The Top 25

Okay, we're going to go in reverse order here, 25 to 1. Some thoughts about this process and how I want to improve it will follow.

25) Jim Kelly (1986-1996) - 338
Five Pro Bowls, led eight teams to the playoffs, four to the Super Bowl. Didn't win once. That's the defining thing of Kelly's career, and it always will be. Other quarterbacks are remembered for losing Super Bowls, but not four, and not in a row. He's in the Hall because he was a fine quarterback for a solid decade, sometimes brilliant, and always at least good.

24) Dan Fouts (1973-1987) - 345
Led the league in passing yardage four times in a row, but never led the league in passer rating. More or less useless on the ground, and fumbled a lot despite not being sacked very often at all. Like Kelly, is in the Hall despite never having won the big one, and was much less of a factor in the postseason, going just 3-4 in his career. His run of eye popping numbers from 1979-1982 was part of a much longer run of being one of the most effective and efficient passers in the league.

23) Donovan McNabb (1999-Present) - 355
Made 6 Pro Bowls, and got the Eagles to a single Super Bowl. Won 9 playoff games. Right now it looks very much like he will fade away without making much more noise after a very disappointing 2010 season, though of course that could change. His Hall of Fame case will be interesting, though give the perception of him right now, I can't see it being successful. The two men just behind him here made the Hall because there was almost universal acceptance of them as premiere quarterbacks in their era. With McNabb, there have always been a number of detractors who had a low opinion of him; not just relatively low, but objectively low.

22) Bert Jones (1973-1982) - 359
Definitely a surprising name to see on the list. Jones won an MVP in 1976, his only Pro Bowl season, which was legitimately terrific by any measure. How does he make this list? Well, he has a lifetime Rate+ of 110, which puts him in a very good position to do well on this list straight away. He ran quite well, which helps. Of course, his scores for longevity and winning are terrible, since he's credited with only six seasons as a starter, and never won a postseason game. Here's the interesting bit. Of his 359 points, 68 of them come from avoiding fumbles. The next closest guy gets 44, and the mean is just under 13. It's bizarre, and I can't blame anyone if they think that the score is suspect on the face of it. But I honestly can't see at this point how it makes sense to give him less credit than that. Bert Jones was an absolute marvel when it came to not dropping the football. Despite being sacked a bit more than average, and despite running the ball a decent amount, he fumbled 16 times in his career. To put that in perspective, Warren Moon fumbled more than 16 times in a season. Twice. Bert Jones is to quarterback fumbles what Jim Brown was to NFL running backs in the late '50s and early '60s. The point values assigned could certainly be off; most quarterback fumbles are recovered by the offense; a slim majority in the case of those that occur on sacks, and the vast majority of those that take place on botched exchanges. Since all fumbles are not created equal, all fumblers are not created equal. However, I also made no attempt to account for the yardage and downs lost as the result of quarterback fumbles. While I could be wrong, I believe I'm estimating the impact of fumbles relatively accurately, and assigning Bert Jones a number of points that are in line with his ridiculously good record of hanging onto the football.

21) Drew Brees (2001-Present) - 359
Has an excellent chance to get a lot higher, obviously. Does everything well as a passer, but gets a big hidden bonus for his ability to avoid sacks, which followed him from San Diego to New Orleans.

20) Kurt Warner (1998- 2009) - 365
Certainly a player about whom opinions will be terribly varied. Criticized for only succeeding when he had terrific groups of receivers, and for having an already relatively short career marred by a five year period in which he disappeared as a premiere player. The court of public opinion seems to have swung around to a consensus that he's Hall of Fame worthy, though it will be interesting to see if it actually happens, since it's always harder to reach the Hall of Fame than people think. Statistically, his inconsistency is defined by his fumbling, and his taking lots of sacks from 2002 to 2006, problems which are obviously related. The fumbling is fairly easy to explain based on his hand injuries. The increase and then decrease in sacks, on the other hand, is more of an enigma. Like the man he replaced in St. Louis, Trent Green, Warner is a fascinating "what-if" subject, based on his late start in the NFL and the freakish nature of his injuries.

