgers 94 38 10 2 0 2 5 4 3 0 1 29
Bernie Kosar 93 36 10 5 0 4 10 5 0 0 1 22
Jake Delhomme 93 42 15 2 0 1 10 1 2 0 1 19
Neil O'Donnell 78 32 13 1 1 0 8 2 1 0 5 15
Mark Sanchez 65 34 5 3 0 3 3 0 1 0 1 15
Jeff Garcia 64 28 7 1 0 3 7 0 1 0 3 14
Colin Kaepernick 60 18 11 1 0 1 5 2 1 0 2 19
Matt Ryan 53 19 5 0 0 2 7 1 3 1 4 11
Tony Romo 45 18 6 2 0 3 2 1 2 1 3 7
Russell Wilson 42 14 9 0 0 1 1 2 1 0 4 10
AVG 135.8 38.8% 12.9% 3.5% 0.1% 3.1% 9.2% 3.1% 1.9% 0.4% 3.9% 23.2%

From there we can compile some interesting numbers like how well each quarterback's kickers have done (splits by distance on the to-do list) and the breakdown of lost fumble luck:

Playoff Field Goal Percentage Playoff Fumble Rates
Rk Quarterback FGM FGA FG% Rk Quarterback Drives FUM-TM FUM-QB FUM-TM% FUM/Dr
1 Russell Wilson 9 9 100.0% 1 Matt Hasselbeck 133 2 0 100.0% 1.50%
2 Matt Ryan 5 5 100.0% 2 Mark Sanchez 65 0 1 0.0% 1.54%
3 Joe Flacco 19 20 95.0% 3 Jeff Garcia 64 0 1 0.0% 1.56%
4 Neil O'Donnell 13 14 92.9% 4 Eli Manning 119 1 1 50.0% 1.68%
5 Colin Kaepernick 11 12 91.7% 5 Philip Rivers 96 1 1 50.0% 2.08%
6 Ben Roethlisberger 19 21 90.5% 6 Tom Brady 282 6 3 66.7% 3.19%
7 Mark Brunell 17 19 89.5% 7 Jake Delhomme 93 1 2 33.3% 3.23%
8 Jake Delhomme 15 17 88.2% 8 Dave Krieg 105 3 1 75.0% 3.81%
9 Jeff Garcia 7 8 87.5% 9 Neil O'Donnell 78 2 1 66.7% 3.85%
10 Tom Brady 40 46 87.0% 10 Jim Kelly 195 5 3 62.5% 4.10%
11 Matt Hasselbeck 19 22 86.4% 11 Kurt Warner 144 3 3 50.0% 4.17%
12 Peyton Manning 41 48 85.4% 12 John Elway 239 5 5 50.0% 4.18%
13 Donovan McNabb 26 31 83.9% 13 Troy Aikman 164 5 2 71.4% 4.27%
14 Aaron Rodgers 10 12 83.3% 14 Peyton Manning 229 8 2 80.0% 4.37%
15 Steve Young 14 17 82.4% 15 Colin Kaepernick 60 2 1 66.7% 5.00%
16 Troy Aikman 22 27 81.5% 16 Brett Favre 274 9 5 64.3% 5.11%
Rk Quarterback FGM FGA FG% Rk Quarterback Drives FUM-TM FUM-QB FUM-TM% FUM/Dr
17 Drew Brees 17 21 81.0% 17 Ben Roethlisberger 154 6 2 75.0% 5.19%
18 Randall Cunningham 16 20 80.0% 18 Joe Montana 248 12 1 92.3% 5.24%
19 Jim Kelly 29 37 78.4% 19 Mark Brunell 113 3 3 50.0% 5.31%
20 Eli Manning 18 24 75.0% 20 Bernie Kosar 93 5 0 100.0% 5.38%
21 Dave Krieg 9 12 75.0% 21 Drew Brees 130 5 3 62.5% 6.15%
22 Tony Romo 6 8 75.0% 22 Steve Young 141 4 5 44.4% 6.38%
23 Brett Favre 34 47 72.3% 23 Joe Flacco 152 7 3 70.0% 6.58%
24 John Elway 28 40 70.0% 24 Tony Romo 45 1 2 33.3% 6.67%
25 Warren Moon 16 23 69.6% 25 Russell Wilson 42 2 1 66.7% 7.14%
26 Dan Marino 19 28 67.9% 26 Aaron Rodgers 94 4 3 57.1% 7.45%
27 Steve McNair 16 24 66.7% 27 Warren Moon 107 4 4 50.0% 7.48%
28 Bernie Kosar 10 15 66.7% 28 Donovan McNabb 187 6 8 42.9% 7.49%
29 Kurt Warner 15 23 65.2% 29 Matt Ryan 53 1 3 25.0% 7.55%
30 Joe Montana 25 39 64.1% 30 Steve McNair 117 7 2 77.8% 7.69%
31 Mark Sanchez 5 8 62.5% 31 Dan Marino 205 11 5 68.8% 7.80%
32 Philip Rivers 10 17 58.8% 32 Randall Cunningham 125 5 5 50.0% 8.00%

Well, no missed kicks yet for Wilson and the Seahawks. Flacco's kickers being 19-of-20 must really sting for Baltimore fans when everyone knows what the one miss was. Montana would have even better numbers if his kickers didn't miss 14 times in an era of weaker kicking, but at least none of the misses ever directly cost him a game. The choking legacy of Nate Kaeding helps Philip Rivers bring up the rear.

