I've been wrong about a lot of Rams related things so far this year, and I'm planning on writing more about that in the near future. However, I'm going to take a gleeful look at something I was attacked for earlier this year.
May 12, 2010
I'm so down on Barron that I'd have celebrated if they'd cut him outright. Whether Carpenter becomes useful or not is almost irrelevant to me. Barron wasn't useful, and he was a major contributing factor to a lousy offensive line. They weren't going to get better while keeping him around. Smith and Saffold have a much higher ceiling than Barron...
On Sunday evening, playing at right tackle for the Cowboys, Alex Barron committed three holding penalties, including a truly hideous effort on the final play of the game, negating an apparent game winning TD pass by Tony Romo.
I have prepared the following statement, intended for Cowboys fans who considered that they won the Barron/Carpenter trade, and mocked the Rams after they cut Carpenter.
I told you so, I told you so, I so totally told you so. Neener neener neener, you have Alex Barron and we don't, ha ha ha.
Monday, September 13, 2010
Thursday, September 9, 2010
The Bottom 10
Just for the heck of it, here are the bottom 10 quarterbacks in each category.
Y/A+
117) Joe Ferguson - 94
118) Roman Gabriel - 94
119) Trent Dilfer - 94
120) Tobin Rote - 94
121) Bubby Brister - 93
122) Kordell Stewart - 92
123) David Carr - 90
124) Dan Pastorini - 88
125) Rick Mirer - 82
126) Joey Harrington - 80
TD%+
117) Jim Zorn - 93
118) Dan Pastorini - 93
119) Kordell Stewart - 92
120) Tony Banks - 92
121) Kerry Collins - 91
122) Rodney Peete - 91
123) Jack Kemp - 90
124) Joey Harrington - 89
125) David Carr - 85
126) Rick Mirer - 82
INT%+
117) Rodney Peete - 92
118) Rick Mirer - 92
119) Vinny Testaverde - 91
120) Richard Todd - 91
121) Ken Stabler - 88
122) Marc Wilson - 88
123) Trent Dilfer - 86
124) Lynn Dickey - 86
125) Mike Tomczak - 86
126) Steve Grogan - 85
Rushing Y/G
117) Bill Kenney - 1.8
118) Mark Rypien - 1.6
119) Jim Hart - 1.0
120) Steve DeBerg - 1.0
121) George Blanda - 1.0
122) Joe Namath - 1.0
123) Lynn Dickey - 0.8
124) Ken Stabler - 0.5
125) Dan Marino - 0.4
126) Norm Van Brocklin - 0.3
Sack%+
117) Daunte Culpepper - 87
118) Steve Beuerlein - 86
119) Neil Lomax - 85
120) Rick Mirer - 85
121) Rodney Peete - 84
122) Randall Cunningham - 81
123) Ben Roethlisberger - 80
124) Jim McMahon - 78
125) Greg Landry - 76
126) David Carr - 73
Dropbacks per fumble
117) Steve Beuerlein - 43.34
118) Jon Kitna - 42.66
119) Trent Dilfer - 41.73
120) Kurt Warner - 40.81
121) Dave Krieg - 37.46
122) Rodney Peete - 36.85
123) David Carr - 36.47
124) Daunte Culpepper - 36.41
125) Don Meredith - 35.92
126) Tony Banks - 35.64
Longevity
117) Elvis Grbac - 4
118) Scott Mitchell - 4
119) Doug Flutie - 4
120) Rick Mirer - 4
121) Bill Kenney - 3
122) Wade Wilson - 3
123) Bubby Brister - 3
124) Marc Wilson - 3
125) Mike Tomczak - 2
126) Erik Kramer - 2
Winning
117) Tony Banks - 2
118) Steve Beuerlein - 1
119) Bill Kenney - 1
120) Jeff Blake - 1
121) Norm Snead - 1
122) Jon Kitna - 1
123) Rick Mirer - 1
124) Archie Manning - 0
125) David Carr - 0
126) Joey Harrington - 0
Interesting that only one list has a Hall of Famer on it. And that list has four.
Y/A+
117) Joe Ferguson - 94
118) Roman Gabriel - 94
119) Trent Dilfer - 94
120) Tobin Rote - 94
121) Bubby Brister - 93
122) Kordell Stewart - 92
123) David Carr - 90
124) Dan Pastorini - 88
125) Rick Mirer - 82
126) Joey Harrington - 80
TD%+
117) Jim Zorn - 93
118) Dan Pastorini - 93
119) Kordell Stewart - 92
120) Tony Banks - 92
121) Kerry Collins - 91
122) Rodney Peete - 91
123) Jack Kemp - 90
124) Joey Harrington - 89
125) David Carr - 85
126) Rick Mirer - 82
INT%+
117) Rodney Peete - 92
118) Rick Mirer - 92
119) Vinny Testaverde - 91
120) Richard Todd - 91
121) Ken Stabler - 88
122) Marc Wilson - 88
123) Trent Dilfer - 86
124) Lynn Dickey - 86
125) Mike Tomczak - 86
126) Steve Grogan - 85
Rushing Y/G
117) Bill Kenney - 1.8
118) Mark Rypien - 1.6
119) Jim Hart - 1.0
120) Steve DeBerg - 1.0
121) George Blanda - 1.0
122) Joe Namath - 1.0
123) Lynn Dickey - 0.8
124) Ken Stabler - 0.5
125) Dan Marino - 0.4
126) Norm Van Brocklin - 0.3
Sack%+
117) Daunte Culpepper - 87
118) Steve Beuerlein - 86
119) Neil Lomax - 85
120) Rick Mirer - 85
121) Rodney Peete - 84
122) Randall Cunningham - 81
123) Ben Roethlisberger - 80
124) Jim McMahon - 78
125) Greg Landry - 76
126) David Carr - 73
Dropbacks per fumble
117) Steve Beuerlein - 43.34
118) Jon Kitna - 42.66
119) Trent Dilfer - 41.73
120) Kurt Warner - 40.81
121) Dave Krieg - 37.46
122) Rodney Peete - 36.85
123) David Carr - 36.47
124) Daunte Culpepper - 36.41
125) Don Meredith - 35.92
126) Tony Banks - 35.64
Longevity
117) Elvis Grbac - 4
118) Scott Mitchell - 4
119) Doug Flutie - 4
120) Rick Mirer - 4
121) Bill Kenney - 3
122) Wade Wilson - 3
123) Bubby Brister - 3
124) Marc Wilson - 3
125) Mike Tomczak - 2
126) Erik Kramer - 2
Winning
117) Tony Banks - 2
118) Steve Beuerlein - 1
119) Bill Kenney - 1
120) Jeff Blake - 1
121) Norm Snead - 1
122) Jon Kitna - 1
123) Rick Mirer - 1
124) Archie Manning - 0
125) David Carr - 0
126) Joey Harrington - 0
Interesting that only one list has a Hall of Famer on it. And that list has four.
