≡ Menu

These two men look important

The two best regular season quarterbacks of all time?

Yesterday, I explained the methodology behind the formula involved in ranking every quarterback season since 1960. Today, I’m going to present the career results. Converting season value to career value isn’t as simple as it might seem. Generally, we don’t want a player who was very good for 12 years to rank ahead of a quarterback who was elite for ten. Additionally, we don’t want to give significant penalties to players who struggled as rookies or hung around too long; we’re mostly concerned with the peak value of the player.

What I’ve historically done — and done here — is to give each quarterback 100% of his value or score from his best season, 95% of his score in his second best season, 90% of his score in his third best season, and so on. This rewards quarterbacks who played really well for a long time and doesn’t kill players with really poor rookie years or seasons late in their career. It also helps to prevent the quarterbacks who were compilers from dominating the top of the list. For visibility reasons, the table below displays only the top 25 quarterbacks initially, but you can change that number in the filter or click on the right arrow to see the remaining quarterbacks.1

Here’s how to read the table. Manning’s first year was in 1998, and his last in 2013. He’s had 8,740 “dropbacks” in his career, which include pass attempts, sacks, and rushing touchdowns. His career value — using the 100/95/90 formula2 is 12,769, putting him at number one. His strength of schedule has been perfectly average over his career; as a reminder, the SOS column is shown just for reference, as SOS is already incorporated into these numbers (so while Tom Brady has had a schedule that’s 0.25 ANY/A tougher than average, that’s already incorporated into his 10,063 grade). Manning is not yet eligible for the Hall of Fame, of course, but I’ve listed the HOF status of each quarterback in the table. Note that I only have quarterback records going back to 1960; therefore, for quarterbacks who played before and during (or after) 1960, only their post-1960 record is displayed. In addition, SOS adjustments are only for the years beginning in 1960.