Speaking of Montana having better numbers, it's pretty amazing that he only lost one fumble compared to 12 by his teammates. Everyone remembers the Roger Craig fumble in the 1990 NFC Championship, but Montana was already out of the game by then with an injury, so that's not one of the 12. Craig had four other lost fumbles in the playoffs with Montana behind center. As for "The Fumble" with Earnest Byner and the Cleveland Browns, Bernie Kosar joins Matt Hasselbeck as the only quarterbacks studied to not lose a playoff fumble.

Starting Field Position and Touchdowns

Here's a crucial element to success that the offense has no direct control over. While there are field position battles involving punts throughout the course of most games, once a team scores that battle is over and it goes back to the special teams to decide where the next drive (and battle) begins. An offense can hurt its defense greatly in field position with a turnover, but can also be aided by one from its own defense. It's a part of the game that gets overlooked too often, but some quarterbacks have definitely had a bigger advantage than others based on how far they had to drive to score points.

The following table shows each quarterback's average starting field position (LOS/Dr). You can see the number of drives where the offense started at exactly the 20, followed by the percentage of drives that started at the 20 or worse, their own 21-35, own 36-49 and drives that started in opponent territory. The "Worst GP" is the worst average starting field position the quarterback had in any one playoff game. The top figure in each column is in bold, while the worst figure is in red and bold.

Average Field Position in Playoffs
Quarterback Drives LOS/Dr 20 (exact) Pct. <20 Pct. 21-35 Pct. 36-49 Pct. Opp. 50+ Pct. Worst GP
Russell Wilson 42 34.81 (1) 9 21.4% 13 31.0% 12 28.6% 10 23.8% 7 16.7% 29.6
Neil O'Donnell 78 33.33 (2) 5 6.4% 15 19.2% 36 46.2% 20 25.6% 7 9.0% 27.6
Brett Favre 274 33.22 (3) 23 8.4% 79 28.8% 102 37.2% 46 16.8% 47 17.2% 24.1
Jim Kelly 195 33.18 (4) 26 13.3% 55 28.2% 70 35.9% 37 19.0% 33 16.9% 21.5
Troy Aikman 164 33.12 (5) 25 15.2% 51 31.1% 57 34.8% 30 18.3% 26 15.9% 23.1
Kurt Warner 144 32.69 (6) 19 13.2% 44 30.6% 49 34.0% 29 20.1% 22 15.3% 22.9
Steve McNair 117 32.34 (7) 8 6.8% 28 23.9% 49 41.9% 25 21.4% 15 12.8% 26.3
Mark Sanchez 65 32.28 (8) 6 9.2% 17 26.2% 28 43.1% 12 18.5% 8 12.3% 26.6
Randall Cunningham 125 32.09 (9) 24 19.2% 41 32.8% 44 35.2% 25 20.0% 15 12.0% 22.4
Matt Hasselbeck 133 32.08 (10) 9 6.8% 28 21.1% 65 48.9% 25 18.8% 15 11.3% 22.1
Donovan McNabb 187 31.99 (11) 14 7.5% 51 27.3% 68 36.4% 40 21.4% 28 15.0% 20.0
Joe Flacco 152 31.92 (12) 26 17.1% 58 38.2% 44 28.9% 23 15.1% 27 17.8% 20.0
Ben Roethlisberger 154 31.82 (13) 23 14.9% 47 30.5% 56 36.4% 30 19.5% 21 13.6% 22.8
John Elway 239 31.79 (14) 35 14.6% 82 34.3% 82 34.3% 36 15.1% 39 16.3% 19.3
Joe Montana 248 31.62 (15) 36 14.5% 85 34.3% 89 35.9% 36 14.5% 38 15.3% 20.2
Dave Krieg 105 31.57 (16) 10 9.5% 33 31.4% 37 35.2% 19 18.1% 16 15.2% 22.4
Quarterback Drives LOS/Dr 20 (exact) Pct. <20 Pct. 21-35 Pct. 36-49 Pct. Opp. 50+ % Worst GP
Steve Young 141 31.44 (17) 19 13.5% 51 36.2% 43 30.5% 18 12.8% 29 20.6% 16.7
Philip Rivers 96 31.25 (18) 13 13.5% 29 30.2% 34 35.4% 17 17.7% 16 16.7% 19.0
Mark Brunell 113 31.16 (19) 8 7.1% 35 31.0% 45 39.8% 19 16.8% 14 12.4% 18.0
Tom Brady 282 30.80 (20) 37 13.1% 94 33.3% 101 35.8% 50 17.7% 37 13.1% 16.1
Dan Marino 205 30.56 (21) 24 11.7% 61 29.8% 86 42.0% 32 15.6% 26 12.7% 19.0
Jake Delhomme 93 30.35 (22) 9 9.7% 29 31.2% 35 37.6% 20 21.5% 9 9.7% 21.5
Eli Manning 119 30.12 (23) 16 13.4% 41 34.5% 44 37.0% 17 14.3% 17 14.3% 24.1
Jeff Garcia 64 29.92 (24) 6 9.4% 16 25.0% 32 50.0% 9 14.1% 7 10.9% 27.1
Bernie Kosar 93 29.14 (25) 12 12.9% 34 36.6% 37 39.8% 13 14.0% 9 9.7% 22.6
Tony Romo 45 28.64 (26) 4 8.9% 18 40.0% 12 26.7% 10 22.2% 5 11.1% 23.5
Warren Moon 107 28.01 (27) 23 21.5% 50 46.7% 35 32.7% 6 5.6% 16 15.0% 20.7
Drew Brees 130 27.51 (28) 18 13.8% 53 40.8% 48 36.9% 19 14.6% 10 7.7% 21.6
Aaron Rodgers 94 27.37 (29) 21 22.3% 38 40.4% 34 36.2% 15 16.0% 7 7.4% 17.6
Peyton Manning 229 27.34 (30) 37 16.2% 92 40.2% 86 37.6% 35 15.3% 16 7.0% 15.7
Colin Kaepernick 60 27.20 (31) 19 31.7% 37 61.7% 10 16.7% 5 8.3% 8 13.3% 22.1
Matt Ryan 53 25.72 (32) 14 26.4% 24 45.3% 18 34.0% 7 13.2% 4 7.5% 21.7
AVG 135.8 30.82 18.1 13.9% 44.7 33.5% 49.6 36.3% 23 17.1% 18.6 13.2% 21.8