Top 25 Quarterbacks by Winning
And here's the biggie. How do you quantify winning? Games won? Well, that's not really what people are generally talking about, is it? After all, Vinny Testaverde won 28 more games as a starting quarterback than Joe Namath did. Try convincing anyone that Testaverde is a bigger winner than Namath. (Not to rag on Testaverde too much. Jason Lisk did a great post over at the Football Reference blog earlier this year, arguing quite persuasively that Testaverde was more a victim of his teams than of his own talent level.)
No, what people mean when they talk about winning is championships. Dan Marino won a ton of games, including quite a few playoff games. He gets criticized for not being a winner all the time. Is it fair? Probably not. Does it reflect the way people tend to think? Sure.
I'm not going to argue that it always makes sense, but football fans as a group tend to put a lot of emphasis on winning championships, and I'm not going to fight against that. Still, there needs to be a way to differentiate between, say, Vinny Testaverde and Jim Kelly. Neither won a Super Bowl, but clearly, Kelly needs to be given credit for his accomplishments. So, I devised a simple system to do just that.
Quarterbacks are awarded points for three things: Leading their team to a winning record during the regular season, winning playoff games, and winning championships. One point is awarded for starting the majority of a team's games, and having a winning record in those games. One point is awarded for winning a postseason game. And points are awarded for championships based on the number of other teams in the league.
Look, supporters of Otto Graham are fond of arguing that in his 10 seasons, he got his team to the championship game 10 times, and won 7 titles. That's very impressive, of course. However, let's not lose sight of the fact that some of those titles came in leagues with a whopping total of 8 teams, and that he never played in a league with more than 13 teams. Should winning the 1948 AAFC championship really count as much as winning the 2004 NFL championship? Can you really make that argument with a straight face?
There's no way that, say, a 1963 AFL title should count as heavily as the 2009 NFL title, given that one comes in an 8 team league, and the other a 32 team league. However, clearly that 1963 AFL title is worth something. The San Diego Chargers beat out everyone they were asked to beat out. That it was only 7 other teams isn't their fault. They did what they could, and Tobin Rote deserves his share of credit for that. It's just not as much credit as Tom Brady deserves for the Patriots 2003 title.
So we've got a system that rewards titles, scaled by the size of the league, and rewards winning seasons and playoff wins at a lower rate. There's room to quibble, but I think the results look nice.
25) Jim McMahon - 36
24) Jeff Hostetler - 37
23) Eli Manning - 38
22) Trent Dilfer - 38
21) Joe Theismann - 38
20) Drew Brees - 39
19) Brad Johnson - 41
18) Phil Simms - 41
17) Ken Stabler - 42
16) Steve Young - 42
15) Kurt Warner - 45
14) Peyton Manning - 50
13) Brett Favre - 57
12) Bob Griese - 64
11) Johnny Unitas - 64
10) Jim Plunkett - 65
9) Roger Staubach - 71
8) Ben Roethlisberger - 75
7) Otto Graham - 76
6) John Elway - 84
5) Troy Aikman - 101
4) Bart Starr - 102
3) Tom Brady - 114
2) Terry Bradshaw - 127
1) Joe Montana - 134
Yeah, there are some guys here who aren't that highly regarded. However, good luck coming up with any kind of list of winners where Jim Plunkett isn't right up there. Is he as good a quarterback as Dan Marino? No, but he definitely deserves his credit for winning a couple of Super Bowls. Is the much maligned Trent Dilfer as good as Joe Theismann? I certainly don't think so, but if you're going to natter on about how a quarterback is good because he's a winner, don't be shocked when a guy with a 5-1 postseason record ranks higher than you might like.
No, what people mean when they talk about winning is championships. Dan Marino won a ton of games, including quite a few playoff games. He gets criticized for not being a winner all the time. Is it fair? Probably not. Does it reflect the way people tend to think? Sure.
I'm not going to argue that it always makes sense, but football fans as a group tend to put a lot of emphasis on winning championships, and I'm not going to fight against that. Still, there needs to be a way to differentiate between, say, Vinny Testaverde and Jim Kelly. Neither won a Super Bowl, but clearly, Kelly needs to be given credit for his accomplishments. So, I devised a simple system to do just that.
Quarterbacks are awarded points for three things: Leading their team to a winning record during the regular season, winning playoff games, and winning championships. One point is awarded for starting the majority of a team's games, and having a winning record in those games. One point is awarded for winning a postseason game. And points are awarded for championships based on the number of other teams in the league.
Look, supporters of Otto Graham are fond of arguing that in his 10 seasons, he got his team to the championship game 10 times, and won 7 titles. That's very impressive, of course. However, let's not lose sight of the fact that some of those titles came in leagues with a whopping total of 8 teams, and that he never played in a league with more than 13 teams. Should winning the 1948 AAFC championship really count as much as winning the 2004 NFL championship? Can you really make that argument with a straight face?
There's no way that, say, a 1963 AFL title should count as heavily as the 2009 NFL title, given that one comes in an 8 team league, and the other a 32 team league. However, clearly that 1963 AFL title is worth something. The San Diego Chargers beat out everyone they were asked to beat out. That it was only 7 other teams isn't their fault. They did what they could, and Tobin Rote deserves his share of credit for that. It's just not as much credit as Tom Brady deserves for the Patriots 2003 title.
So we've got a system that rewards titles, scaled by the size of the league, and rewards winning seasons and playoff wins at a lower rate. There's room to quibble, but I think the results look nice.