RkQuarterbackFirst YrLast YrDBValueSOSRecordWin %HOF
1Peyton Manning199820138740127690167-730.696not el.
2Tom Brady200020136943100630.25148-430.775not el.
3Dan Marino1983199986379850-0.01147-930.613YES
4Steve Young19851999455084870.0194-490.657YES
5Fran Tarkenton19611978708982890.18124-109-60.531YES
6Joe Montana1979199457247948-0.05117-470.713YES
7Drew Brees2001201370827947-0.01110-750.595not el.
8Johnny Unitas19561973558478830.1191-50-40.641YES
9Dan Fouts19731987593677600.1286-84-10.506YES
10Ken Anderson19711986489369180.0491-810.529NO
11Norm Van Brocklin1949196030466620-0.3310-20.833YES
12Brett Favre19912010107086503-0.03186-1120.624not el.
13Otto Graham194619552861650000YES
14Roger Staubach1969197932915994-0.0185-290.746YES
15Sonny Jurgensen1957197445895891-0.267-70-70.49YES
16Aaron Rodgers2005201332055784-0.0158-290.667not el.
17John Brodie19571973476856830.3569-73-80.487NO
18Sammy Baugh193719523049563200YES
19Sid Luckman193919501752543000YES
20Len Dawson19571975412948650.0394-56-80.62YES
21Philip Rivers2004201343604843-0.0979-490.617not el.
22John Hadl19621977499647250.1282-76-90.518NO
23Tony Romo20052013399347140.0163-450.583not el.
24John Elway19831998779946870.07148-82-10.643YES
25Y.A. Tittle1948196447954655-0.3335-14-30.702YES
26Joe Namath1965197739424354-0.0462-63-40.496YES
27Daryle Lamonica1963197427804197066-16-60.784NO
28Roman Gabriel19621977487841790.0786-64-70.57NO
29Boomer Esiason1984199755304173-0.0280-930.462NO
30Jim Hart19661984537541200.0287-88-50.497NO
31Bart Starr19561971350240440.1987-41-50.673YES
32Ben Roethlisberger20042013474739600.1395-470.669not el.
33Trent Green19972008400539440.0756-570.496NO
34Warren Moon19842000730338450.03102-1010.502YES
35Donovan McNabb19992011581238270.0298-62-10.612not el.
36Kurt Warner1998200943333809-0.2166-490.574not el.
37Jeff Garcia1999200938833695-0.1158-580.5not el.
38Troy Aikman19892000498335580.0794-710.57YES
39Steve McNair1995200748353553-0.0591-620.595NO
40Cecil Isbell19381942828334900NO
41Bert Jones1973198227973258-0.1647-490.49NO
42Bobby Layne1948196239973196-0.2514-150.483YES
43Rich Gannon1987200445293172-0.0976-560.576NO
44Charlie Conerly1948196130403100-0.228-3-10.708NO
45Daunte Culpepper19992009353130000.1341-590.41not el.
46Matt Schaub20042013335729970.0446-440.511not el.
47Jim Everett19861997518429030.0364-890.418NO
48Vinny Testaverde19872007713328680.0890-123-10.423NO
49Bob Griese1967198037712835-0.1392-56-30.619YES
50Carson Palmer20042013493328120.1864-730.467not el.
51Joe Theismann19741985395927570.0677-470.621NO
52Jim Kelly1986199651092745-0.18101-590.631YES
53Randall Cunningham19852001480827350.0782-52-10.611NO
54Arnie Herber193219451177269000YES
55Tommy Thompson194019501455265300NO
56Earl Morrall1956197629602651-0.0554-24-30.685NO
57Mark Brunell1994201150452635-0.0778-730.517not el.
58Terry Bradshaw1970198342402577-0.08107-510.677YES
59Greg Landry19681984263625770.2844-51-30.464NO
60Mark Rypien19882001271825200.0847-310.603NO
61Matt Ryan20082013345024790.0960-340.638not el.
62Johnny Lujack19481951877237500NO
63George Blanda1949197541922350-0.0645-380.542YES
64Milt Plum19571969262923130.0438-32-60.539NO
65Craig Morton1965198242032243-0.1881-62-10.566NO
66Ken Stabler1970198440782220-0.2196-49-10.661NO
67Bob Waterfield194519521667217300YES
68Chad Pennington20002010264021580.0644-370.543not el.
69Billy Kilmer1961197832212156-0.3161-52-10.539NO
70Brad Johnson1994200845852143-0.0172-530.576NO
71Michael Vick2001201333532132-0.0158-48-10.547not el.
72Dave Krieg19801998581820980.1998-770.56NO
73Steve Grogan1975199038752035-0.2975-600.556NO
74Doug Williams1978198926062020-0.0438-42-10.475NO
75Bernie Kosar1985199636431978-0.1353-54-10.495NO
76Frank Ryan1958197023441959-0.0257-24-30.696NO
77Danny White19761988319918450.1362-300.674NO
78Charley Johnson1961197536681797-0.0659-57-80.508NO
79Tommy Kramer19771990390817900.1554-560.491NO
80Doug Flutie19862005226816910.0338-280.576NO
81Ed Danowski19341941641168400NO
82Phil Simms1979199351301681095-640.597NO
83Matt Hasselbeck1999201353801676-0.0880-720.526not el.
84Steve DeBerg19781998532716680.0453-86-10.382NO
85Brian Sipe1974198336741613-0.0557-550.509NO
86Jim McMahon1982199628151610-0.0567-300.691NO
87Bernie Masterson19341940415160800NO
88Neil Lomax19811988352516080.0347-52-20.475NO
89Billy Wade19541966274415560.1128-25-20.527NO
90Russell Wilson2012201388215140.3124-80.75not el.
91Ron Jaworski1974198944961502-0.0373-69-10.514NO
92Wade Wilson19811998265414730.136-330.522NO
93Bill Munson19641979217814250.2327-34-50.447NO
94Jeff Hostetler1986199725621381-0.0151-320.614NO
95Cam Newton2011201316171361-0.0725-230.521not el.
96Joe Ferguson1973199048421340-0.0879-920.462NO
97Tom Flores1960196918881215031-32-40.493NO
98Bob Monnett19341938294120200NO
99Colin Kaepernick2011201370311680.3617-60.739not el.
100Jeff George1990200143271159-0.0446-780.371NO
101Steve Bartkowski19751986382311420.0859-680.465NO
102Jim Zorn19761987337911330.0644-620.415NO
103George Ratterman194719561464113100NO
104Bobby Thomason194919571483111900NO
105Elvis Grbac1994200125811090-0.0540-300.571NO
106Neil O'Donnell1991200334921031-0.1155-450.55NO
107Bobby Hebert1985199633001024-0.0456-440.56NO
108Keith Molesworth19321936227100600NO
109Ace Parker1937194673199900YES
110Ed Brown1954196522029870.0721-21-40.5NO
111Bill Kenney198019882630983-0.1834-430.442NO
112Jay Schroeder1985199430219630.1461-380.616NO
113Don Meredith196019682581943-0.3248-33-40.588NO
114Chris Chandler198820044397929-0.1267-850.441NO
115Ray Mallouf1941194932791300NO
116Bill Nelsen196319722057905-0.3140-31-30.561NO
117James Harris196919791260893-0.0825-160.61NO
118Pat Haden1976198114848810.0635-19-10.645NO
119Erik Kramer1987199924268750.0731-360.463NO
120Eli Manning200420135264874-0.0585-660.563not el.
121Babe Parilli1952196936228700.0345-33-70.571NO
122Norm Snead196119764710863-0.1752-99-70.351NO
123Frankie Albert19461952167184800NO
124Ed Hargett196919714608010.411-5-10.214NO
125David Garrard200220102477777-0.0639-370.513not el.
126Marc Bulger200220093433707-0.1841-540.432not el.
127Virgil Carter19681976848683-0.0616-140.533NO
128Steve Beuerlein198820033665680-0.1247-550.461NO
129Tuffy Leemans1941194322963200YES
130Scott Mitchell1992200125126280.0932-390.451NO
131Jake Plummer199720064651623-0.0369-670.507NO
132Brian Griese199820082994609-0.0445-380.542NO
133Stan Humphries198919972667607-0.0450-310.617NO
134Jim Harbaugh198720004297607-0.0766-740.471NO
135Dutch Clark1936193813459900YES
136Jake Delhomme199920113107594-0.1656-400.583not el.
137Nick Foles20122013634585-0.489-70.563not el.
138Frank Filchock1938195068458000NO
139Tobin Rote195019663203578-0.0613-5-10.711NO
140Paul Governali1946194850456400NO
141Drew Bledsoe1993200671945560.0198-950.508NO
142Mike Tomczak1985199924605510.0642-310.575NO
143Glenn Presnell1932193632449200NO
144Gus Frerotte199420083319489-0.1145-47-10.489NO
145Harry Newman1933193423546700NO
146Robert Griffin20122013924465-0.3412-160.429not el.
147Gary Danielson197719882122454-0.0928-31-10.475NO
148Gene Ronzani1934194517143200NO
149Matthew Stafford200920132622427-0.0724-370.393not el.
150Don Strock19741988818423-0.0416-60.727NO
  • Perhaps the most shocking name at the top of the list is Fran Tarkenton, but he gets a big strength of schedule boost (which is consistent with Doug’s findings from years ago). Tarkenton played during some terrible years for quarterbacks, faced a brutal schedule, and played outdoors in Minnesota and New York. His numbers today may look unimpressive, but I’m not too bothered by his top-five ranking here, even if it doesn’t jive with popular opinion. He played forever, which helps, but his efficiency numbers look outstanding once you account for strength of schedule. Had Minnesota won a couple of Super Bowls, his legacy would be quite different, but his regular season numbers would be the same. He does get a bit of a boost by having so many dropbacks, but Tarkenton wasn’t really a compiler. He had three seasons with over 1,000 yards of value added, four more with 800+ yards, three more with 700+ yards, and two more with 600+ yards above average. That’s a dozen years of very good to excellent play.
  • Here’s what I said about Joe Montana last time, when he checked in at #4: “[Montana’s] reputation as the game’s best clutch passer has overshadowed how elite of a quarterback he was whenever he was on the field. His high placement validates this system and shows why football analysts need to ignore volume based passing stats in favor of refined efficiency models.” Well, Brady’s 2012 vaulted him over Montana, and Tarkenton moved up because of SOS, but Montana still sits as the 6th best regular season quarterback ever. That’s pretty darn good, and it’s before you begin to add in his 2800+ yards of postseason value added.
  • I’m not too comfortable with how this list handles older players who simply didn’t get to throw enough passes. But frankly, how do you compare Otto Graham and his 2,861 dropbacks to Drew Brees and his 7,082? I’m pretty lost, but leave your suggestions in the comments and I’ll see what I can do. Surely Brees has provided more value to his teams because well, he’s been involved in another 4000+ plays. But Graham, Sid Luckman, Norm Van Brocklin, and Sammy Baugh could rank in your top ten and I wouldn’t argue with you. Roger Staubach played in more modern times, but he barely finished with more dropbacks than Baugh. Comparing rate versus volume guys is frustrating enough in modern times (say, Matthew Stafford vs. Russell Wilson), and that difficulty is only magnified when you compare across eras.
  • Brett Favre was a great quarterback, but I have a hard time putting him in my top ten. This list seems to confirm that feeling, and that’s before putting Luckman, Baugh, and Graham ahead of him (and, I think, the better argument probably goes in favor of doing just that). He’s pretty clearly behind Dan Marino in my eyes, but I’m willing to hear counter-arguments.
  • Ken Anderson, to nobody’s surprise, ranks as the top quarterback passed over by the Hall of Fame. He even moved up from #12 to #10 this time around, after including the SOS adjustment. Anderson faced a slightly harder than average schedule, which was enough to move him ahead of Van Brocklin (note that for Brocklin, his -0.33 SOS grade consists of just one season — 19603 — but Anderson was only 16 yards behind NVB last time around) and Favre, who had a slightly easier than average schedule.
  • The Hall of Fame needs more 49ers quarterbacks like it needs more Steelers, but John Brodie is another glaring omission.
  • Tony Romo doesn’t just rank ahead of Troy Aikman, but he also has abetter career winning percentage. I bet you didn’t know that.4
  • Regarding John Elway’s 24th-place ranking, you can read some more thoughts I have on Elway here.
  • The anti-Tarkenton is Kurt Warner, who comes in at #34 on this list. Warner had one of the easiest schedules and played in a passer-friendly era, but I doubt he’ll ever make one of my top 25 lists. Frankly, he’s probably too high at 34, which ignores that he played the vast majority of his games in domes and had four Hall of Fame caliber receivers (and Marshall Faulk) on his teams.

What stands out to you? What suggestions/modifications do you have?

  1. Note that while yesterday’s list was just from 1960 to 2013, the career list reflects every season in history, using the same methodology as used in GQBOAT IV. []
  2. And including negative seasons. []
  3. In all seasons before 1960, there was no SOS adjustment. []
  4. Of course, Aikman provided slightly more value than Romo in the postseason. []
  • Arif

    What stands out to me are Joe Namath’s numbers precisely because he’d be higher on the list if the AFL data included sacks (because including rushing data doesn’t help when you count all rushing yards above 4.0 per carry).

    C’est la vie

    • Chase Stuart

      Well, I did include sack data for Namath. So his ranking reflects his great sack rate (well, for the most part; for pre-’69, it’s estimated sack rate, but that’s close enough).

      • Arif

        Oh, alright then.

        I’ll just sort by the wins column and be done with it. Ahem.

  • Nate

    Is there a way to rate-adjust dropbacks similar to how they compare ERA throughout baseball history using ERA+?

    • Chase Stuart

      There are ways to make the adjustment, but the devil is always in the details. Not sure the best way, but I’m sure some of the commenters here will think of something.

  • They way I read this, there is more of a gap between the values of #1 and #2 than the values of #2 and #9. That means to me that Manning is ranked #1 by a substantial margin.