Kaepernick's placement near the bottom is a direct result of the kickoff rule change in 2011 that has created many more touchbacks and worse field position overall. Since all six of Kaepernick's playoff starts have been since last year, it's no surprise to see his numbers be what they are. Then again, that does little to explain Wilson having the best field position, other than it being a tiny four-game sample and a sign the Seahawks are really good on defense and special teams. Wilson's never had a worse game than getting to start at the 29.64, which fittingly was his only playoff loss (2012 Falcons). Matt Ryan only started at his own 25.10 that day, but he's quickly grown used to bad field position in big games.

Still, no one can take the crown for notoriously bad field position in the playoffs from Peyton Manning. He's ranked 30th despite only playing three of his 22 games in the post-2011 kickoff era. He's had a study-low 7.0 percent of his drives start in opponent territory. The average is 13.2 percent. Not only is Manning's average field position of 15.67 against the 2008 Chargers the worst for any studied playoff game, his 16.63 mark in Super Bowl XLIV is the third worst. Tom Brady's 16.11 in Super Bowl XLVI rests between the two Manning games and this era's greatest quarterbacks went 0-3 in those games with 17 points on the board in all three.

The correlation coefficient between LOS/Dr and Pts/Dr may only be 0.306 (based only on this data), but no one can deny the increased difficulty associated with having to drive long fields for points against the best competition in the game.

When breaking down the touchdown drives each quarterback has led in the playoffs, those differences in field position start to show up more.

Playoff Touchdown Drive Splits
Quarterback Drives TD TD/Dr Avg. Length 70+ Pct. 80+ Pct. <50 Pct. <20 Pct. Avg. Lead
Tony Romo 45 7 .156 (30) 72.7 5 71.4% 4 57.1% 1 14.3% 1 14.3% 2.86 (4)
Peyton Manning 229 52 .227 (16) 72.2 36 69.2% 22 42.3% 3 5.8% 1 1.9% 1.90 (5)
Jeff Garcia 64 14 .219 (17) 71.4 9 64.3% 6 42.9% 1 7.1% 0 0.0% -7.21 (32)
Aaron Rodgers 94 29 .309 (3) 71.0 18 62.1% 14 48.3% 1 3.4% 0 0.0% -2.38 (20)
Philip Rivers 96 21 .219 (18) 70.7 14 66.7% 5 23.8% 3 14.3% 0 0.0% -4.43 (29)
Colin Kaepernick 60 19 .317 (1) 69.4 13 68.4% 11 57.9% 4 21.1% 1 5.3% -4.26 (26)
Warren Moon 107 22 .206 (23) 68.5 14 63.6% 10 45.5% 5 22.7% 1 4.5% -1.55 (16)
Drew Brees 130 36 .277 (6) 68.4 22 61.1% 13 36.1% 6 16.7% 1 2.8% -3.00 (22)
Eli Manning 119 23 .193 (25) 67.3 12 52.2% 8 34.8% 3 13.0% 1 4.3% 0.13 (11)
Bernie Kosar 93 22 .237 (13) 67.0 12 54.5% 9 40.9% 4 18.2% 1 4.5% -3.50 (25)
Jake Delhomme 93 19 .204 (24) 66.2 9 47.4% 5 26.3% 2 10.5% 1 5.3% -2.37 (19)
Matt Ryan 53 11 .208 (22) 65.9 6 54.5% 4 36.4% 2 18.2% 0 0.0% -1.27 (14)
Steve McNair 117 19 .162 (29) 64.8 7 36.8% 1 5.3% 1 5.3% 0 0.0% -3.26 (23)
Brett Favre 274 65 .237 (12) 64.6 36 55.4% 16 24.6% 12 18.5% 4 6.2% -1.22 (13)
Russell Wilson 42 10 .238 (11) 64.3 4 40.0% 2 20.0% 1 10.0% 0 0.0% -6.80 (31)
John Elway 239 61 .255 (9) 63.7 28 45.9% 21 34.4% 14 23.0% 3 4.9% 0.08 (12)
Quarterback Drives TD TD/Dr Avg. Length 70+ Pct. 80+ Pct. <50 Pct. <20 Pct. Avg. Lead
Dan Marino 205 44 .215 (20) 63.0 20 45.5% 9 20.5% 10 22.7% 2 4.5% -4.41 (28)
Tom Brady 282 68 .241 (10) 62.9 31 45.6% 13 19.1% 13 19.1% 3 4.4% 0.97 (9)
Kurt Warner 144 42 .292 (5) 61.7 19 45.2% 8 19.0% 9 21.4% 2 4.8% 0.95 (10)
Mark Brunell 113 19 .168 (28) 61.7 9 47.4% 3 15.8% 3 15.8% 2 10.5% -1.53 (15)
Dave Krieg 105 16 .152 (31) 61.2 5 31.3% 3 18.8% 3 18.8% 1 6.3% -4.31 (27)
Joe Montana 248 66 .266 (8) 60.8 34 51.5% 16 24.2% 18 27.3% 6 9.1% 4.15 (1)
Donovan McNabb 187 34 .182 (27) 60.2 15 44.1% 7 20.6% 9 26.5% 3 8.8% -2.29 (17)
Randall Cunningham 125 19 .152 (32) 60.1 10 52.6% 5 26.3% 5 26.3% 2 10.5% -6.47 (30)
Mark Sanchez 65 15 .231 (15) 59.8 6 40.0% 3 20.0% 5 33.3% 1 6.7% -2.33 (18)
Ben Roethlisberger 154 41 .266 (7) 59.6 12 29.3% 5 12.2% 9 22.0% 2 4.9% -2.59 (21)
Steve Young 141 44 .312 (2) 58.8 17 38.6% 9 20.5% 17 38.6% 1 2.3% 3.84 (2)
Neil O'Donnell 78 15 .192 (26) 58.5 5 33.3% 1 6.7% 2 13.3% 1 6.7% 1.67 (6)
Jim Kelly 195 46 .236 (14) 58.4 13 28.3% 5 10.9% 12 26.1% 2 4.3% 1.48 (7)
Troy Aikman 164 48 .293 (4) 58.3 22 45.8% 10 20.8% 16 33.3% 6 12.5% 3.10 (3)
Joe Flacco 152 32 .211 (21) 57.3 14 43.8% 8 25.0% 10 31.3% 5 15.6% 1.16 (8)
Matt Hasselbeck 133 29 .218 (19) 57.0 11 37.9% 2 6.9% 8 27.6% 2 6.9% -3.34 (24)
AVG 135.8 31.5 .228 64.0 15 49.2% 8.1 27.0% 6.6 19.5% 1.8 5.4% -1.44