19) Ken Anderson (1971-1986) - 366
The highest ranking non-Hall of Famer who is eligible for induction. Led the league in passer rating four times, so he obviously does well in terms of passing efficiency. MVP in 1981. Though not remembered much for it, was an efficient runner who averaged 5.6 yards per carry while totaling more than 2000 career yards. Is not in the Hall based entirely on not having won it all, and it's interesting to compare him to Fouts, his almost exact contemporary. Anderson does better on this list because of his superior efficiency in avoiding interceptions, and because he was a superior runner. Fouts had gaudier career numbers, in large part because of the offense in which he played. I don't actually feel comfortable putting Anderson ahead of Fouts, and would probably reverse their relative positions based on my subjective opinion, but it's an interesting comparison.

18) Len Dawson (1957-1975) - 376
Led the Dallas Texans to the AFL title in 1962, after coming over from the NFL Browns. Won another AFL title in 1966, but the Chiefs were blown out by the Packers in Super Bowl I. Three years later, brought the Chiefs back to the Super Bowl, where they clobbered the Vikings to even the AFL-NFL Super Bowl record before the merger. Benefits here from defaulting to average in terms of sacks; in the years we actually have number, he was sacked an awful lot. Led the AFL in passer rating six times in its 10 year independent existence, and was actually in the other league two of the four seasons he didn't.

17) Dan Marino (1983-1999) - 383
One of the hardest men to sack in football history, you can actually see his decrease in mobility in the sack numbers. Or you can choose to read his sack trends as reflecting his decreased mobility, at any rate. Marino was efficient in every area as a passer, and his numbers look even better when you consider how little he was sacked. Actually fumbled a lot, considering how few chances he had on sacks or runs. It would be interesting to see why, if we had a breakdown of how he fumbled. Was he prone to botched snaps, or is there potentially some kind of effect where quarterbacks can drastically limit sacks, but are still prone to the type of sacks that lead to fumbles? Anecdotal evidence would be welcomed.

16) Fran Tarkenton (1961-1978) - 388
The highest ranking QB of the Super Bowl era who never won one. It's easy to see why. Though Tarkenton is now most often referenced as a scrambler, he held a number of important passing records when he retired. The closest analog to Randall Cunningham's drastic stylistic reinvention in Minnesota in 1998 is probably Tarkenton. As he aged, Tarkenton ran a good deal less, though he remained dangerous on the ground. For the last three years of his career, however, beginning in 1976, he became ineffective as a runner. While this happened, however, he became more accurate, putting up four of his five best Cmp%+ seasons in his last four seasons, leading the league twice.

15) Troy Aikman (1989-2000) - 391
Without the credit for winning, ranks lower than any Hall of Famer besides Blanda. Passing numbers are good, but not great, but are augmented by his taking few sacks and hanging onto the football very well. It's pretty easy to build the case that Aikman's career was the result of a pretty good QB surrounded by a lot of talent, but impossible to argue that that career isn't worth of the Hall. His raw postseason passing numbers are terrific as well; many quarterbacks who were successful in the playoffs still see their efficiency go down compared to the regular season, which is natural, since they face generally better defenses in postseason games. Aikman's numbers go up significantly.

14) Ben Roethlisberger (2004-Present) - 393
Surprised? Well, it's pretty much impossible to keep him off a list like this. Roethlisberger's career Rate+ of 112 is elite, and though he takes a big knock for the absolutely ridiculous rate at which he's sacked, he's done everything else so well that his two Super Bowl wins propel him into rare air. If he used his physical gifts to limit sacks instead of running into them, he'd be up with Brady and Manning as the best of the era. Even as things are, he's elite.