    • Chase Stuart

      Yes, that is a pretty good takeaway.

  • Kibbles

    I know the NFL counts them, but I’m not at all a fan of including Graham’s AAFC numbers in any analysis. To me, they have every bit as many quality-of-competition concerns as the early AFL years that you hit with a penalty. Graham averaged 9.5 YPA and a better than 2:1 TD:INT ratio against the AAFC. He averaged 8.5 YPA and threw more INTs than TDs against the NFL (which was quite common of the era). His passer rating dropped by 20 points on his transition. I don’t mean to suggest he was bad- he was a deserving multiple All Pro and easily worthy of the HoF. I’m just saying that his numbers are inflated in a way that, say, Jim Kelly’s or Warren Moon’s never were because the NFL chose to recognize his stats from an inferior league and it didn’t do the same for theirs.

    • Chase Stuart

      I didn’t explicitly say it, but I did adjust Graham’s AAFC numbers. If we didn’t, I’m sure he’d be in the top ten. It looks like the adjustment I gave Graham was very similar to what I did for the early AFL days, so Graham’s ’47 and Blanda’s ’61 are pretty significantly devalued. I don’t quite remember the adjustment, because it’s the adjustment I made two years ago (I didn’t re-run any numbers for pre-1960 QBs.)

      • Chase Stuart

        On second thought, I see what I did for the AAFC. They get a modifier of 0.5, but that then gets pro-rated to a 16-game schedule. So yeah, a pretty significant deflation in value for Graham’s numbers, which I think is appropriate (0.5 may not not be the right number, but I think it’s in the ballpark).

        • Kibbles

          Thanks Chase. I figured something must be at play if Van Brocklin was coming out ahead of Graham (the two had nearly identical numbers during their NFL tenures, but obviously Graham’s AAFC stats juiced his career totals so they look substantially better than NVB’s). I think the 0.5 modifier sounds about right- intuitively, it seems to make sense that the AAFC would have been about as strong as the early AFL, since both were upstart leagues. The AFL got strong enough to eventually force a merger. The AAFC did not, but the Browns got strong enough to eventually shoe-horn their way into the league, anyway.

  • Chase, good work. In regard to your question on how to compare rate versus volume players, perhaps you should try to develop a statistic that compares a QB to a precisely average NFL QB during the years in which he played. The average QB would have a number of zero so that a positive number would reflect how much above-average a QB was during his years of play and a negative number would reflect a below-average QB. Lee Sinins has developed a similar statistic for hitters (Runs Created Against Average) and pitchers (Runs Saved Against Average) in regard to Major League Baseball and they are quite helpful in comparing players from different eras.

    • Chase Stuart


      That’s exactly what this formula does. The issue is the number of dropbacks. For example, if baseball rules changed and the average hitter had only 300 plate appearances a year, it would be impossible for a hitter to provide as many runs created against average to say, Alex Rodriguez in his prime.

      • bill

        The formula treats all games the same. I strongly disagree with this. Great or would be great quarterbacks get an a opportunity to play about 40 really important games in their career. This is regular and post season together. You have guys on this list that really suffered when it mattered most. Marino, Tarkenton, Fouts, Favre, Tittle and Romo come to mind. Tarkenton had 3 awful Super Bowls that he shouldn’t even have been in based on his other playoff performances. Marino’s passer rating dropped a full 10 points in the post season. That matters. I have watched full replays of Tittle’s 3 championship games in the early 60’s. He was terrible when it mattered most. So was Fouts. Favre may be the worst of them all when it mattered most. Hadl, Brodie and Jurgensen barely even played any real important games and were all mostly bad when they did. Fouts is defined by losing at home to Gifford Nielsen’s Oilers, not by passing for 40,000 mostly meaningless yards. Favre is defined by simply throwing away virtually every big game he played. My favorite quarterback is Bart Starr. He himself has said that the measure of a quarterback is how he plays when his team has to win. I have a degree in math and the greatness of a quarterback is not in a formula. It is in big game performances.

  • David

    Ken Anderson stands out. You mentioned that in other analyses, he comes out as a top 15-20 QB as well. So what’s up, you have to win a SB to get into the HOF? I guess not, because Dan Fouts is ranked right ahead of Anderson at #9.

    Aaron Rogers sort of stands out, in that, we’re keeping an eye on him. In a post a month or so ago, we were speculating that he’s due for a big jump in the rankings if he keeps doing what he should be doing.

    • Chase Stuart

      The best explanation for Anderson’s absence is that he just wasn’t regarded as HOF caliber during is playing days. One reason for that may have been that his best years were nonconsecutive: As you can see in yesterday’s post, his top five seasons were ’73, ’74, ’75 and then ’81 and ’82. That’s a bit like Warner, except Warner’s numbers look much better because they are in the modern era. As for why Anderson struggled from ’76 to ’80, well, Jason Lisk wrote a very good article about that: http://thebiglead.com/2012/06/21/ken-anderson-still-waiting-for-the-hall-of-fame-but-he-should-have-been-in-long-ago/

      • Kibbles

        Actually, in my mind, the best explanation for Anderson’s absence is the fact that three quarterbacks worked with Bill Walsh for several years, and those quarterbacks are ranked 4th, 6th, and 10th all-time by this formula. Which do you think is more likely- that Bill Walsh happened to coach three quarterbacks for a sustained period of time, and those three guys just happened to be among the 10 best quarterbacks in the history of the game? Or that Bill Walsh made all of those quarterbacks look better than they really were?

        Now, to me, this is a silly distinction. In my mind, the Hall of Fame doesn’t honor might-have-beens, it honors what actually happened. What actually happened is that Anderson was a star who easily outpaced most of the rest of the league. I think he greatly benefited from working with Bill Walsh in a cutting-edge offense, but I think Otto Graham greatly benefited from working with arguably the biggest football innovator in history in Paul Brown, and yet Graham still sailed into the Hall of Fame.

        Still, accusations that he’s a “system player” have also kept the equally-deserving Terrell Davis out of the Hall of Fame, and Terrell Davis has a better resume than Anderson (in my opinion). Plus, Davis was the best player to ever play in his particular system, whereas Anderson was “just” the third-best quarterback in his.

    • Richie

      SOMEBODY has to be the best QB to not make the HOF. If Anderson ever makes it, then it’s going to be Brodie or Lamonica or Rivers who keeps popping up on these lists. However, since Anderson really ends up sticking out like a sore thumb on these lists, his oversight really seems like a mistake.

      He led the league in yards in 74 and 75, and then went to the Super Bowl in 81. He won 10 more games than he lost for his career. When you look at his PFR page, his numbers don’t look great by todays standards. But they seem to be in line with what was happening in the 70s.

      I think he just kind of had bad timing and lacked “fame”. Had he played 10 years later, he probably would have had the more impressive numbers that guys like Elway and Kelly put up. Instead, he had numbers that were dwarfed shortly after he retired.

      And his contemporaries who made the HOF were:
      Montana (barely a contemporary), won many Super Bowls
      Fouts, put up huge numbers
      Bradshaw, won many Super Bowls
      Griese, won Super Bowls
      Staubach, won Super Bowls, played for America’s Team
      Tarkenton, played in many super bowls, played a long time, played for the Giants
      Namath, (barely a contemporary), was JOE NAMATH

      I think when you compare him to those guys, Anderson just kind of looks like a second-class citizen.

      Does Anderson have any comps in other positions – or other sports? A guy who is overlooked for the HOF, largely because you have to era-adjust his stats to see that he is an all-time great? A guy who played for a low-profile franchise?

      • David

        Looking at list of MLB player not in the HOF; Bagwell, Tramell. I recall that it took Jim Rice a while to get voted in. Guys like Tramell and Rice come from the same era as Anderson, funny.

        Chase, thx for the article link.

      • I think Cliff Branch, Harold Jackson, and Harold Carmichael fall in that category.