Manning's right near the top with his average touchdown drive traveling 72.2 yards. He's led the most 80-plus yard touchdown drives (22) in postseason history, breaking John Elway's record of 21. Believe it or not, Kaepernick is already halfway to that total in six games. He's No. 1 in touchdowns per drive. Current players will have more opportunities, but it's still impressive to watch an offense go the long way for a touchdown.

Beyond Kaepernick, San Francisco quarterbacks again create a lot of intrigue here with Montana's average touchdown drive coming with the biggest lead (4.15). Steve Young's right behind him and no quarterbacks had more drives start in opponent territory than those two. Then there's Jeff Garcia with the long scoring drives, but they came with the biggest deficit (7.21).

Seeing Mark Sanchez in between Manning and Jim Kelly in TD/Dr is confounding, but these are just drive stats. Something like Expected Points Added, which we'll look at next week, can do a much better job to gauge the individual's contribution to scoring.

Average Scoring Margin

This final table focuses on the scoreboard. The Avg. Lead is again the average score margin at the start of the drive (negative for a deficit). Then it’s broken down by the number of drives when the offense was behind, tied or ahead. "3SD-" is the number of drives that started with the offense trailing by at least three scores (17-plus points since 1994 or 15-plus prior). "3SL+" is the number of drives that started with the offense holding at least a three-score lead.