13) Bob Griese (1967-1980) - 394
Two Super Bowl wins, a solid record of passing efficiency, and didn't do anything else badly enough to ding him down much. Sometimes feels like a forgotten Hall of Famer, which is odd given that he's remained a high profile broadcaster. My guess is because there's always somebody better in any given category... Griese was a winner, but Bradshaw won more. He was efficient, but Staubach was more efficient. He was accurate, but only led the league once, and his numbers look pedestrian by today's standards. His career numbers aren't eye popping because he only ever played a full season twice, and he played in run first offenses that asked him to play consistently well, but rarely to shoulder the bulk of the load. That last is a legitimate criticism, of course, and a list like this is going to be too kind to quarterbacks like Griese, and by extension, too harsh on Marino and Fouts and Warner and their like. However, Griese ranks as a no doubt Hall of Famer.

12) Sammy Baugh (1937-1952) - 394
Baugh tests the limits of this system. Can you really effectively compare Baugh, who played before our modern conception of the quarterback position existed, to Peyton Manning? Is that even a realistic goal? I can't answer that question, but I can say that Baugh does extremely well here, and his ranking seems defensible when compared to expert subjective rankings. Baugh ranks as much, much better than his contemporaries as a passer. That those contemporaries were not pure passers isn't the point, since neither was Baugh. He was perhaps the best punter of his generation, and was a gifted defensive back early in his career as well. Is the comparison absolutely apt across eras? No, but the statistical result here tallies well with recorded observation. Baugh was considered a great passer in his time, and he certainly shows up as such here.

11) John Unitas (1956-1973) - 396
Often brought up as a potential best QB of all time, Unitas falls short here because while he was quite efficient compared to his contemporaries, (112 career Rate+) there are others with a significant advantage over him. Shouldered the load far more than many of his contemporaries, as he led the league in attempts and yardage four times each. He declined dramatically after his lost 1968 season, and is probably underrated in many fans' minds as a result, since the pre Super Bowl history of the NFL is so often ignored. In 1968, he was replaced by Earl Morrall, who went on to win the league MVP. In his Super Bowl win, Unitas didn't play well, got hurt, and was replaced by Morrall again. When he left the Colts for the Chargers, he was a pitiable mess with a passer rating of 40.0. Based on those facts and nothing else, it's hard to avoid the sneaking suspicion that the great Johnny U was a bit of a fraud. Based on his career prior to the Super Bowl era, that sneaking suspicion is utter crap.

10) Brett Favre (1991-2010) - 397
It's hard to think of anyone, in football or anywhere else, who is as defined by contradictions. Favre is known for toughness, and as a diva who craves attention. He's known for tremendous clutch play, and for clutch disasters. He's been adored and despised, often by the same people and without much time in between. As a passer, he's good everywhere but the interceptions, where he's a bit below average. Naturally, he does very well with longevity. With any luck, he's actually done now.

9) John Elway (1983-1998) - 410
The difference here between John Elway and Donovan McNabb is the two Super Bowl wins. Without the points for those two wins, Elway would be just behind McNabb, and just ahead of Fouts. Still a Hall of Famer? Yes, definitely, but it's amazing how much the Super Bowl wins affect a legacy.

8) Bart Starr (1956-1971) - 426
Get the fourth most points for winning, behind Montana, Bradshaw, and Brady. His reputation now is based almost entirely on his leadership of the dynastic Packers, it seems, and very little on his abilities as a passer. To some degree, that's understandable. While Starr's rate statistics look very good compared to his contemporaries, he never threw for even 2500 yards in a season, in part because he was hurt in many of his seasons as a starter, and in part because Vince Lombardi's offense never required him to do the lion's share of the heavy lifting. He led the league in passer rating three times, completion percentage three times, and a few other rate stats a number of times, but he never led the league in any counting stat. He was a very efficient passer, and his postseason performance raised his passer rating 24 points over his regular season rating. A winner? Absolutely.