    • Andrew Healy

      I agree with the lacking fame and not being regarded as HOF caliber during his career. If Anderson plays in NYC, I think he’d get in (although Charlie Conerly ranked fairly high there and didn’t get in). Maybe if he was more well-known from college (he played at Augustana, a Div III school now), he would have been more famous to start with and that would have made a difference. Also, maybe if he had Anthony Munoz earlier, his mid-career swoon wouldn’t have happened (Munoz comes in ’80 the year before Anderson’s late career renaissance).

      And maybe there’s a bias against his early 80s mustache.

    • bill

      I put Anderson ahead of Fouts, but I put neither in the hall of fame. The hall of fame has too many guys. For quarterbacks I would take out Moon, Fouts, Tittle, Jurgensen, Namath and Tarkenton to start with. None of them were truly great. I would put Plunkett in because I like winning.

  • eag97a

    Good work Chase ! Would you add in postseason value for a total value ranking ?

    • Chase Stuart

      I think so. Now I only crunched postseason data back to 1966, so Bart Starr and Sid Luckman are undervalued here. Without thinking too much about the appropriateness of just adding the two numbers together…

      1____Peyton Manning______________12769___1619___14388
      2____Tom Brady___________________10063____921___10984
      3____Joe Montana__________________7948___2812___10760
      4____Dan Marino___________________9850____635___10485
      5____Steve Young__________________8487___1003____9490
      6____Drew Brees___________________7947___1237____9184
      7____Fran Tarkenton_______________8289_____63____8352
      8____Dan Fouts____________________7760____301____8061
      9____Johnny Unitas________________7883_____88____7971
      10___Brett Favre__________________6503___1422____7925
      11___Ken Anderson_________________6918____501____7419
      12___Roger Staubach_______________5994____705____6699
      13___Norm Van Brocklin____________6620______0____6620
      14___Aaron Rodgers________________5784____828____6612
      15___Otto Graham__________________6500______0____6500
      16___John Brodie__________________5683____243____5926
      17___Sonny Jurgensen______________5891______0____5891
      18___Sammy Baugh__________________5632______0____5632
      19___Sid Luckman__________________5430______0____5430
      20___John Elway___________________4687____710____5397
      21___Len Dawson___________________4865____494____5359
      22___Kurt Warner__________________3809___1543____5352
      23___Philip Rivers________________4843____460____5303
      24___Daryle Lamonica______________4197____872____5069
      25___Troy Aikman__________________3558___1408____4966
      26___John Hadl____________________4725______0____4725
      27___Joe Namath___________________4354____342____4696
      28___Bart Starr___________________4044____638____4682
      29___Y.A. Tittle__________________4655______0____4655
      30___Tony Romo____________________4714___-173____4541
      31___Terry Bradshaw_______________2577___1815____4392
      32___Warren Moon__________________3845____454____4299
      33___Roman Gabriel________________4179______0____4179
      34___Jim Hart_____________________4120______0____4120
      35___Ben Roethlisberger___________3960____147____4107
      36___Boomer Esiason_______________4173___-211____3962
      37___Trent Green__________________3944______0____3944
      38___Donovan McNabb_______________3827_____12____3839
      39___Jeff Garcia__________________3695_____57____3752
      40___Steve McNair_________________3553_____99____3652

      I really do need to get around to re-running the numbers for pre-SB era, but this post has some info you might enjoy: http://www.footballperspective.com/the-greatest-qb-of-all-time-iv-part-iii-playoff-results/

      • Red

        Are these values weighted by playoff round or do they all count as one game? Are they adjusted for SOS? For the life of me, I cannot figure out how Montana’s playoff value changed so much since last year. In the article you linked, Montana’s straight value was 1810 and his weighted value was 3777. But in this chart he has 2812, nowhere near either figure from last year.

      • Tanner

        I really enjoy this total data crunch, gives you more variables to punch in during debates. I’m just really curious why Peyton Manning has one of the highest “Post Season Values.”

  • Jim

    Tony Romo has had a better career than John Elway did?????back to the drawing board…

    • Chase Stuart

      Yeah, I probably should have mentioned Elway. Some thoughts on his stats here: http://www.footballperspective.com/john-elway-the-quarterback-version-of-dick-butkus/

    • David

      For the regular season: Elway only threw for 20+ td’s once in his first 10 seasons. Romo (in 8 seasons as a starter) has thrown for 30+ three times and 20+ six times. I know there some adjustment for era there, but I think that’s Romo over Elway in a nutshell.

      Remember that Elway lost 3 Super Bowls? What if he never made it back to the big one? A big what if, I know.

    • Ron

      Pretty stunning that people just can’t buy how good a QB Romo has been, even stacked against the greats. Cut through the media, the nation’s hatred for the Cowboys, and the idea that the QB is the one who wins and loses all the games — disregarding that there are 52 other players who must play well also — and you see raw talent and production that is not only good enough to win a championship, but that most NFL teams would kill for.

      Now if only his supporting cast was any good in the last 7 years….

  • mike carlson

    The question of QBs from the Graham-Baugh-Luckman era is interesting. John Maxymuk basically compared them to the average of their eras, which I would have thought at least possible with your figures. His approach confirmed that those guys (and Arnie Hebner) were way ahead of their time, but that doesn’t compare like for like across the decades. I don’t doubt, even without adjusting for growth, training, steroids and the like, that those guys would have been great playing one-platoon (or do you give Baugh points for interceptions and for his punting?) with a smaller ball in a passer-friendly environment, but I can’t see in principle any way to prove it statistically. What you do with AAFC stats, I don’t know, though they probably are as legitimate as AFL stats in that league’s first 4 years, and USFL stats are probably about the same.

    • Chase Stuart

      Thanks for dropping by, Mike.

      The real issue seems to be the number of dropbacks. We can adjust for it, but the results tend to be pretty funky. I suppose it’s worth investing some more.

  • Man Steve Young needs to be in more “greatest QB’s ever” conversations. Has the ring, the stats and some of the better single seasons in NFL history as well. I’m obviously exaggeration but I think he was much better than many give him credit for.

    • Chase Stuart

      No doubt. He gets penalized because people want to give all the credit to other people. That may be fair, but from a statistical perspective, Young was arguably the GOAT (consider his value in related to his dropbacks).

    • Kibbles

      He’s already in them. For my money, he had the most dominant peak of any quarterback in history, and if it weren’t for Peyton Manning, he’d probably be at the top of my personal all-time rankings.

  • Me

    Tony Romo got to sit on the bench and learn for years – Aikman got thrown into the fire as a rookie on a terrible team.

    • Chase Stuart

      I keep forgetting about how many advantages Romo has compared to what Aikman had to deal with.

      • David

        I know, Aikman had it rough. Bad head coach, no running game behind him, nobody to throw the ball to, bad pass protection…oh, and the owner had tight pockets!

        • Steve

          The Cowboys were 3-13 the year before Aikman got there before bottoming out at 1-15 his rookie season. Jimmy Johnson had had great success as a college HC but had zero NFL experience before taking the Cowboys’ job. On the other hand, the Cowboys were 9-7 the year before Romo threw his first pass in an NFL game that counted and his HC was someone who everybody had down as a HoF lock.

          Look, I agree that Romo’s gotten more overall criticism/hate than deserved, but the apologizing for him here is in the extreme.

      • Kibbles

        I keep looking for the “like” or “favorite” button, but I can’t seem to find it…

      • Steve

        The Cowboys were a mess when Aikman first got there and then his head coach undermined him by drafting the starting qb he had in Miami.

    • richintx

      RIGHT. Romo had the huge advantage of learning from the greats like Drew Henson, Chad Hutchinson, Quincy Carter, Drew Bledsoe, Vinny Testaverde, etc…

      • Steve

        As opposed to Aikman learning from a fellow rookie in Steve Walsh and Babe Laufenberg

  • OPBolt

    So, the way I read your list, these are, in order, the top 10 QBs in the game today:
    1 – Peyton Manning
    2 – Brady
    3 – Brees
    4 – Rogers
    5 – Rivers
    6 – Romo
    7 – Roethlisberger
    8 – Schaub
    9 – Palmer
    10- Ryan

    • Matt

      I wouldn’t read it like that. If you take a look at overall careers then yes you are right but the “current” top 10 would be a bit different with rankings skewed more towards recent production.