Average Score Margin
in Playoff Drives
Quarterback Drives Avg. Lead Behind Pct. Tied Pct. Ahead Pct. 3SD- Pct. 3SL+ Pct.
Neil O'Donnell 78 3.26 24 30.8% 10 12.8% 44 56.4% 1 1.3% 8 10.3%
Joe Montana 248 2.64 93 37.5% 38 15.3% 117 47.2% 7 2.8% 29 11.7%
Kurt Warner 144 2.50 47 32.6% 24 16.7% 73 50.7% 7 4.9% 16 11.1%
Jim Kelly 195 1.73 63 32.3% 48 24.6% 84 43.1% 16 8.2% 25 12.8%
Aaron Rodgers 94 1.67 37 39.4% 15 16.0% 42 44.7% 7 7.5% 9 9.6%
Joe Flacco 152 1.61 55 36.2% 33 21.7% 64 42.1% 4 2.6% 11 7.2%
Troy Aikman 164 1.59 59 36.0% 30 18.3% 75 45.7% 19 11.6% 22 13.4%
Tom Brady 282 1.49 80 28.4% 74 26.2% 128 45.4% 13 4.6% 16 5.7%
Steve Young 141 0.66 61 43.3% 23 16.3% 57 40.4% 16 11.4% 19 13.5%
Peyton Manning 229 0.64 79 34.5% 52 22.7% 98 42.8% 12 5.2% 15 6.6%
Brett Favre 274 0.18 102 37.2% 60 21.9% 112 40.9% 23 8.4% 15 5.5%
Ben Roethlisberger 154 0.14 56 36.4% 31 20.1% 67 43.5% 6 3.9% 2 1.3%
John Elway 239 -0.10 77 32.2% 67 28.0% 95 39.8% 22 9.2% 16 6.7%
Eli Manning 119 -0.57 51 42.9% 35 29.4% 33 27.7% 3 2.5% 2 1.7%
Steve McNair 117 -0.88 52 44.4% 32 27.4% 33 28.2% 1 0.9% 1 0.9%
Jake Delhomme 93 -0.96 32 34.4% 22 23.7% 39 41.9% 14 15.1% 4 4.3%
Quarterback Drives Avg. Lead Behind Pct. Tied Pct. Ahead Pct. 3SD- Pct. 3SL+ Pct.
Warren Moon 107 -1.13 42 39.3% 22 20.6% 43 40.2% 7 6.5% 3 2.8%
Tony Romo 45 -1.36 22 48.9% 10 22.2% 13 28.9% 4 8.9% 6 13.3%
Donovan McNabb 187 -1.56 89 47.6% 32 17.1% 66 35.3% 15 8.0% 4 2.1%
Russell Wilson 42 -1.60 26 61.9% 2 4.8% 14 33.3% 3 7.1% 0 0.0%
Bernie Kosar 93 -1.62 42 45.2% 29 31.2% 22 23.7% 7 7.5% 3 3.2%
Matt Hasselbeck 133 -1.77 61 45.9% 30 22.6% 42 31.6% 14 10.5% 6 4.5%
Mark Sanchez 65 -1.86 31 47.7% 13 20.0% 21 32.3% 3 4.6% 0 0.0%
Philip Rivers 96 -1.88 47 49.0% 25 26.0% 24 25.0% 6 6.3% 0 0.0%
Drew Brees 130 -2.26 69 53.1% 33 25.4% 28 21.5% 4 3.1% 5 3.9%
Colin Kaepernick 60 -2.42 30 50.0% 12 20.0% 18 30.0% 4 6.7% 0 0.0%
Mark Brunell 113 -3.29 62 54.9% 21 18.6% 30 26.6% 11 9.7% 2 1.8%
Matt Ryan 53 -3.74 27 50.9% 9 17.0% 17 32.1% 7 13.2% 0 0.0%
Dave Krieg 105 -5.62 68 64.8% 18 17.1% 19 18.1% 17 16.2% 1 1.0%
Dan Marino 205 -5.62 128 62.4% 32 15.6% 45 22.0% 48 23.4% 12 5.9%
Randall Cunningham 125 -6.58 88 70.4% 14 11.2% 23 18.4% 14 11.2% 2 1.6%
Jeff Garcia 64 -6.97 43 67.1% 13 20.3% 8 12.5% 11 17.2% 0 0.0%
AVG 135.8 -1.05 57.6 44.9% 28.4 20.3% 49.8 34.8% 10.8 8.1% 7.9 5.1%

Hard to say I was expecting Neil O'Donnell to jump to the top of the list with the largest average lead per drive in the postseason and the highest percentage of drives with a lead. With a 3-4 record as a starter, O'Donnell joins Peyton Manning as the only quarterbacks with a positive scoring margin per drive to not have a winning playoff record. O'Donnell is doing it on the back of two easy wins over the 1994 Browns and 1995 Bills in which the Steelers seized control early and never let up.

Tom Brady (28.37 percent) still has the lowest rate of drives playing from behind, but 2014 entries Bernie Kosar and Russell Wilson are the new high and low for drives in a tied game.

Three years later and the fact that still remains most staggering from this table: Dan Marino trailed by at least three scores on 23.4 percent of his postseason drives. Suddenly 8-10 with some less than stellar stats makes more sense.

Eight quarterbacks with potential to be included in next year's study, if I reduce the minimum number of drives to be considered: Andrew Luck, Alex Smith, Drew Bledsoe, Mark Rypien, Chad Pennington, Brad Johnson, Kerry Collins and Trent Dilfer. That would give us a top 40.

Drive Stats for Super Bowl XLVIII

I have intentionally given Manning and Wilson a lot of focus, but here's some more to close things out. Consider it an appetizer to the Super Bowl previews this week. The Broncos and Seahawks are perfect examples of why we need a heavier usage of drive stats. Some will look at each team and conclude the following:

  • In the playoffs, Denver is averaging 25.0 points scored per game and 16.5 points allowed per game.
  • In the regular season, Denver averaged 37.9 points scored per game and 24.9 points allowed per game.
  • The offense is down 12.9 points per game and the defense is improved by 8.4 points per game -- a net decrease of 4.5 points per game.
  • In the playoffs, Seattle is averaging 23.0 points scored per game and 16.0 points allowed per game.
  • In the regular season, Seattle averaged 26.1 points scored per game and 14.4 points allowed per game.
  • The offense is down 3.1 points per game and the defense is allowing 1.6 more points per game -- a net decrease of 4.7 points per game.
  • Each team has decreased by about the same scoring margin, but Denver's swings are more volatile.

No, things should not be worked out this way. For starters, Denver's offense averaged 12 drives per game in the regular season, but has only had the ball eight times in each playoff game and ended both games by running out the clock. The ball-control offense has worked and the scoring efficiency is still just as staggering as the regular season.

The Broncos offense is averaging 3.13 Pts/Dr in the playoffs after averaging 2.98 Pts/Dr in the regular season. While the defense has done great work for the first three quarters of each game, the defense is allowing 2.06 Pts/Dr in the playoffs after allowing 1.88 Pts/Dr in the regular season. In this regard, one can say the Denver offense has been a tad better and the defense has been a tad worse at scoring efficiency.

Seattle's games have featured more drives, so in the playoffs the Seahawks are scoring 2.19 Pts/Dr and allowing 1.39 Pts/Dr. Compared to a regular season with 2.16 Pts/Dr scored and 1.22 Pts/Dr allowed, it would appear Seattle's strength (defense) has taken more of a fall than the fictional decline from Denver's strength (offense).

Of course none of this takes into consideration the opponents faced or that pesky field position, but if the Broncos are anywhere close to 3.0 Pts/Dr on Sunday, then the Seahawks are in serious trouble, even if the Broncos only score 24 points.