7) Tom Brady (2000-Present) - 444
Might pass Bradshaw and Montana for win points even if he doesn't win another Super Bowl, if he plays a few more effective seasons and wins some playoff games. There is no knock to make on Brady's game statistically. He's just fantastic, and will add to his total here before he's done, I'm quite certain. His 2010 season should propel him upwards.

6) Roger Staubach (1969-1979) - 447
Often gets overlooked in the best ever debates. His career stats aren't that impressive because of the era and his short career, but his rate stats are frankly ridiculous compared to his contemporaries. He had a Hall of Fame career if you look only at what he did after he turned 31, and even though he retired at 37 after two seasons of leading the league in passer rating. Took way too many sacks early in his career, but got that sorted out later on, and was better than league average. Isn't often remembered as a runner, since Tarkenton seems to have a death grip on that reputation in the '70s, but his rushing numbers are very, very good.

5) Terry Bradshaw (1970-1983) - 447
Bradshaw is often dismissed as being simply a product of his great teams, but in fact, he compared well to a lot of Hall of Fame QBs here even without his points for winning. His passer rating is unfairly low, given his relatively low completion percentage, and relatively high yards per attempt numbers. Later in his career, the passing game became more important to the Steeler attack, and Bradshaw was up to the challenge.

4) Peyton Manning (1998-Present) - 460
Tremendous regular season numbers. His postseason struggles are somewhat exaggerated, and it's not really that his single championship is out of line with what's expected from a QB of his caliber, but that he is always, and always will be, compared to Tom Brady, the other super QB of his generation.

3) Steve Young (1985-1999) - 488
The rate stats are obscene. He led the league in passer rating six times, yards per attempt 5 times, completion percentage five times, and many others a similar number of times, all in an eight year window as the 49er starter. The knock on him is that he took over on an established, stacked team, and that he never had to suffer through a decline period, because the moment that the 9ers crashed, his career ended. The flip side is that he put up even better numbers than the Hall of Famer he replaced as the 49ers' QB, and that his final team that crashed went 2-1 with him, and 2-11 without him. His passing and rushing numbers are so good that he rates as a top ten QB even without any credit for winning at all.

2) Otto Graham (1946-1955) - 514
Just today I saw someone trumpeting Graham's resume, bringing up how he brought his team to his league's championship game every season of his ten year career, winning seven of them. I've already talked about how misleading that is, given that he played in leagues with only a handful of other teams during some of those years, but the numbers are suspect in other ways as well. Graham's passing rate stats drop off a cliff after the Browns joined the NFL. After posting passer ratings of 112.1, 109.2, 85.6 and 97.5 in the AAFC, Graham posted 64.7, 79.2, and 66.6 in his first three seasons in the NFL. The way we're calculating the rate stats, though, blunts this significantly. While Graham's Rate+ and its component statistics do drop during his NFL years, the drop isn't as precipitous. Graham was still clearly one of the best passers in the NFL, though the NFL was not as much of a passer's league. In all, Graham is rated only 7th in win points, but his rate stats elevate him nearly to the top of the heap.

1) Joe Montana (1979-1994) - 530
Was he the most talented QB of all time? Probably not, but I tend to buy this system's pronouncement that he had the greatest career of all time. Four Super Bowl titles, won against a league that had expanded since some of Bradshaw's. Rate stats that put all but a tiny handful of his fellow passers to shame, and terrific peripherals. Montana was rarely sacked, fumbled hardly at all, and ran well enough until the very tail end of his career. I kept an eye out in the course of this project for a QB who actually deserves the reputation as a "game manager" that Trent Dilfer so manifestly does not, someone who actually did take care of the football very well. In fact, Montana is the best at this by a wide margin, but he's also so good at all the other aspects of quarterbacking that nobody would dream of calling him a game manager. There wasn't anything he did badly, and not a season in his career in which the advanced stats see him as anything but significantly above average.