  • I’m sure one of the more-mathematically-educated people here could explain why this wouldn’t work (I didn’t have to take a ton of statistics as a psych major and law school shockingly does not include any math at all.), but could you add a regressed rate to the sample size of those with a smaller number of dropbacks?

    When I’m trying to write this out it sounds horribly confusing, so I’m going to try to write it as a completely theoretical example with completely made up and simplified numbers.
    Passer A has a value of 5000 over 5000 dropbacks.
    Passer B has a value of 2000 over 2000 dropbacks.
    Both were, per rates, 20% better than average.
    Can we add 3000 dropbacks of some value in between the rate we have and league average (depending on what power we do have)–say, 8% better than average, to get them to equal numbers? It would read rather like a projection at that point instead of being a straight comparison of what they actually did, but I think it at least would do something in between overvaluing the higher usage rates of recent players and overvaluing higher rates over smaller samples, and it seems like it would work. It does still give an advantage to anyone with a higher number of dropbacks because they get regressed less, but it seems to me that we would want that anyway.

    I have a feeling that in 30 years Tony Romo is going to look a lot like John Brodie does now–production that seems to merit Hall of Fame induction, a reputation that does not even merit discussion, and a weird gaping chasm in between. And incidentally, I am surprised at just how high Brodie is, but it makes sense.

    I was glad to see Tarkenton improve. He seems unfairly dismissed in GOAT discussions. (Well, to the extent that anyone can be unfairly dismissed–I really don’t see anymore how there can be a discussion. It’s like discussing who the greatest guitarist ever is–there are tons of contenders for second place but anyone who argues for someone else at the top is either trying to be too clever or doesn’t know what s/he’s talking about.)

    • Ty

      Tom Tango does something like that with his baseball projections. If, for example, a player has 200 plate appearances and is has a OBP of .400, then he would regress the players next 400 (or how many PAs they are expected to have), by combining their current OBP and the league average (which is around .320). What this does is that it prevents any sorts of outliers that can be had from a lower numbers of attempts, and as the player racks up more attempts, their productions gets closer to their true talent than the league average.

      I’m not sure if this is the way to go with older (pre-70s) players, as using a league average weight to expand their number of drop backs might artificially lower their true value. Maybe the older players would produce on a similar level with more attempts, maybe not.

      • I was pretty certain I got the idea from somewhere in baseball. I should have guessed it was something Tango used.

        There may be a better input than league average, but it seems to me that the basic idea of adding to the sample size by creating dropbacks that are a regressed version of what they actually did makes sense as a way to give those with fewer dropbacks more of a chance without overvaluing them.

    • bill

      The reason Tarkenton is dismissed from from the GOAT discussion is because he was the “goat” in 3 super bowls. He was just as bad in most of the playoff games that the Vikings won to get there. I’ll tell you someone that hasn’t dismissed him. Fran himself. He calls himself the GOAT, but he isn’t even close.

  • Chris

    Something that really stands out to me is that Manning is basically alone at the top. Brady is neck and neck with Marino although I expect him to separate himself from Marino. But regardless, Manning is a full tier above everyone else statistically. To me, that’s incredible.

    • Well, there must be a reason why Manning is the only four time MVP winner, and only five time MVP winner, in addition to being either First (7) or Second (3) team All Pro 10 times in his sixteen year career, meaning he was selected either the #1 or #2 QB in the league that often. This study and the awards seem to validate each other.

      • Kibbles

        Scott Kacsmar keeps track of All Pro votes, and I believe Peyton Manning has received over 50% of all first-team AP All Pro votes over the last decade+. Which is insane. In any given year, an All Pro voter was more likely to vote Peyton Manning as the best QB in the league than ALL OTHER QUARTERBACKS COMBINED.

        I often say that it could easily be argued that Manning should have won the MVP award in 8 of his past 10 seasons. You’ve got the 5 he actually won, for which the argument is pretty easy. After that, he lost three MVPs to RBs. In 2005, Shaun Alexander won the MVP with 19 votes to Manning’s 13. When you consider that Shaun Alexander probably wasn’t the most valuable player on his own team (hello, Walter Jones and Steve Hutchinson), the case for Manning over Alexander is an easy one. Last year, of course, was the race between Manning and Peterson, which many consider a coin flip- the case for Manning over Peterson is another easy one, and one that was made in many places. Finally, in 2006 you had LaDainian Tomlinson posting his huge TD season. That was a magical season, but it was a pretty magical season for Peyton, too- statistically speaking, 2006 was Peyton’s second-best season (behind 2004 and ahead of 2013). I’d take the second-best season from Peyton Manning over the best season from LaDainian Tomlinson, but that’s just me.

        Peyton really had no argument for the MVP in 2007 and 2010 (Brady was clearly superior both years), and Peyton was injured in 2011 when Aaron Rodgers won it. But otherwise- and particularly if you believe that quarterbacks are always by definition more valuable than running backs- it’s not a stretch at all to say that Peyton could and should have walked away with the other 8. He’s achieved a level of dominance that is, quite frankly, pretty hard to imagine. Steve Young was better at his peak, but the fact that Peyton’s pretty much been the best player in the entire NFL nearly every season for more than a decade is pretty mind-boggling.

        • Red

          Agreed. With all the discussion about “system QB’s” and who benefitted the most from their surroundings, Peyton Manning transcends that entire argument. He IS the system. He runs the practices, he calls the plays, he designs plays, he game plans like a coach. He basically runs the team. Does anyone really believe that John Fox or Jim Caldwell or even Tony Dungy had more power than Peyton? Look at how abruptly the Colts crumbled when he was injured, and look at how he completely changed the culture in Denver immediately upon arrival. No other player in NFL history had entire organizations revolve around him like Peyton Manning does.

          Brady answers to Belichick. Montana answered to Walsh. Peyton answers to nobody. GOAT.

        • I always sort of thought the one he actually won in 2009 should have gone to Drew Brees, but I don’t have a big problem with Manning winning it.

        • Corey

          I agree with most of this, I’d just add that I would have given the 2005 MVP to Walter Jones, or shared it between Jones and Hutchinson. If there were ever a year a lineman should have won, it was that year.

  • Chase Stuart

    Another thing to consider for fans of giving adjustments to quarterbacks who played in eras with fewer dropbacks. What do you suggest we do with Dan Fouts’ 1982 season? Should that get pro-rated to 16 games?

    • Scott Kacsmar

      I’m always against pro-rating stats.

      • Chase Stuart

        What about players from the 12-game era?

        • I know we’re limited to 1960 (or even a few years later when adding sacks), but couldn’t you calculate the VALUE for each game and produce a per game career average? Though, I guess you still end up with the problems of older QBs having huge min and max games because of the 12 and 14-game seasons having a bigger impact on the numbers than they would in a 16-game season. I know I begrudgingly pro-rated to 16 games in my last calculation of DAPR, but I hate doing that. Even if outliers are getting chopped into 1/12 or 1/14, we’re still adding fake data to the sample. Just think of how many offenses and defenses have smoothed out their numbers by virtue of two games in a season. How would Nick Foles’ 2013 season look if Oakland and one more game simply never happened? We never really know when someone’s going to explode or implode in any given game. That’s part of why I’m so opposed to expanding the regular season. I want to keep historic comparisons as apples-to-apples as possible. Save for two strike seasons, we’ve been going at 16 games since 1978 and that works well.

          • Chase Stuart

            The issue with per-game career averages is that it doesn’t tend to conform with popular opinion. A guy with 3 excellent seasons, 3 bad seasons, and 6 average seasons will be the same as a guy with 12 average seasons, but my current formula — and the way we think about things — will give the former QB more credit. I think players like Aikman, Warner, Bradshaw, Namath, etc. who had really terrible years early or late will be unfairly penalized by this, whereas someone like Aaron Rodgers sat on the bench during his “bad” years.