25) Jim McMahon - 36
24) Jeff Hostetler - 37
23) Eli Manning - 38
22) Trent Dilfer - 38
21) Joe Theismann - 38
20) Drew Brees - 39
19) Brad Johnson - 41
18) Phil Simms - 41
17) Ken Stabler - 42
16) Steve Young - 42
15) Kurt Warner - 45
14) Peyton Manning - 50
13) Brett Favre - 57
12) Bob Griese - 64
11) Johnny Unitas - 64
10) Jim Plunkett - 65
9) Roger Staubach - 71
8) Ben Roethlisberger - 75
7) Otto Graham - 76
6) John Elway - 84
5) Troy Aikman - 101
4) Bart Starr - 102
3) Tom Brady - 114
2) Terry Bradshaw - 127
1) Joe Montana - 134
Yeah, there are some guys here who aren't that highly regarded. However, good luck coming up with any kind of list of winners where Jim Plunkett isn't right up there. Is he as good a quarterback as Dan Marino? No, but he definitely deserves his credit for winning a couple of Super Bowls. Is the much maligned Trent Dilfer as good as Joe Theismann? I certainly don't think so, but if you're going to natter on about how a quarterback is good because he's a winner, don't be shocked when a guy with a 5-1 postseason record ranks higher than you might like.
Top 25 Quarterbacks by Longevity
There are about a thousand ways to figure longevity. Attempts? Is that really fair to quarterbacks from running eras? How about years played? Sure, but what about the George Blandas? Do you give him credit for the years he barely played? How about the years he was a regular kicker, and threw a few times on fakes? Games played. Well, some guys played in leagues where the schedule was 12 games long. Does it really make sense to give them less credit for playing a whole season?
What I'm actually going to do is give credit for the number of seasons someone was the starting quarterback for the majority of his team's games. Start 9 of 16 games? That's a season as a starter. Start 8? Sorry, doesn't count. Is that totally fair? Heck no. Is it good enough for what we're doing? Sure. Total attempts will work as the tiebreaker.
Players will be listed in ascending order, with their years as a starter behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Steve DeBerg - 11
24) Boomer Esiason - 11
23) Dave Krieg - 11
22) Kerry Collins - 11
21) Bob Griese - 12
20) Terry Bradshaw - 12
19) Troy Aikman - 12
18) Joe Montana - 12
17) Peyton Manning - 12
16) Drew Bledsoe - 12
15) Sammy Baugh - 13
14) Bart Starr - 13
13) Bobby Layne - 13
12) YA Tittle - 13
11) Ken Anderson - 13
10) John Hadl - 13
9) Jim Hart - 13
8) Dan Fouts - 13
7) Vinny Testaverde - 13
6) Johnny Unitas - 14
5) Warren Moon - 14
4) John Elway - 16
3) Dan Marino - 16
2) Fran Tarkenton - 18
1) Brett Favre - 18
There really shouldn't be any surprises here. 15 Hall of Famers out of the possible 25. You think this might be measuring something important? Naturally, anyone who sticks around for 10+ seasons as a starter has to be good. The majority of the guys who are good for that long wind up in the Hall.
What I'm actually going to do is give credit for the number of seasons someone was the starting quarterback for the majority of his team's games. Start 9 of 16 games? That's a season as a starter. Start 8? Sorry, doesn't count. Is that totally fair? Heck no. Is it good enough for what we're doing? Sure. Total attempts will work as the tiebreaker.
Players will be listed in ascending order, with their years as a starter behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Steve DeBerg - 11
24) Boomer Esiason - 11
23) Dave Krieg - 11
22) Kerry Collins - 11
21) Bob Griese - 12
20) Terry Bradshaw - 12
19) Troy Aikman - 12
18) Joe Montana - 12
17) Peyton Manning - 12
16) Drew Bledsoe - 12
15) Sammy Baugh - 13
14) Bart Starr - 13
13) Bobby Layne - 13
12) YA Tittle - 13
11) Ken Anderson - 13
10) John Hadl - 13
9) Jim Hart - 13
8) Dan Fouts - 13
7) Vinny Testaverde - 13
6) Johnny Unitas - 14
5) Warren Moon - 14
4) John Elway - 16
3) Dan Marino - 16
2) Fran Tarkenton - 18
1) Brett Favre - 18
There really shouldn't be any surprises here. 15 Hall of Famers out of the possible 25. You think this might be measuring something important? Naturally, anyone who sticks around for 10+ seasons as a starter has to be good. The majority of the guys who are good for that long wind up in the Hall.
Top 25 Quarterbacks by Fumble%
The numbers from our previous posts on quarterback fumbles don't really apply here. There, I was focused on how likely a QB was to put the ball on the ground when he got hit, in large part to see how the recently retired Kurt Warner compared to the rest of the world. Here, we're more worried about how likely a QB is to fumble in general. We're going to go with fumbles divided by (attempts + rushes) as a simple and easy way to get a general idea. However, we'll express it in terms of average number of dropbacks per fumble, because I think that conveys meaning better than a fairly narrow band of low percentages does. We'd include sacks in the calculations, but we don't have them for all players, and they actually wouldn't change the numbers all that much.
In point of fact, we're going to be measuring some of the same thing that the last list did. Quarterbacks who are sacked more tend to fumble more. The majority of QB fumbles come on sacks; a relatively small number of plays. The second largest group, by far, is botched snaps. These fumbles are not created equal. Botched snaps are recovered by the offense 75% to 80% of the time, sack fumbles, a bit less than 50% of the time. However, we don't have a breakdown of what kind of fumbles a player made, so we have to make do.