Manning has averaged 1.97 Pts/Dr in his 23 full-game losses since 2007. That would be the second-worst game of the season for Denver's offense (1.71 in New England) and the third-worst game of the season for Seattle's defense after Indianapolis (3.0) and Tampa Bay (2.18).

Manning has only had one three-and-out drive in his last 35 postseason possessions. The Seahawks rank 15th at forcing three-and-out drives and have done it only four times on 23 drives this postseason. Denver has one punt on 16 playoff drives and has gained at least 37 yards on 14 of its 16 drives. The only times Denver was really stopped were a Julius Thomas fumble against San Diego and a drive that started at the San Diego 30 where no more than 30 yards could be gained. Denver gained 26 before a tipped ball was intercepted in the end zone. The offense might punt more times and go three-and-out more times in the first quarter or first half on Sunday than the last two games combined, but the Broncos should be able to move the ball effectively against Seattle.

Wilson has the lowest standard deviation in Pts/Dr in the playoffs (0.19). He's been able to exceed 2.0 Pts/Dr in each game despite varying degrees of individual success. Again, that significant field position advantage helps. Manning ranks 29th out of 32 quarterbacks in standard deviation (1.36). That's more likely to happen with some of the best games ever and quarterbacks like Montana (1.12) and Young (1.49) also rank near the bottom in standard deviation. Wilson will have to count on that consistency while Manning needs to play efficiently.

It would not be like me to spend this much time on drive stats without mentioning game-winning drives. If Wilson leads one, he'll tie Andrew Luck for the most through two seasons (11). If Manning leads one, he'll move past Dan Marino for the most in NFL history with 52. No pressure, guys.

Posted by: Scott Kacsmar on 28 Jan 2014

35 comments, Last at 30 Jan 2014, 10:04pm by nat

Comments

1
by Aaron Brooks Go... :: Tue, 01/28/2014 - 3:17pm

-redacted-

2
by Scott Kacsmar :: Tue, 01/28/2014 - 3:24pm

Does anyone have the play-by-play for the 2nd quarter of 1984 Rams-Giants playoff game? I believe that's the only thing keeping me from adding Phil Simms to this. The gamebook is missing the 2nd quarter.

17
by Travis :: Wed, 01/29/2014 - 1:51am

Let me know where to send it.

18
by Scott Kacsmar :: Wed, 01/29/2014 - 2:25am

Send it to fofantasyanswerman@gmail.com

Thanks.

3
by Chris West :: Tue, 01/28/2014 - 4:48pm

On the blue Playoff Drive Results chart, I'm surprised that Tony Romo doesn't have a BOTFG. Did his botched hold at Seattle get charted as a FUM-QB?

4
by Scott Kacsmar :: Tue, 01/28/2014 - 5:05pm

Hmm, good point. I have Romo's Seattle drive credited as Downs. Obviously I haven't had any reason to update Romo since the first time I did this, so that might be worth a change. Just wondering if the reason I gave Warner & Moon a botched FG and Romo a Downs is because Romo was the one directly responsible for that drive ending the way it did. Again, the drive stats can be done a little differently when you present them by the QB.

5
by duh :: Tue, 01/28/2014 - 5:57pm

'Manning has only had one three-and-out drive in his last 35 postseason possessions.'

In some ways I find this to be one of the most impressive stats I've seen recently.

8
by Bobman :: Tue, 01/28/2014 - 6:40pm

I THINK that was in clock-killing mode, which is even more amazing.

9
by Perfundle :: Tue, 01/28/2014 - 6:54pm

Nah, it occurred in the Ravens game last year when they were tied 28-28.

Looking through those drives, it seems Denver almost never does two straight runs to start a drive (assuming the first run didn't pick up the first down), even when they're trying to run out the clock. Of course it's Manning doing the throwing, but it's still something coaches should attempt more.

10
by Bobman :: Tue, 01/28/2014 - 6:58pm

Thanks; good catch. I thought it was two weeks ago.

6
by Bobman :: Tue, 01/28/2014 - 6:38pm

Scott, next time you are in Seattle, dinner's on me. This was fantastic.

Surprising take-aways for me: Kaep is pretty damn good! Also, he has not been well-supported by D/ST--I would not have expected that. Brady, for the most part, has been ordinary.

Non-surprises: Manning has not been well-supported by D/ST and Wilson has.

TD drives of less than 50 yards:
Brady: 16 out of 68 (24%)
Manning : 4 out of 52 (8%)
Rodgers: 1 out of 29 (3%)
Marino: 12 out of 44 (27%)
Marino: 11 out of 42 (26%)
Roehlisberger: 11 out of 41 (2%)
Brees: 7 out of 36 (19%)
Favre: 16 out of 65 (25%)
Montana: 24 out of 66 (36% that's just not freakin' fair!)

League average 25%.

7
by nat :: Tue, 01/28/2014 - 6:40pm

Drive stats are not adjusted for opponent.
More importantly, they are not adjusted for playing conditions, which are huge in the playoffs. It gets nasty outside in January.

As with your earlier article, this is going to be very skewed depending on whether a QB plays most of his playoff games inside or in the cold, rain, snow, and wind of northern outdoor stadiums. January weather is one of the things that almost everyone knows makes playoff football "special". It's why straight up comparisons of statistics are usually a bad idea when looking at teams or players in the playoffs.