      • Pro-rating is really only bad when the sample size is small (6 games or less is about the point where you lose integrity) but especially for 12+ games the stats hold pretty well.

    • Kibbles

      I think simple pro-rating is too unfair, but I could see some sort of compromise for strike seasons. Maybe instead of pro-rating his 9-game pace, you give him credit for 7 more games at his career averages (or at his averages in the two seasons before and after). So instead of giving him 7 extra games at 320 ypg and 8.7 YPA, you give him an extra 7 games of about 295 yards and 8.0 ypa (just quick estimates of his averages in the four surrounding seasons).

      I would be against pro-rating injury seasons, but I’m pretty ambivalent about it with respect to strike years.

      • Chase Stuart

        I view the strike season as just a good way to begin the broader discussion of what to do for players in the 12-game era. How do we handle Bart Starr or Johnny Unitas and their 12 game seasons?

    • Red

      Not sure why you changed your method from previous GQBOAT entries, where short seasons were 50% prorated. So Fouts’ 9 game season would be prorated to 12.5, and a 12 game season would be prorated to 14. That seems like the best compromise.

    • eag97a

      For QB’s with fewer dropbacks I’m guessing that’s because the run game was a higher percentage of the total number of plays run then. The question is do we credit a QB for successful run plays since it’s part of his responsibilities for executing the plays whether pass or RUN or do we ignore it and devalue older QB seasons?

      As for 12 and 14 game seasons my view is they are of lesser value than the current 16 game format because of the volume of plays run, the quality of the competition (league expansion) and more recently the salary cap. The only argument for the older QB’s are the rule changes that made life easier for their modern counterparts.

      My suggestion would be to sum the standard deviation of each QBs game against league average QB game performance, apply SOS correction by iteration and then finally apply the 100/95/90… rule for their seasons and add postseason numbers to arrive at total value. As for weather adj., rule changes, home/away, coaching variables I have no idea how to handle them and would very much welcome ideas on how to. But Chase’ methodology works very nicely for me. 🙂

    • bill

      I suggest we wipe out Fouts’ 1982 season because that is what his 5 playoff interceptions in Miami did. Big games matter. That one game is more important than his entire 1982 season.

      Also, I want to say this about “dropbacks”. The Packers had the ball at the Dallas 11 in the Ice Bowl final drive. Starr called a give play right over Bob Lilly after drawing him out of the hole with a fake sweep. No dropback, but it is as great of a play as any a quarterback has made. Quarterbacks who dropback more are quarterbacks who lose more, especially in big games.

  • Dave

    Chris- That’s mostly because Manning has played two more seasons than Brady and thus about 1800 more dropbacks.

    Manning’s value per dropback 1.46
    Brady’s value per dropback 1.44

    Steve Young’s value per dropback 1.86!

    I think trying to compare Qb’s from different time periods simply doesn’t work well no matter what kind of math you try and apply. I’d rather just know who was the best QB from each era.

    • Chris

      Didn’t Chase factor in dropbacks? If not, then that certainly makes the comparison a lot more interesting. I don’t want to go down the Brady-Manning debate road (I take the side of Brady for the record.) but I would like to see a GOAT list based on value per dropback. The only drawback I see is a small sample size, but I’m sure that Chase could come up with a minimum attempt criteria.

      • Dave

        Not the way you are thinking. Remember this is a cumulative ranking that is weighted per season. Its not a rate stat. The more dropbacks you have the more value over average you can acquire in a season.

        1. Brady has two fewer seasons in which he can acquire value granted at about 35-40% weight.

        2. The latter part of his career has been higher value so two additional decent value seasons would have a cascading effect.

    • eag97a

      @Dave Sorry to be particular but Manning played as the starter for 3 more seasons than Brady since Brady rode the bench during his rookie season.

      • Dave

        Whoops, you are right eag97a.

  • Richie

    The funny thing about Romo-Aikman is that most of the people who want to lynch Romo are the types of people who think QB Wins is an important stat.

    Romo is entering his age 34 season. Aikman’s final season was his age 34 season.

    Romo has had 5 seasons where his QB Win% was over .500 and Aikman had 7.

    • bill

      I don’t know about QB Wins as an important stat, but BIG WINS is the only important stat. Romo doesn’t have them. I like Romo and he has skill. However, I think the greatness train has left the station without Romo. He has had the talent to be great and he has had the stats to be great. But he isn’t great. Romo may have more talent than Aikman had and he has better stats. But Aikman attained greatness. He is about #11 all time and Romo maybe about #50.

      • Travis Jones

        And what is a “BIG WIN” exactly? I mean, Romo has a 75-48 W/L record, Aikman: 94-71 – a notable difference (61% vs. 57%) So obv Romo has won more of the time in general. Also, he has more comebacks and game-winning drives. So far, we aren’t seeing a trend that Aikman won any kind of games more than Romo, Did Aikman just not try as hard in the “NON-BIG WIN” games?

        Romo took a team to 13-3 in 2007 (that team was good but he was by far the best player), dragged a 3-5 team(when he took over) in 2006 to the playoffs once he became the starter, win against the previously unbeaten 9-0 Indy team in 2006 (eventual champs), a win against the previously13-0 Saints of ’09 (eventual SB champs), played the best December ever by a QB this past year(BIG WINS) and followed it up by two excellent playoff performances (including a BIGGER WIN) and really did everything he could to beat the Packers and should have were it not for a “Tuck-rule”-type situation.

        Does he have the postseason success Aikman did? Of course not, not even close. But I find it ironic that Romo has better statistics in the postseason now than Aikman, and had his best postseason when he finally had a running game – so maybe some supporting cast again makes the ultimate difference between big wins for the two? Romo still has not had a shutdown defense like Aikman worked with through the prime of his career.. Let us also not forget Aikman lost big to the Lions in his first playoff game, and finished his playoff career with a real stinker against the Cardinals, a heavy underdog.

        Finally, implying Romo doesn’t win big games would imply he doesn’t have the stuff to do things like lead 4th-quarter comebacks, which he did in the playoffs this year and has done more than anyone in the league since 2011. So obviously, you were wrong initially and this past year you have been proven even moreso.

        • Steve

          “By far the best player” in 2007? Weren’t there three future HoFers in their primes on that team?

    • Travis Jones

      This is a very silly stmt. Chase already mentioned Romo’s winning percentage is higher than Aikman. Number of winning seasons? Well, that might have something to do with by the time they were both 34, Aikman had played 3 more seasons. Also Aikman had 6 winning seasons, including the one where he only played 11 games. And if we’re going to discount supporting cast (in which Romo’s performance becomes even more dominant over Aikman, who played with the #1 rusher and #1 defense and #1 OLines), Romo never went 0-11. or even 6-10(with a pretty decent team) like Aikman. Romo’s worst season was 1-5, a very small sample size. Other than that, 8-8. The only valid point one could make about Aikman over Romo is post-season performance – but the funny thing is, after this year’s postseason, Romo actually has better numbers across the board there, too.

  • I tried a whole bunch of weirdo ways of getting the right blend of rate and volume. They all left a bad taste in my mouth. Recently, I used AY/A+ as the basis for rankings, and I used a bunch of different methods to weight attempts differently. I only did post merger stuff because I didn’t want to estimate sacks, but I came up with these two tables:

    http://nflsgreatest.co.nf/2014/06/qb-career-values/ – which uses simple AY/A+
    http://nflsgreatest.co.nf/2014/06/qb-value-above-replacement/ – which uses AY/A+ above replacement, with 1 SD below average (AY/A+ of 85) as replacement (somewhat arbitrary, I know).

    Nothing felt “right,” but messing with the math yielded some pretty neat results, I think.