Players will be listed in ascending order, with their Dropbacks per fumble behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Carson Palmer - 77.17
24) Drew Brees - 77.42
23) Milt Plum - 77.56
22) Marc Bulger - 78.40
21) Dan Marino - 78.73
20) Jim Hart - 80.55
19) Jim Everett - 80.94
18) Elvis Grbac - 81.50
17) Matt Hasselbeck - 84.04
16) Fran Tarkenton - 85.02
15) Y.A. Tittle - 85.14
14) Daryle Lamonica - 86.47
13) Otto Graham - 86.60
12) Troy Aikman - 86.97
11) George Blanda - 88.15
10) Sonny Jurgensen - 88.88
9) Ken Anderson - 93.69
8) Joe Theismann - 94.21
7) Jim McMahon - 97.2
6) John Brodie - 98.46
5) Joey Harrington - 107.32
4) Joe Montana - 110.34
3) Joe Namath - 116.15
2) Peyton Manning - 127.04
1) Bert Jones - 174.88
Bert Jones, the paragon of football protection. In point of fact, older QBs apparently have an advantage here. At some point, the NFL switched from charging the fumble to the running back on botched handoffs, to charging the quarterback with the fumble. This leads to oddities in fumble rates at both positions. However, since some more recent quarterbacks are able to make this list, I don't think the problem is strong enough to warrant throwing out the data. If I had the whole story on when the change took place, and some time, the data could probably be normalized to account for the change in some way, but it wouldn't be perfect by any means. These are the frustrations of working with these statistics. With an army, time and money, you could do a lot better. I'm just one guy interested in trying to learn a few things.
In point of fact, we're going to be measuring some of the same thing that the last list did. Quarterbacks who are sacked more tend to fumble more. The majority of QB fumbles come on sacks; a relatively small number of plays. The second largest group, by far, is botched snaps. These fumbles are not created equal. Botched snaps are recovered by the offense 75% to 80% of the time, sack fumbles, a bit less than 50% of the time. However, we don't have a breakdown of what kind of fumbles a player made, so we have to make do.
Players will be listed in ascending order, with their Dropbacks per fumble behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Carson Palmer - 77.17
24) Drew Brees - 77.42
23) Milt Plum - 77.56
22) Marc Bulger - 78.40
21) Dan Marino - 78.73
20) Jim Hart - 80.55
19) Jim Everett - 80.94
18) Elvis Grbac - 81.50
17) Matt Hasselbeck - 84.04
16) Fran Tarkenton - 85.02
15) Y.A. Tittle - 85.14
14) Daryle Lamonica - 86.47
13) Otto Graham - 86.60
12) Troy Aikman - 86.97
11) George Blanda - 88.15
10) Sonny Jurgensen - 88.88
9) Ken Anderson - 93.69
8) Joe Theismann - 94.21
7) Jim McMahon - 97.2
6) John Brodie - 98.46
5) Joey Harrington - 107.32
4) Joe Montana - 110.34
3) Joe Namath - 116.15
2) Peyton Manning - 127.04
1) Bert Jones - 174.88
Bert Jones, the paragon of football protection. In point of fact, older QBs apparently have an advantage here. At some point, the NFL switched from charging the fumble to the running back on botched handoffs, to charging the quarterback with the fumble. This leads to oddities in fumble rates at both positions. However, since some more recent quarterbacks are able to make this list, I don't think the problem is strong enough to warrant throwing out the data. If I had the whole story on when the change took place, and some time, the data could probably be normalized to account for the change in some way, but it wouldn't be perfect by any means. These are the frustrations of working with these statistics. With an army, time and money, you could do a lot better. I'm just one guy interested in trying to learn a few things.
Wednesday, September 8, 2010
Top 25 Quarterbacks by Sack%+
Eligibility is the same, but we have a problem with sacks. They aren't kept through all of NFL history, and some of our quarterbacks have careers that took place entirely before we have any sack numbers at all, and some more have careers that span the dividing line.
What to do? Well, for those quarterbacks we don't have sack data for, we'll assume they were average. That's hardly fair, and it's not a good solution, but the alternative is to throw out the perfectly good data we have for the majority of quarterbacks on the list, or to eliminate older quarterbacks entirely. I don't like either of those options, so we'll have to go with it. For the quarterbacks we have partial data for, I've chosen one of two options, based on the percentage of their career we have data for. For those, like Jim Hart, who we have sack data for the vast majority of their career, we'll count it as if the career number applied to their whole career, rather than just part. For those like Johnny Unitas, where the opposite is true, we'll count them as being average, as we do with those for whom we have no data. It's not perfect, or even good, but the alternatives stink at least as badly.
Players will be listed in ascending order, with their Sack%+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Carson Palmer - 110
24) Bobby Hebert - 110
23) Rich Gannon - 110
22) Tom Brady - 110
21) Troy Aikman - 110
20) Kerry Collins - 110
19) Erik Kramer - 111
18) Elvis Grbac - 111
17) Billy Kilmer - 111
16) Joe Montana - 111
15) Brett Favre - 111
14) Joey Harrington - 112
13) Jeff Garcia - 112
12) Jim Everett - 112
11) Steve DeBerg - 112
10) Doug Flutie - 114
9) Mike Tomczak - 115
8) Jim Hart - 115
7) Dan Fouts - 115
6) Joe Namath - 117
5) Drew Brees - 118
4) Mark Rypien - 121
3) Peyton Manning - 122
2) Doug Williams - 125
1) Dan Marino - 125
Five Hall of Famers. This would be about what we'd expect with the usual grouping. (List of 25, 25 Hall of Famers, 126 eligible players) However, the numbers change with the sack issue. 25 of our 126 quarterbacks weren't given Sack%+ numbers, since their careers were all or mostly before sacks were recorded. And we're not drawing from a pool of 25 Hall of Famers anymore. 10 of our Hall of Famers are in the group of 26 who don't have numbers for this list, which is kind of mind blowing.
All in all, from just chance, we'd be expecting to see 3.5 Hall of Famers in this group by chance, so the 5 we do see is pretty good. It's clear that some not very good quarterbacks are in this group, like Mike Tomczak and Joey Harrington. However, the numbers are very interesting when it comes to which quarterbacks are best. Dan Marino and Peyton Manning I'd have called. Mark Rypien, I'm hardly surprised. But Doug Williams? His success isn't all from playing behind the Redskins line that Rypien succeeded him behind, either. His numbers from Tampa Bay are also phenomenal. Doug Williams was just a very, very hard man to sack.
What to do? Well, for those quarterbacks we don't have sack data for, we'll assume they were average. That's hardly fair, and it's not a good solution, but the alternative is to throw out the perfectly good data we have for the majority of quarterbacks on the list, or to eliminate older quarterbacks entirely. I don't like either of those options, so we'll have to go with it. For the quarterbacks we have partial data for, I've chosen one of two options, based on the percentage of their career we have data for. For those, like Jim Hart, who we have sack data for the vast majority of their career, we'll count it as if the career number applied to their whole career, rather than just part. For those like Johnny Unitas, where the opposite is true, we'll count them as being average, as we do with those for whom we have no data. It's not perfect, or even good, but the alternatives stink at least as badly.
