Let's be honest. Arguing about who is the best QB really has very little to do with those particular guys. It always ends up actually being an argument about how you measure quarterback greatness, or flat out what is you're expectation of the quarterback position. If you think a QB who can't make big plays isn't worth his salt, and another guy thinks that a QB has to throw 30-40 passes a game to be good, you'll never agree on a who's best. You may have exactly the same opinions about who has the strongest arm, the best mobility, and so on, but you value different skills and abilities. Then again, you may agree on what you expect the QB to do, but disagree on how to measure it. Eye test or stats? If stats, which stats? Well, I think you can make a pretty good case for the one I've developed.
First of all, let me make clear what I'm actually trying to measure. The Quarterback Overall Rank, QBOR as I call it, is a statistical measure of "eliteness." It is designed to eliminate as much as possible factors that aren't under the quarterback's control, and focus on the basic skills of quarterbacking that will give a player the highest possible ceiling for success assuming all other phases of the game are operating effectively. Passer rating is still probably the best evaluator of how good a game a given QB has, but who is a better QB is better described by comparing the eliteness in the basic tasks the QB is responsible for. QBOR is used for just that, to determine who has a more elite skill set. Now that that's clear, let me be specific about what is not being measured and why...
THE WINS FALLACY: A lot of people take into account wins, especially playoff wins, in their evaluation of QB's, and I think that's ridiculous. There are way too many factors besides the QB involved in winning and losing. There are groups of players known as the "defense," and "special teams," for example. There's another problem when you look at playoff performance, tiny sample sizes. The most obvious example is Tony Romo, who has been branded "can't win in the playoffs" after playing only four postseason games (one of which he won with a 104 passer rating, and one of which was against the 2007 Giants, incidentally who beat the previously undefeated Patriots also). Nobody would pass judgment on a QB after four regular season games; how does it make sense to evaluate a guy's postseason ability from the same tiny sample size? On the other hand, Eli Manning has two rings, Joe Flacco, Trent Dilfer, Mark Rypien, Brad Johnson, and Jeff Hostetler have all won SuperBowls also (Hostetler has like a 112 postseason passer rating, btw), which goes to show there is not that strong a correlation between playoff success and overall quarterbacking ability. Winning games only shows that a given QB is elite at being on a good team.
THE TOUCHDOWNS FALLACY: I'm probably going to catch a lot of flack for this one, but the fact is that scoring plays is a pretty weak surrogate for quarterback skill. Think about it. How many TD's would Ben have if the coaches hadn't been calling 36's number everytime they got to the redzone or got a 10 point lead in the second half? Conversely, how much of Tom Brady's production has come under the assumption that a 40 point lead with three minutes doesn't necessarily guarantee a win, and 1st and inches on the goal line is a great passing down? Total touchdowns reflects the coaching philosophy more than anything else. The NFL's passer rating stat uses TD%, and I don't even know what that's supposed to have to do with anything. If you've ever heard someone say "he may not have passed for a lot of yards, look what a high percentage of his passes were TD's!" let me know. Obviously, TD% is incredibly skewed toward certain systems as well. Throwing lots of TD's shows that a quarterback is elite at not having a goal line back or a defense that can protect a lead (I'm looking at you Tom Brady).
THE CLUTCH FALLACY: Actually, I'm probably going to catch flack for this as well, but game winning come from behind drives mean a whole bunch of nothing. The thing is, TD's count the same at the beginning of the game as they do at the end. Really, breaking a 0-0 tie in the first quarter has arguably an even bigger impact on the game than breaking a 31-31 tie in the fourth quarter. Do you think Seattle having a big early lead had no impact on the past Superbowl? Sure, you want a guy who can handle pressure late in games, but if he's just as good during the rest of the game as the other guy, then his superior ability in the fourth quarter will give him better stats. If he has the same stats, that means his fourth quarter ability is counterbalanced by being not as good in the first quarter, which means he's just better at making games more tense by delaying scoring until the end rather than putting away teams early. All else being equal, game winning drives only show that a QB is elite at drama.
SO, WHAT THEN? I'm going to suggest that the QB's job is not actually to win games or score touchdowns. Those are things that happen if the QB does his job. His job is to put the ball in the hands of his receiver as frequently as possible, as far down the field as possible, while avoiding the players on the opposite team. The stats that really matter are whether he's moving the ball down the field or killing drives with incompletions, stretching the field to create space for the running game and score in a hurry if necessary, and of course whether or not he's committing football suicide by giving the ball to the other team. Also noteworthy is that we're looking at these things on per attempt basis since we want to know how effective the QB is, not how effective the OC thinks he is compared to his runningback. Basically, we're looking at on any given play how likely is this guy to fulfill the basic requirements of a quarterback? That said, it's only fair to point out that it's a lot harder to do the things mentioned above when the other team know that you're passing on 90% of your plays versus when the defense is really only concerned with trying to stop Marshawn Lynch. So we are going to actually give the pass-wacky boys a bit of a bump by looking at yards/game as well.
CAN WE FINALLY SEE SOME RESULTS NOW? No. First I have to point out the special spin that makes this metric different. Rather than dealing with raw stats, I'm actually looking at ranking relative to all other quarterbacks. That means rather than messing around with a QB's actual completion percentage of 59.7, I'm working with his all-time rank in that category of 35 or whatever. This puts all the categories under examination in the same range so they can be weighted equally. All stats were obtained from pro-football-reference.com.