Your "Average Scoring Margin" is perhaps the only table here that isn't largely spoiled by the playoff weather problem. In that table, you're indirectly comparing QBs to their opposing numbers in each game. That should remove the skew due to weather, although not perfectly.

Still, QBs (and their teams) who play many more of their playoff drives with the lead than behind are probably outplaying their opponents' QBs in whatever conditions they are playing in, no matter how high or low their other stats are in those drives.

Much of the rest of this article should be modified with the phrase "if we pretend weather doesn't affect this stat" scattered throughout. Do we really believe it as easy to lead a TD drive in a 20 degree blizzard as it is indoors with perfect conditions? Or even in a 40 degree day with 10 mph winds with gusts to 20 mph? I don't think we do.

13
by dmstorm22 :: Tue, 01/28/2014 - 9:38pm

I havent' done any of the requisite analysis, but my guess is the effect is less than you think, both on a game-by-game basis and the overall dataset.

Far fewer playoff games were played in domes than outdoors. There just aren't that many dome teams. Most of the consistent playoff team. Also, terrible playing conditions are rare. I woudl guess the correlation between temperature and offense isn't as strong as people believe. Wind and precipitation seems to be a much bigger issue.

'It gets nasty outside in January'. Well, define nasty? How many playoff games in recent years were really affected by weather conditions? The last one was probably either the 49ers @ Packers this year (no precipitation, but frigid cold) or the '11 NFC Championship Game (rain, lots and lots of rain).

I'm sure there is a study that has looked over this and possibly statistically proves I'm wrong, but I happen to think weather is an overrated variable.

14
by Scott Kacsmar :: Tue, 01/28/2014 - 9:57pm

The sample of blizzard playoff games in terrible conditions might not be any bigger than the sample of Colin Kaepernick's playoff career. I would think the Fog Bowl would rank highly on a list of playoff games impacted by weather, but didn't Randall Cunningham still somehow throw for 400 yards? Sunday will only be the 504th playoff game ever. At least a quarter of those have been played indoors. We can pretty much eliminate the pre-1978 days if we're worried about comparisons to a good offensive league. So that's not going to leave a ton of examples for truly bad weather conditions.

I feel like I know exactly how to study the weather vs. dome effect on the passing game, but it would take time and methods out of my reach right now in life. A big part of the research involves eliminating YAC and focusing on the distance of throws. That requires video analysis and game charting of countless games.

16
by nat :: Wed, 01/29/2014 - 1:33am

It doesn't need to be a blizzard. Double digit wind speeds or temps near or below freezing would be plenty.

You're just making excuses. We routinely use stats from just sixteen games. There are many more games with double-digit winds or sub-forty temps than that.

You don't need to do something exotic with YAC, either. Just look at playoff drive stats for indoor games, nice weather games, and bad weather games for each QB. Bad doesn't have to be a blizzard. Wind and/or cold should be enough.

But really, we all know that domes are easier on offenses than outdoor stadiums than outdoor stadiums in bad weather. No one hears "cold and windy" and thinks "great, a high scoring game then!" Even if they do happen sometimes.

20
by Aaron Brooks Go... :: Wed, 01/29/2014 - 10:54am

Depends how far back you go. The 1960s NFL was actually quite free-wheeling. The Packers were the exception, not the rule. They were sort of like the modern 49ers -- a contrarian, defense-first, power-running team.

22
by tuluse :: Wed, 01/29/2014 - 11:19am

Look at pass attempts, and you'll see that every team tried to be a running team in the 60s. It's just then when they passed, they passed deep.

27
by Aaron Brooks Go... :: Wed, 01/29/2014 - 7:33pm

Look before the AFL merger.

29
by tuluse :: Wed, 01/29/2014 - 8:01pm

I think you have under-estimated how pass happy the current NFL is.

The leader in pass attempts in 1963 was the Cardinals, they had 31 pass attempts per game. Last year the leader in pass attempts was the Lions. They had 46 attempts per game.

The average team had 28 pass attempts in 1963, in 2012 it was 34.

Yeah, they were more open than the dead ball 70s, but by 1980 the 49ers lead the league with 37 pass attempts per game and the average team threw 30 times per game.

21
by nat :: Wed, 01/29/2014 - 10:59am

In the last article, we found that points for and against of QBs in games against playoff teams have a -.02 correlation during the regular season and a whopping .48 correlation in the postseason. Somehow, offensive efficiency is going up or down in a kind of lockstep in the playoffs in a way that doesn't happen in the regular season.

Given that stark difference between the playoffs and the rest of the year, we can say with near certainty that the effect is real. It's caused by something that impacts both offenses in each game, but that doesn't happen much in warmer months. It also affects some QBs' games but not others. What could it be.... duh. It's the weather. What else could it be? A decades long league-wide conspiracy against some QBs' stats, but not against their winning?

It's an open question of how large of an adjustment playing most of his playoff games in great conditions should give a QB on stats like these. But we know from the last article that stats like "Completion %" are well correlated with wins against playoff teams in the regular season, but poorly correlated in the playoffs. This is true even though they remain well correlated with points scored. I'd expect the same to hold true for drive stats.

The NFL's Freakonomics feature did a small study on this topic. http://www.nfl.com/features/freakonomics/episode-12 In it you'll see that just playing outdoors causes a large drop in completion percentage from September through the year to January. Indoor playoff games have completion percentages of 61.1%. Outdoors, it's 56.4%. Roughly, that's the difference between having Tom Brady and Geno Smith taking snaps. And that's looking at all outdoors stadiums, not just bad weather ones.