Players will be listed in ascending order, with their Sack%+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Carson Palmer - 110
24) Bobby Hebert - 110
23) Rich Gannon - 110
22) Tom Brady - 110
21) Troy Aikman - 110
20) Kerry Collins - 110
19) Erik Kramer - 111
18) Elvis Grbac - 111
17) Billy Kilmer - 111
16) Joe Montana - 111
15) Brett Favre - 111
14) Joey Harrington - 112
13) Jeff Garcia - 112
12) Jim Everett - 112
11) Steve DeBerg - 112
10) Doug Flutie - 114
9) Mike Tomczak - 115
8) Jim Hart - 115
7) Dan Fouts - 115
6) Joe Namath - 117
5) Drew Brees - 118
4) Mark Rypien - 121
3) Peyton Manning - 122
2) Doug Williams - 125
1) Dan Marino - 125
Five Hall of Famers. This would be about what we'd expect with the usual grouping. (List of 25, 25 Hall of Famers, 126 eligible players) However, the numbers change with the sack issue. 25 of our 126 quarterbacks weren't given Sack%+ numbers, since their careers were all or mostly before sacks were recorded. And we're not drawing from a pool of 25 Hall of Famers anymore. 10 of our Hall of Famers are in the group of 26 who don't have numbers for this list, which is kind of mind blowing.
All in all, from just chance, we'd be expecting to see 3.5 Hall of Famers in this group by chance, so the 5 we do see is pretty good. It's clear that some not very good quarterbacks are in this group, like Mike Tomczak and Joey Harrington. However, the numbers are very interesting when it comes to which quarterbacks are best. Dan Marino and Peyton Manning I'd have called. Mark Rypien, I'm hardly surprised. But Doug Williams? His success isn't all from playing behind the Redskins line that Rypien succeeded him behind, either. His numbers from Tampa Bay are also phenomenal. Doug Williams was just a very, very hard man to sack.
Top 25 Quarterbacks by rushing Y/G.
Eligibility remains the same as on the other lists. In this case, the quarterbacks will be listed in order by their rushing yards divided by games played. This is far from a perfect way of evaluating a quarterback's contribution on the ground. Roger Staubach averaged 3.1 yards per carry, while Greg Landry averaged twice that, at 6.2. Does it make sense to give them essentially the same credit for their very similar number of yards and games? Probably not, although the question of how each passer's style affected their teams is a complicated one. In the end, the number of yards per game gained by even the highest players on the list gives you an idea of how much a quarterback's legs count compared to his arm. It would be great to get into greater detail, but it isn't so bad to have to make do.
25) Jim McMahon - 13.6
24) Mark Brunell - 13.8
23) Bert Jones - 14.0
22) Bobby Layne - 14.0
21) Rick Mirer - 14.1
20) Archie Manning - 14.5
19) David Carr - 14.6
18) Steve Grogan - 14.6
17) John Elway - 14.6
16) Fran Tarkenton - 14.9
15) Rich Gannon - 15.6
14) Jim Harbaugh - 15.7
13) Aaron Brooks - 16.5
12) Jeff Blake - 16.9
11) Jeff Garcia - 17.1
10) Roger Staubach - 17.3
9) Doug Flutie - 17.8
8) Greg Landry - 18.2
7) Tobin Rote - 21.0
6) Donovan McNabb - 22.0
5) Steve McNair - 22.3
4) Kordell Stewart - 22.8
3) Steve Young - 25.1
2) Daunte Culpepper - 25.3
1) Randall Cunningham - 30.6
Probably the placement of some of these fellows will surprise you. Fran Tarkenton's current reputation rests so much on his merit as a scrambler, that it almost seems shocking to see him this far down the list, given his status as a Hall of Famer. In fact, Tarkenton was a tremendous passer, and held most of the career passing records at his retirement, and while his scrambling was legendary, there was much more to him than that.
Beyond that, there are just four hall of famers on this list, less than we'd expect just from random chance. This is also the only list so far to have bad quarterbacks on it. Rick Mirer, anyone? Clearly, while being able to run is helpful to a quarterback, it's hardly necessary for success.
25) Jim McMahon - 13.6
24) Mark Brunell - 13.8
23) Bert Jones - 14.0
22) Bobby Layne - 14.0
21) Rick Mirer - 14.1
20) Archie Manning - 14.5
19) David Carr - 14.6
18) Steve Grogan - 14.6
17) John Elway - 14.6
16) Fran Tarkenton - 14.9
15) Rich Gannon - 15.6
14) Jim Harbaugh - 15.7
13) Aaron Brooks - 16.5
12) Jeff Blake - 16.9
11) Jeff Garcia - 17.1
10) Roger Staubach - 17.3
9) Doug Flutie - 17.8
8) Greg Landry - 18.2
7) Tobin Rote - 21.0
6) Donovan McNabb - 22.0
5) Steve McNair - 22.3
4) Kordell Stewart - 22.8
3) Steve Young - 25.1
2) Daunte Culpepper - 25.3
1) Randall Cunningham - 30.6
Probably the placement of some of these fellows will surprise you. Fran Tarkenton's current reputation rests so much on his merit as a scrambler, that it almost seems shocking to see him this far down the list, given his status as a Hall of Famer. In fact, Tarkenton was a tremendous passer, and held most of the career passing records at his retirement, and while his scrambling was legendary, there was much more to him than that.
Beyond that, there are just four hall of famers on this list, less than we'd expect just from random chance. This is also the only list so far to have bad quarterbacks on it. Rick Mirer, anyone? Clearly, while being able to run is helpful to a quarterback, it's hardly necessary for success.