    Something I am currently working on is trying to come up with something like a PassAttempts+ which would be an adjusted Z-score like all of PFR’s Index scores but based on the average pass attempts per team each season…or something. So something like Bledsoe’s 1994, Moon’s 1991, or Stafford’s 2012 would stick out as a high volume+ season. I’m still not sure this is fair for modern players.

  • Chris

    Dave- A valid point. I guess this is the closest we’re going to get unless Chase makes a big breakthrough.

  • james

    Will your article tomorrow adjust for home/road and weather conditions (dome, outdoors, cold weather, etc.)?

  • Red

    Suggestion for rating the older players – Leave them out completely. I would exclude any QB who retired before 1950, and there are several reasons why: a) the QB position in the 30’s and 40’s bears very little resemblence to the modern QB position, as passing wasn’t even their primary job in many cases; b) sample sizes are too small both for individuals and for league baselines, easily distorted by one outlier; c) one season wonders like Johnny Lujack and Cecil Isabel have no place on an all-time list, as each had under 1000 dropbacks and Lujack was a kicker/punter for most of his career.

    I would add two rules to your existing methodology:
    1) Must have played at least one full season in 1950 or later
    2) Must have registered at least 1,000 career dropbacks

    These rules eliminate the impossible-to-measure pre-modern QB’s and the current players who have only logged one or two seasons (Wilson, Kaep, and Foles have no place on an all-time list yet, IMO).

  • Dan

    Here is one way of calculating something similar to a per-game career average or a per-dropback career average, which deals with some of the standard problems of using a career average.

    Average together two seasons of league average performance and the quarterback’s n best seasons, where n is selected separately for each quarterback in order to maximize his score. (I am not sure how many seasons of league average performance to include, but two seems to be in the right ballpark.)

    You can think of this as a step-by-step process, where first you only include the player’s best season (averaged together with two imaginary league average seasons). Then you also include his second best season in the average, but only if doing so increases his score (otherwise, you stop with just his one best season). Then you also include his third best season in the average, but only if doing so increases his score. Then also include the fourth best, etc.

    This way, a quarterback’s bad or average seasons don’t hurt him, peak performance is emphasized, and long peaks are valued over short peaks (moreso if you choose a value larger than two).

  • To me trying to rank QB’s who played in different eras is comparing apples and oranges and a few lemons thrown in for guys on the beginning or end of the eras. I think the longest comparison should be no more than a decade. And should include some common sense qualities and not just statistics…For instance I’d love to see you rank then as to how you would draft them for your team of the 40’s thru the year 2010. Would you really choose Fran Tarkington over Roger Staubach? I guess what I’m suggesting is find a formula that measures not only stats but the intangibles as well. Yeah right I can hear you now…but I must tell you I do not agree with almost your entire list. but I suppose you knew this when you published it. Sincerely Jane

  • BringBackTheFlex

    “Surely Brees has provided more value to his teams [than Otto Graham]”

    Hmm. Graham played ten years, took his team to the Championship EVERY year and won seven times. I think he added more value to his team than Brees does to his.

    • welshpete

      Well, plenty have weighed in, so why not me?
      The main point of conjecture here seems to be stats Vs big game wins. I understand the validity of the argument of which is more important; and I don’t think it will ever please both sides – if you came out with a formula that included every single factor, somebody will start arguing about finer points that simply cannot be quantified like ‘poise’ and leadership. It would be never ending.
      From a statistical point of view, I’d pretty much agree with what’s there – it takes into account many factors, including all of the important ones. The fact that it doesn’t skew figures because someone won the big one is, in this case, to be accepted; and as a starting point to be debated on the GQBOAT when you add in the intangibles there are bound to be changes.

      Here’s how I’ve been looking at it. Comparing over different eras is hard, and gets harder the further apart you move it – for starters Brady and Manning would have to play defense in the 30’s too…would that mean they’d be great QB’s in between making tackles and interceptions? However, certainly from the 60’s on (accepting some differences) I’d always try and weigh up what a player would do if you slotted them into another team. I’ll use an example I know best being a long time Dolphins fan.
      Take Marino out of the 80s/90s Dolphins and give him 15 years between 1970 and 1985 with the ‘fins. Before the arguments start about how Shula might have changed how they played because they had a rocket arm back there (you’d still give Csonka the ball though, right?); you’d have to admit the team would most likely be at least as successful; if not more so….and that Marino’s post season numbers and career would most likely improve….he also wins at least one ring if you directly pick him up in 1985 and put him with the Bears; Payton in the backfield and THAT defense. Ditto 86 Giants, Ravens SB teams, the Steelers at almost any time, the current hawks……
      But take it further – give him the 1990(ish) Eagles defense – does he become a world beater? Put Terrell Davis, Eric Dickerson, Barry Sanders with him (so you can’t key on pass). With a different fall of the cards – Marino not only breaks all those records; but also has a handful of rings (and lets face it if he’d won in 85 and/or 86 he’d be top 5 all time without question).
      BUUUUT. Take him out and give him the 76 Bucs; the early 2010’s lions or the Browns or Raiders for the last 15 years….does he now get the same stats? Probably not, very few if any playoffs and not even in the top 20?
      That’s a single case. Now for some of the guys at the top – notably Brady and Montana – they’ve had amazing teams around them – admittedly they might now be considered great without them; and may have dragged weaker teams to bigger things.
      So this leads me to my method for picking the GOAT outside the stats….take the one being debated and in a straight swap, would you change them for someone else? Would I pick Marino over Elway (mid 80’s or late 90’s Broncos) sure I would….would I pick him over Bradshaw? Griese? Warner? Flacco? Dilfer? Manning (E)? any number of SB winners? In a freakin’ heartbeat.
      Then comes tougher ones – pick him over Starr? Young? Manning (P)? Rogers? Brady? hmm tough decisions.
      The worst thing of all is that the one guy in the SB era I can’t find a way to say I’d prefer Marino is his nemesis from XIX….Montana. And yes, it’s completely based on intangibles…but on that day, Montana was winning and nothing was stopping him….same in XVI and XXIII for that matter (I could have QB’d the 49ers to a win in XXIV).
      Am I saying Joe Cool is the best? No, I still go with Otto for that – even though I didn’t bear witness to it….10 years = 10 Champ games….

      As for how you rank the usual contenders up there, I always think it helps to consider what they might have done with a better team – or a worse one; and where that would leave them. Starr with a weak? Still great but not all time….Brady, Montana, Young – same thing.
      Marino with a better supporting cast? Does this now mean 1/2/3 Superbowls and we’re in a very different debate. Ditto Drew Brees to an extent, and probably more than anyone, Peyton Manning. Could you imagine him having 15 years with Belichik’s Patriots? (I’m not saying Brady did a bad job, just that Manning was close with a weaker team) you could argue for 3/4/5/6 Superbowls – just because the days Manning isn’t up to standard (and they are pretty few and far between) the team picks up the slack.

      Anyway, that’s just my opinion, and an interesting way to think about things – take your potential #1 and try swapping with the big boys, and do it objectively….

      QB is really tricky in this circumstance….but work it the same way for other positions and it usually points to someone obvious but who WOULDN’T you swap for Rice at WR?
      At RB there are a few candidates but it’s hard to see past Brown, Payton, Sanders and a few others.
      An outside linbacker? You can have LT or any other to play the game…..sure you get the gist anyway…just my £0.02


  • BringBackTheFlex

    Does your formula take into account the strength of the running game and defense? A QB with a poor running game may have had to take many more risks to compensate. Or a poor defense might make a QB take more risks, while a great defense can make a QB play more relaxed knowing he doesn’t have to do it all, he can make mistakes and be bailed out by his defense.

    Know what Trent Dilfer and Joe Montana have in common? Neither of them every lost a SB with a top ten defense.

  • Mike Cross

    What stands out to me is Eli Manning coming in at #120. I’m a Cowboy fan, but I believe he has 2 Super Bowls & consistently at top of many of the yards & other metrics. To me that is completely absurd!