That article even gives a link to a nice listing of December conditions for different NFL locals. http://www.advancednflstats.com/2007/11/weather-and-home-field-advantage... December is not playoff time. But the effect of being in the elements only gets worse in January.

The evidence is all around you. Playing outdoors harms offensive efficiency. The effect is larger the colder (and windier) it gets.

It is not overrated at all.

23
by JIPanick :: Wed, 01/29/2014 - 11:38am

"The last one was probably either the 49ers @ Packers this year (no precipitation, but frigid cold) or the '11 NFC Championship Game (rain, lots and lots of rain)."

Broncos-Ravens '12.

25
by dmstorm22 :: Wed, 01/29/2014 - 12:41pm

Yeah, I like to pretend that didn't happen. But that one's along the lines of the 49ers @ Packers this year. How much did the weather really effect the game, and how much was Manning losing McGahee, getting some terrible officiating and the Ravens playing really well?

11
by RickD :: Tue, 01/28/2014 - 7:09pm

"Tom Brady's 16.11 in Super Bowl XLVI rests between the two Manning games and this era's greatest quarterbacks went 0-3 in those games with 17 points on the board in all three."

Really a great stat there. Even the greatest QBs can only do so much.

12
by Bobman :: Tue, 01/28/2014 - 8:38pm

Or, to paraphrase that great Brazilian sports philosopher, "he can't throw AND catch the balls himself!"

Yes, I know that has nothing to do with field position, but it fit the theme.

Some day I'd like to see Manning in a football movie about himself, huddling up on the 5 yard line against the Chargers in the playoffs. "Okay, guys, I won't shit you. That $%#@ Scifers nailed us real bad and our own special teams %^$#ed us real good. Now we gotta go 95 yards, and we can do this, if only to win a ring for those d-bags who couldn't manage to catch a goddamn punt, or tackle Sproles, or block Tim HOF Dobbins. Wait, where was I?"

Wayne: "Peyton, that's great, but what's the play, man? Clock ticking."

Manning: "Hell if I know. I'm dying here, Reg, just looking at this crappy field position. My soul has shriveled to a wizened speck in the last 30 seconds. What the hell, everybody go long and I'll chuck it up. Works for my brother...."

15
by Sixknots :: Tue, 01/28/2014 - 11:58pm

Yep, I'd pay to see that movie. And Peyton could pull it off.

19
by CBPodge :: Wed, 01/29/2014 - 8:46am

A random thought before I go to actually reading it - how difficult would it be to have sortable tables for this sort of thing? Would be really helpful to have.

24
by Scott Kacsmar :: Wed, 01/29/2014 - 12:36pm

I mention this to every site I have written for. I think the only one that actually had sortable tables was the Pro-Football-Reference blog (see link at the beginning).

Hopefully that's something that will be a real possibility in the future.

26
by tuluse :: Wed, 01/29/2014 - 12:45pm

Wikipedia has sortable tables.

28
by Aaron Brooks Go... :: Wed, 01/29/2014 - 7:35pm

Wikipedia has tables where I can change the data I don't like. =)

31
by Bobman :: Thu, 01/30/2014 - 4:22am

THAT explains the Wikipedia article I read about the first president of the US being named Aaron Brooks Go..., and of course, the fine national capital named for him, and the state in the Pacific NW.

I was wondering about that.

30
by formido :: Thu, 01/30/2014 - 4:09am

I suppose it would be a problem for Seattle if Denver averaged 3 p/dr, but that doesn't seem very likely now, does it? Also, maybe it wouldn't be as big a problem as you think, since Seattle is averaging 3.3 p/dr when Percy Harvin is on the field.

32
by Bobman :: Thu, 01/30/2014 - 4:23am

Harvin on the field is the very definition of too small of a sample size.

Of course, at this stage of the season, all it takes is one game, one big performance, etc. to make history. Sample size be damned.

33
by nat :: Thu, 01/30/2014 - 10:54am

For the sixteen QBs listed both here and in the previous article, the correlation between Points/Drive and Points Against(!) is 0.29. This level of correlation does not happen in the regular season.

Since those two aren't causing each other, there is something about playing in January that is affecting offensive stats for both teams in each game. That mysterious cause is corrupting this data for the use you are trying to put it to.

That cause is bad weather and/or playing inside.

Unless you know that two QBs played in a similar mix of indoor, moderate, and bad weather locations, or somehow account for the difference in stadium types, comparing them for playoff stats is a waste of time.

34
by LionInAZ :: Thu, 01/30/2014 - 8:43pm

So prove it.

The one thing that's more annoying than someone coming on here to claim that FO is biased for/against certain teams is the one who claims an analysis is faulty for specific reasons, but can't be bothered to back up his claims with data and independent analysis. If you put as much effort into supporting your claims as you put into arguing them, you might be taken seriously.

Talk is cheap... so it is said. You shouldn't expect someone else to prove your case for you.

35
by nat :: Thu, 01/30/2014 - 10:04pm

Sigh.
I've been giving statistical evidence left and right. I also referenced an independent NFL freakonomics study on the effect of weather on passing efficiency. But I've been mostly restricting myself to Kacsmar's data, because that's what the article is about.
I really don't want to rehash this. Look at the correlations I point out above. Think about what they mean, and especially about why they might happen in January but not in the warmer months.
Something is affecting both offenses in the same game, that spans each QB's career, but depends on the time of year. It helps dome and warm locale QBs' playoff stats relative to QBs who play in more bad weather, without helping them win.
If you don't want to call it weather, call it the X factor. But the X factor is acting just like weather.