Top 25 Quarterbacks by INT%+
Eligibility requirements remain the same. Players will be listed in ascending order, with their INT%+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Sonny Jurgensen - 109
24) Steve McNair - 109
23) Phil Simms - 109
22) Dan Marino - 109
21) Joe Theismann - 110
20) Len Dawson - 110
19) Greg Landry - 112
18) Tom Brady - 112
17) Fran Tarkenton - 112
16) Steve Young - 113
15) Bart Starr - 114
14) Jeff Garcia - 113
13) Rich Gannon - 114
12) Ken Anderson - 114
11) Mark Brunell - 114
10) Ken O'Brien - 116
9) Donovan McNabb - 116
8) Roger Staubach - 117
7) Neil Lomax - 117
6) Bernie Kosar - 117
5) Neil O'Donnell - 118
4) Joe Montana - 118
3) Roman Gabriel - 119
2) Sammy Baugh - 120
1) Otto Graham - 121
10 Hall of Famers on this list, like the last, but do you get a sense that the list is not quite as strong? Me too. Greg Landry, Mark Brunell, Neil O'Donnell... this supporting cast isn't quite up to the level of the non-HOFers that rounded out the previous two lists. In truth, avoiding interceptions is great, but it's not enough to make a great quarterback. Bernie Kosar is a good example. He's 6th on this list, but 98th and 110th on the other two passing lists, and he's not often brought up as a great NFL passer.
On the other hand, Joe Montana makes his highest appearance of the three lists here, coming in at #4. But the genius of Montana was that he was so very good at avoiding mistakes while at the same time making enough big plays to keep the chains moving and the scoreboard rolling. Just avoiding mistakes isn't enough. But it sure helps.
25) Sonny Jurgensen - 109
24) Steve McNair - 109
23) Phil Simms - 109
22) Dan Marino - 109
21) Joe Theismann - 110
20) Len Dawson - 110
19) Greg Landry - 112
18) Tom Brady - 112
17) Fran Tarkenton - 112
16) Steve Young - 113
15) Bart Starr - 114
14) Jeff Garcia - 113
13) Rich Gannon - 114
12) Ken Anderson - 114
11) Mark Brunell - 114
10) Ken O'Brien - 116
9) Donovan McNabb - 116
8) Roger Staubach - 117
7) Neil Lomax - 117
6) Bernie Kosar - 117
5) Neil O'Donnell - 118
4) Joe Montana - 118
3) Roman Gabriel - 119
2) Sammy Baugh - 120
1) Otto Graham - 121
10 Hall of Famers on this list, like the last, but do you get a sense that the list is not quite as strong? Me too. Greg Landry, Mark Brunell, Neil O'Donnell... this supporting cast isn't quite up to the level of the non-HOFers that rounded out the previous two lists. In truth, avoiding interceptions is great, but it's not enough to make a great quarterback. Bernie Kosar is a good example. He's 6th on this list, but 98th and 110th on the other two passing lists, and he's not often brought up as a great NFL passer.
On the other hand, Joe Montana makes his highest appearance of the three lists here, coming in at #4. But the genius of Montana was that he was so very good at avoiding mistakes while at the same time making enough big plays to keep the chains moving and the scoreboard rolling. Just avoiding mistakes isn't enough. But it sure helps.
Top 25 Quarterbacks by TD%+
The same eligibility requirements apply here, of course. Players will be listed in ascending order, with their TD%+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Ken Stabler - 109
24) Sonny Jurgensen - 109
23) Randall Cunningham - 109
22) Boomer Esiason - 109
21) Otto Graham - 110
20) Carson Palmer - 110
19) Drew Brees - 110
18) Dave Krieg - 110
17) Steve Grogan - 111
16) Joe Montana - 111
15) Roger Staubach - 112
14) Bob Griese - 112
13) Jim Kelly - 112
12) Dan Marino - 112
11) Kurt Warner - 113
10) Danny White - 114
9) Terry Bradshaw - 114
8) Brett Favre - 114
7) Ben Roethlisberger - 115
6) Daryle Lamonica - 115
5) Len Dawson - 115
4) Frank Ryan - 116
3) Tom Brady - 116
2) Peyton Manning - 119
1) Steve Young - 120
Again, trying to anticipate reactions here. Frank Ryan is #4? Really? Who the heck is Frank Ryan? Steve Grogan is popping up again? What's going on?
Well, Frank Ryan was the quarterback for the Browns in the early to mid '60s, and he made a decent little career for himself handing off to Jim Brown and Leroy Kelly. His numbers aren't that great overall, but he got to throw a relative lot of TD passes since the Browns' running game was so effective. He might have gotten on this list by circumstance, but he wasn't a bad QB, and he deserves his place.
Of the top 25, 10 are Hall of Famers, and some more will be eventually. The quarterbacks whose names are most surprising generally were from high risk, high reward offenses, which produced yards, and TDs, but also INTs, and while they do well on this list, they'll be punished on the next.
25) Ken Stabler - 109
24) Sonny Jurgensen - 109
23) Randall Cunningham - 109
22) Boomer Esiason - 109
21) Otto Graham - 110
20) Carson Palmer - 110
19) Drew Brees - 110
18) Dave Krieg - 110
17) Steve Grogan - 111
16) Joe Montana - 111
15) Roger Staubach - 112
14) Bob Griese - 112
13) Jim Kelly - 112
12) Dan Marino - 112
11) Kurt Warner - 113
10) Danny White - 114
9) Terry Bradshaw - 114
8) Brett Favre - 114
7) Ben Roethlisberger - 115
6) Daryle Lamonica - 115
5) Len Dawson - 115
4) Frank Ryan - 116
3) Tom Brady - 116
2) Peyton Manning - 119
1) Steve Young - 120
Again, trying to anticipate reactions here. Frank Ryan is #4? Really? Who the heck is Frank Ryan? Steve Grogan is popping up again? What's going on?
Well, Frank Ryan was the quarterback for the Browns in the early to mid '60s, and he made a decent little career for himself handing off to Jim Brown and Leroy Kelly. His numbers aren't that great overall, but he got to throw a relative lot of TD passes since the Browns' running game was so effective. He might have gotten on this list by circumstance, but he wasn't a bad QB, and he deserves his place.
Of the top 25, 10 are Hall of Famers, and some more will be eventually. The quarterbacks whose names are most surprising generally were from high risk, high reward offenses, which produced yards, and TDs, but also INTs, and while they do well on this list, they'll be punished on the next.