    • Chase Stuart

      He’s definitely consistently at the top of the interceptions metric!

    • richintx

      what? Eli is horrible. Unless you get points for throwing INT’s

  • Chase Stuart

    One other Aikman note: I did not exclude negative seasons, which really hurts Aikman. He jumps from 38 to into the top 30 on the career list if you exclude below average seasons for all quarterbacks.

  • Scott

    Eli Manning 2 Super Bowl Rings…beat Brady both times…#120
    Rivers…same draft classs…no Super Bowl games…#20
    this list is about as stupid a time waste as it gets

    • Shaun

      Manning also has 67 more interceptions than Rivers, and I beleive this is based on regular season not post season.

      • Peyton also has way more attempts than Rivers. It’s why the completion to interception ratio Must be considered.

    • bill

      Scott, finally someone who makes sense. I have Eli in my top 20 and Rivers nowhere close.

  • Shaun

    I think the Super Bowl arguement has always been overweighted, it takes far more than just the QB for a team to get to and win a SB. You have to have a good owner, front office, coaches and good TEAM, no one player has ever won a SB. I have always looked at post season as more of a tie-breaker kind of thing, if QB A and B have similar regular season stats but QB B has won two SB then I think that elevates him.

    • eag97a

      Agree on all points. I also add you need a healthy dose of luck to win a SB. But personally I believe SOS is a bigger factor than SB’s in qb evals. Winning SB’s specially before the salary cap was very dependent on how good a qb’s teammates are.

    • bill

      That is good Shaun. You get guys like Fouts, Tarkenton and Marino that sucked in the post season. But it doesn’t matter. They threw for a lot of yards against bad teams in September and October. And isn’t greatness defined by big stats in meaningless games? I don’t think so.

  • Tim Truemper

    So many comments in proportion to other subjects. Just mentioning GOAT list of QB’s and the urge to put in one’s two cents soars. I see a data analysis prospect here. Pardon if I missed a comment on this, but it was stated with surprise about Fran Tarkenton’s standing on the list. I believe one statement said, “his numbers don’t look impressive…”. Lets remember that when he retired he was the career leader in yards and TD’s, passing John Unitas. And while “Scrambling Fran” did play a lot in the low yardage years of the early to mid 70’s, he also played during the high flying times of the early to mid 60’s. Overall, this is a great list and it is appreciated that the methodology is spelled out clearly. It would be interesting to see if any changes in weighting certain variables (yards for touchdowns as an example), has a major effect on standing.

    • bill

      Tarkenton is a good example of a quarterback that compiled numbers that didn’t matter. He was awful in the post season. A passer rating of 58 with 17 interceptions in a 11 games. These are numbers that matter.

  • Shaun

    I would love to see a similar list for the RB position as well

    • Chase Stuart

      Thanks, Shaun. I would too. Unfortunately, they are not next on the to-do list: a less-exciting position has priority.

  • bill

    Greatness in quarterbacks is more mental than physical, and it is also intangible. Career stats are mostly compiled against weak teams in meaningless games during garbage time. If you want to access a quarterback by stats, look at how he performed in the 40 biggest games (regular and post season) of his career. You’ll find that guys like Marino, Tarkenton, Fouts, Favre, Brodie, Hadl, Romo and Tittle were mostly bad in big games. None of those guys belong in the top 25. Starr, a guy not even is this top 25, is #1 and second is not close. He was about 36-2-2 in the biggest games of his career with 66 td’s and 13 ints in an era when those stats were unheard of. He is tops in everything that counts: championships; post season rating (still); post season winning %; big game performances; bad weather games; games vs legendary defenses; games that mattered in December and January. His play calling, poise and decision making set him apart. Way apart. That is what greatness is. Not a string of 4000 and 5000 yard seasons that end abruptly with a disaster in the playoffs. Here is a top 25:
    1 starr, 2 montana, 3 graham, 4 brady, 5 luckman, 6 unitas, 7 bradshaw, 8 p manning, 9 brees, 10 Rodgers, 11 aikman, 12 dawson, 13 warner, 14 baugh, 15 stabler, 16 eli mainning, 17 roethlisberger, 18 young, 19 elway, 20 van brocklin, 21 flacco, 22 staubach, 23 b layne, 24 plunkett, 25 a herber.

    • If you look very closely at the teams some of your top guys played with, you will find that they were surrounded by great defensive teams, which contributed a lot to their offensive play calling, and the amount of pressure on their shoulders to carry the team. QBs like Bart Starr, Joe Montana, Terry Bradshaw, Troy Aikman, and Tom Brady were surrounded by highly ranked defenses in their championship and Super Bowl winning runs. Other QBs, Eli Manning, Peyton Manning, and Drew Brees, have been surrounded by mediocre defenses. In a team game, which NFL football is, that matters a great deal.

  • bill

    In NFL games, the team with the most passing yards doesn’t usually win. Its because a team plays bad and gets behind. Then they have to throw, and teams let them complete passes while protecting their lead. So much of NFL passing yardage is garbage. Matt Stafford passes for 5000 yards and isn’t even very good. In 2013, RG3 would have about 40 passing yards in the first half and end up with 300 meaningless passing yards against soft defenses.

    I point this out because this is how you can come up with a top 25 with Hadl, Rivers, Romo, Tittle, Brodie and Fouts. You can add Marino, Favre, Jurgensen and Tarkenton who are all overrated and don’t belong in a top 25. If I’m not mistaken, those 10 guys have 1 total championship.

    You can out think yourself with a formula.

    Here are 11 guys left out that can replace them: Starr, Bradshaw, Aikman, Warner, Stabler. Eli M, Roethlisberger, Bobby Layne, Plunkett, Flacco and Herber. That’s 11 guys with 25 titles.

  • bill

    I want to bring up a forgotten quarterback named Tobin Rote. In 1963 ant the age of 35 turning 36, he led the Chargers to the AFL championship. His backup was a young guy named John Hadl. Rote also led the 1957 Lions to the NFL title, with a playoff comeback from 27-7 down against a 49er team quarterbacked by Y A Tittle. He is the only quarterback ever to win both the NFL and AFL championships.

    I have to wonder why the Chargers played an old guy that nobody remembers, when they had one of the 25 greatest ever on the bench. I’m sure that over his career, Hadl and Tittle averaged more yards per dropback against strength of schedule opponents that were better with a supporting cast that was weaker and whatever and who cares. Rote was better than Hadl and Tittle. The Chargers knew that in 1963, the 49ers knew it in 1957 and I know it now. He was also better than Fouts, Brodie, Rivers and Romo. I’m not making a case to put Tobin Rote in the top 25, but I would put him ahead of these 6 guys. Also, it really doesn’t matter if any of these 6 guys was actually better or more talented. He got it done when it mattered. That is greatness.

  • Tim Truemper

    Bill you make many good points. You say however that you can “out think” yourself with a formula. Technically you are referring to the validity of the measure. And you make a good case questioning the validity of Chase’s formula as one of greatness. I would surmise that it is a good measure of regular season performance and not QB “greatness.” You provide a basis for a metric that would evaluate the most important games of a QB’s career (playoffs, regular season opponents quality, and importance of regular season games) as a way to numerically test that.

  • William Berry

    Again I have a point that you, or anyone, can not compensate for.
    Elite receiver helps the QB.

    For example, Unitas had Raymond Berry, the receiver who invented timing routes. Is the fact that his best years were those before Berry (no relation) retired coincidence or is it just because of his natural aging? To answer we need a Unitas without Berry when he was young. 🙂

  • Richie

    Where did all these commenters come from? Was this thread cross-posted someplace else? Have all these guys (and trolls) been lurking on this blog all along?

    • I assume in response to the Wisdom of Crowds thread. And, of course, the new comment system!

  • Clint

    In Kurt’s defense, he was still great outside of the dome in Arizona. If we’re going with dome talk, then you gotta put Peyton Manning in there as well… although he too is successful outside of the dome.
    Not sure if I understand how a dome necessarily makes a QB better.