Top 25 Quarterbacks by Y/A+
A quick note, first off. In all these rankings, we'll only be considering quarterbacks who reached a threshold of 2000 career pass attempts. This is arbitrary, but necessary to keep the numbers from becoming unmanageable. In fact, there have been 126 passers who have reached 2000 attempts in NFL/AFL/AAFC history, and will be considered for these lists. In fact, a list of the top 25 passers in any category represents almost 20% of the eligible quarterbacks.
Players will be listed in ascending order, with their Y/A+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Tom Brady - 109
24) Dan Marino - 109
23) Ken Stabler - 110
22) Ken Anderson - 110
21) Johnny Unitas - 110
20) Earl Morrall - 111
19) Lynn Dickey - 111
18) Bart Starr - 111
17) Bob Griese - 111
16) Joe Montana - 111
15) Craig Morton - 112
14) Jim Kelly - 112
13) Steve Grogan - 113
12) Len Dawson - 113
11) Daunte Culpepper - 114
10) Norm Van Brocklin - 115
9) Roger Staubach - 115
8) Sammy Baugh - 115
7) Trent Green - 115
6) Dan Fouts - 116
5) Peyton Manning - 116
4) Ben Roethlisberger - 120
3) Kurt Warner - 120
2) Steve Young - 123
1) Otto Graham - 129
Some names will probably surprise people. Lynn Dickey right next to Bart Starr? That will certainly shock some Packer fans. Steve Grogan? Yes, I think that will surprise some people.
Still, there are 25 Hall of Famers who are eligible for these lists. 13 of them made the list, and the group that just missed includes more, including Joe Namath and Terry Bradshaw, who have the same 109 as Marino and Brady. (Ties are broken in favor of the player with more attempts, favoring those with greater longevity.) Of the other 14 on the list, there are a couple of slam dunk HOFers, a few more current/recent players who have a good argument, and some good QBs who didn't or won't make the Hall of Fame cut. This is the passing stat that is considered to be the most important, and it certainly looks as though we're measuring something pretty important here.
As for the surprises, well, this is why you shouldn't use just one stat when you can use many. Is Steve Grogan one of the top 25 quarterbacks of all time? Probably not, but he does very well with this particular piece of the puzzle, and that's worth knowing.
Players will be listed in ascending order, with their Y/A+ rating behind them. Hall of Famers are in bold. Players who are currently active, or were active recently enough that they are not yet eligible for the Hall are in italics.
25) Tom Brady - 109
24) Dan Marino - 109
23) Ken Stabler - 110
22) Ken Anderson - 110
21) Johnny Unitas - 110
20) Earl Morrall - 111
19) Lynn Dickey - 111
18) Bart Starr - 111
17) Bob Griese - 111
16) Joe Montana - 111
15) Craig Morton - 112
14) Jim Kelly - 112
13) Steve Grogan - 113
12) Len Dawson - 113
11) Daunte Culpepper - 114
10) Norm Van Brocklin - 115
9) Roger Staubach - 115
8) Sammy Baugh - 115
7) Trent Green - 115
6) Dan Fouts - 116
5) Peyton Manning - 116
4) Ben Roethlisberger - 120
3) Kurt Warner - 120
2) Steve Young - 123
1) Otto Graham - 129
Some names will probably surprise people. Lynn Dickey right next to Bart Starr? That will certainly shock some Packer fans. Steve Grogan? Yes, I think that will surprise some people.
Still, there are 25 Hall of Famers who are eligible for these lists. 13 of them made the list, and the group that just missed includes more, including Joe Namath and Terry Bradshaw, who have the same 109 as Marino and Brady. (Ties are broken in favor of the player with more attempts, favoring those with greater longevity.) Of the other 14 on the list, there are a couple of slam dunk HOFers, a few more current/recent players who have a good argument, and some good QBs who didn't or won't make the Hall of Fame cut. This is the passing stat that is considered to be the most important, and it certainly looks as though we're measuring something pretty important here.
As for the surprises, well, this is why you shouldn't use just one stat when you can use many. Is Steve Grogan one of the top 25 quarterbacks of all time? Probably not, but he does very well with this particular piece of the puzzle, and that's worth knowing.
All Time QB Rankings
In honor of the NFL starting out a new season tomorrow night, I'm going to do a ranking of the top 25 QBs in league history.
Now, how do you go about doing something like that? Anyone's list is ultimately going to be based on opinions, but nobody's opinion will be based on seeing the entire careers of more than a couple of players. Broad opinions are necessarily going to be based on secondhand reports, hearsay, and guesswork. This is why, while statistics are not capable of conveying the whole story, they are fundamental to any attempt to rank players, even across eras.
What goes into making up a good quarterback? Well, to get the whole picture, you need to know how good one is as a passer, clearly. Some quarterbacks make a huge contribution to their team with their legs as well, so evaluating them as runners is also very important. If you compare John Elway to Dan Marino and don't take Elway's running into consideration, you've failed miserably. While it's hard to quantify, a quarterback's qualities as a leader are also important, though impossible to quantify. Some passers are better at avoiding sacks and fumbles than others, and while they aren't a huge part of the picture, they are significant. How long a quarterback plays is also important. A player who plays at a high level for fifteen years has had a better career than one who plays for seven.
And then there's winning. While evaluating one player, even one as important as a quarterback, by a team accomplishment might seem ridiculous, it's generally a hugely important part of how people evaluate quarterbacks. And really, while many of the lists people produce purport to rank players by their ability, in practical terms, they rank them according to the quality of their actual careers, which are influenced by a myriad of other factors.
But how to compare quarterbacks across eras? A young modern fan, upon looking up the statistics of a luminary of yesteryear, say, a Bart Starr or a Roger Staubach, is unlikely to be impressed by their passing accomplishments. The game has changed. What I will be relying upon will be the "+" statistics available from the indispensable Pro Football Reference site. Staubach's completion percentage of 57.0 might not look terribly good in the context of the modern game, but his completion percentage +, or Cmp%+, of 114 tells us that he was actually well above the standards set by his contemporaries in this area.
And what categories will we use? Well, in passing, we'll use the + statistics derived from three of the four elements that go into the NFL's official passer rating.
1) Yards per attempt plus (Y/A+)
2) Touchdown percentage plus (TD%+)
3) Interception percentage plus (INT%+)
Now, why not the fourth statistic, completion percentage? Well, if you think about it, yards per attempt can be seen as a combination of two other statistics, completion percentage and yards per reception. One of the major drawbacks of the NFL's official passer rating is that it drastically overvalues completion percentage, counting it once as a full quarter of the system by itself, and then as one of the component parts of yards per attempt. Advanced methods of analyzing passing statistics always come to the same conclusion: yards per attempt is the most important simple statistic you can look at to determine the quality of a passing game, and in practice, it is the least important of the four categories in the NFL's official rating. Now, since the Football Reference site doesn't have a category for Yards Per Reception Plus, we'll simply eliminate completion percentage from our calculation and count Y/A+ more heavily.
Next, we'll cover the running ability of quarterbacks. That will be accomplished by looking at their average yards per game.
4) Rushing Yards per game.
This isn't perfect, or even close to it, but it will give us a rough estimate of how much a player's running ability impacted his team.
Next, we'll cover two often overlooked and very much entwined aspects of a quarterback's game, his fumbling, and his ability to avoid sacks. People often assume that a quarterback being sacked is the fault of his offensive line, if blame must be assigned. In fact, a quarterback's style and habits have an awful lot to do with how much he winds up being sacked. Good or bad sack percentages follow quarterbacks from year to year and team to team with more regularity than their TD% or INT%, indicating strongly that there is a real ability involved. Quarterback fumbles have been discussed in this space before. There is a wide range to the spectrum of how good quarterbacks are at holding onto the football. How likely they are to be hit is also a major factor in how often they drop the ball, so their proclivity for running or being sacked has an impact here. So, we'll have:
5) Sack%+
6) Fumble%
Then we'll also have to give quarterbacks points for winning and for longevity.
7) Winning
8) Longevity
Leadership, as we discussed, is important, but we'll have to assume that good leadership must lead to more positive results on the field, so we're already measuring it via the other categories.
I'll do lists in all eight categories, showing the top 25 in each. The tenth post in this series will compile them all, and reveal my picks for the top QBs of all time.
Now, how do you go about doing something like that? Anyone's list is ultimately going to be based on opinions, but nobody's opinion will be based on seeing the entire careers of more than a couple of players. Broad opinions are necessarily going to be based on secondhand reports, hearsay, and guesswork. This is why, while statistics are not capable of conveying the whole story, they are fundamental to any attempt to rank players, even across eras.
What goes into making up a good quarterback? Well, to get the whole picture, you need to know how good one is as a passer, clearly. Some quarterbacks make a huge contribution to their team with their legs as well, so evaluating them as runners is also very important. If you compare John Elway to Dan Marino and don't take Elway's running into consideration, you've failed miserably. While it's hard to quantify, a quarterback's qualities as a leader are also important, though impossible to quantify. Some passers are better at avoiding sacks and fumbles than others, and while they aren't a huge part of the picture, they are significant. How long a quarterback plays is also important. A player who plays at a high level for fifteen years has had a better career than one who plays for seven.
And then there's winning. While evaluating one player, even one as important as a quarterback, by a team accomplishment might seem ridiculous, it's generally a hugely important part of how people evaluate quarterbacks. And really, while many of the lists people produce purport to rank players by their ability, in practical terms, they rank them according to the quality of their actual careers, which are influenced by a myriad of other factors.
But how to compare quarterbacks across eras? A young modern fan, upon looking up the statistics of a luminary of yesteryear, say, a Bart Starr or a Roger Staubach, is unlikely to be impressed by their passing accomplishments. The game has changed. What I will be relying upon will be the "+" statistics available from the indispensable Pro Football Reference site. Staubach's completion percentage of 57.0 might not look terribly good in the context of the modern game, but his completion percentage +, or Cmp%+, of 114 tells us that he was actually well above the standards set by his contemporaries in this area.
And what categories will we use? Well, in passing, we'll use the + statistics derived from three of the four elements that go into the NFL's official passer rating.
1) Yards per attempt plus (Y/A+)
2) Touchdown percentage plus (TD%+)
3) Interception percentage plus (INT%+)
Now, why not the fourth statistic, completion percentage? Well, if you think about it, yards per attempt can be seen as a combination of two other statistics, completion percentage and yards per reception. One of the major drawbacks of the NFL's official passer rating is that it drastically overvalues completion percentage, counting it once as a full quarter of the system by itself, and then as one of the component parts of yards per attempt. Advanced methods of analyzing passing statistics always come to the same conclusion: yards per attempt is the most important simple statistic you can look at to determine the quality of a passing game, and in practice, it is the least important of the four categories in the NFL's official rating. Now, since the Football Reference site doesn't have a category for Yards Per Reception Plus, we'll simply eliminate completion percentage from our calculation and count Y/A+ more heavily.
Next, we'll cover the running ability of quarterbacks. That will be accomplished by looking at their average yards per game.
4) Rushing Yards per game.
This isn't perfect, or even close to it, but it will give us a rough estimate of how much a player's running ability impacted his team.
Next, we'll cover two often overlooked and very much entwined aspects of a quarterback's game, his fumbling, and his ability to avoid sacks. People often assume that a quarterback being sacked is the fault of his offensive line, if blame must be assigned. In fact, a quarterback's style and habits have an awful lot to do with how much he winds up being sacked. Good or bad sack percentages follow quarterbacks from year to year and team to team with more regularity than their TD% or INT%, indicating strongly that there is a real ability involved. Quarterback fumbles have been discussed in this space before. There is a wide range to the spectrum of how good quarterbacks are at holding onto the football. How likely they are to be hit is also a major factor in how often they drop the ball, so their proclivity for running or being sacked has an impact here. So, we'll have:
5) Sack%+
6) Fumble%
Then we'll also have to give quarterbacks points for winning and for longevity.
7) Winning
8) Longevity
Leadership, as we discussed, is important, but we'll have to assume that good leadership must lead to more positive results on the field, so we're already measuring it via the other categories.
I'll do lists in all eight categories, showing the top 25 in each. The tenth post in this series will compile them all, and reveal my picks for the top QBs of all time.
Subscribe to:
Posts (Atom)