YES, NOW WE CAN GET STARTED. A great QB pretty much has to be able to distinguish between offensive and defensive players. A QB who is prone to throwing the ball to the other team isn't going to accomplish much no matter how good he may be at other stuff. If you look at the top active QB's in Int%, you see Aaron Rodgers at #1 all-time a head above the crowd at a measly 1.8%, truly cherishing the football. Behind him, Tom Brady stands a head above everyone else at 2.0%. Sam Bradford, Andrew Luck, Matt Ryan, David Garrard, Jason Campbell, Shaun Hill, Joe Flacco, and Philip Rivers tied for 13th all time at 2.5% round out the top 10 active QB's. Clearly, a lot of those guys avoid picks by throwing at the grass a lot or taking sacks. If you get right down to it, I could actually go out and post an Int% of 0.0. As important as it is, it's not a be all and end all unto itself. The QB obviously has to also be able to be able to, you know, accomplish something with the football.
The most obvious thing on that list is that he is putting the ball in the hands of his receivers. That's a pretty important sign of skilled quarterback play (as opposed to, say, running three yards past the LOS and throwing a backwards jump pass to nowhere for example). The QB has to do more than just complete a bunch of short passes, though. You want to see him throwing down the field. Yards/Attempt basically gives you a weighted completion percentage that accounts for the fact that throwing farther downfield is harder and rewards players accordingly, so you can account for both consistency and big play ability. If you have a guy who almost never completes a pass farther than 10 yards, but connects on 70% of his passes overall, you're in pretty good shape. At the same time, if you have a guy who only completes 50% of his passes, but has an average of over 15 yards per/completed pass you're sitting pretty (remember I'm assuming for the purposes of this comparison that you have an effective running game). If you look at the top active QB's on the all-time Y/A list, you see the "Killer R's" at the top: Rodgers(3), Rivers(7), Roethlisberger(T11), Romo(T11), Manning(T17), Newton(T17), Schaub(22), Brees(T29), Brady(T29), and Cutler(57). Number one overall in case you were wondering is Otto Graham, who played back when the Browns were good (1955).
Finally, you may have noticed the name "Schaub" in the above list. Passing effectively when there are 9 men in the box is one thing and passing effectively when the opposing defensive coordinator doesn't even know the name of your runningback is quite another. We could look at Att/game here, but we don't want to reward incompletions. Yds/game gives us just as good a look at how focused defenses are on stopping a particular QB and adds his level of success at still passing in spite of them. The top ten active QB's on the all-time Yds/G list are Stafford(1), Brees(2), Manning(3), Rodgers(5), Luck(6), Brady(7), Ryan(9), Rivers(10), Palmer(11), and Roethlisberger(12).
So you there you have it. Add up each player's rank in each of these categories and you get his overall rank (QBOR), the most effective quarterbacks in the game today. 1 - Rodgers(9), 2 - Rivers(32), 3 - Brady(38), 4 - Manning(43), 5 - Brees(54), 6 - Roethlisberger(58), 7 - Newton(71), 8 - Romo(75), 9 - Ryan(92), 10 - Palmer(130), 11 - Stafford(136), 12 - Cutler(159), T13 - Flacco(160), T13 - Dalton(160), 15 - E. Manning(176). I might add that I toyed around with various other formula that included Comp%, weighted Int% more heavily, or excluded Yds/G. Interestingly, it didn't end up significantly changing the players' rankings relative to each other. So without further ado, here are your franchise quarterbacks in order of eliteness.
1 - Aaron Rodgers (9) is really, really, really good. Which Steelers fans know perhaps better than anyone else after the clinic he put on against us in the Superbowl. Still, he's actually even better statistically than I ever knew. Ranking #1 all-time in Int% and #3 all-time in Comp% despite much less reliance on short passes (#3 active QB in Yds/Comp, behind Newton and Vick), Rodgers isn't just best in combined skills, he's pretty much the best at everything. His QBOR of 9 is a fraction of the other elite quarterbacks, meriting him a clear place in the "best ever" discussion (coming soon!).
2 - Philip Rivers (32). We all know Rivers isn't actually this good. But he is. Rivers throws farther down the field than Brady and fewer of his passes hit dirt, that makes him demonstrably better at moving the ball down the field. What else does he need to do? Win playoff games? Call me crazy, but I just can't buy that the point and half difference between Brady and Rivers' playoff passer ratings is the difference between four SB's and none (incidentally, Rivers' postseason passer rating is a point and a half higher than Big Ben's). Brady is better at protecting the ball, which brings him mighty close to Rivers, but not better overall. This is a classic example of rings not making you necessarily a better passer.
3 - Tom Brady (38). I've never been a fan of Brady's because so much of his production comes from easy throws. He has the most boring highlight reels you can find on Youtube (unless you count tortoise racing highlights), mostly clips of runs after the catch or Moss/Gronk abusing substantially smaller defensive backs. The thing is, Brady allows his teammates to succeed by not shooting his team in the foot. He's a master at disaster avoidance whether that be interceptions or costly sacks (how often I've wished Ben would spike the ball at the feet of a RB in some situations), and he's as good as anyone at finding and hitting the open receiver. I still don't like him, but I have to admit he's effective.
4 - Peyton Manning (43).
5 - Drew Brees (54). Big shockers here. Brady, Manning, and Brees play very similar styles, very similar success, and you know they're going to end up very near each other on a list like this. Manning and Brees are actually tied in Int% and within one of each other in Yds/G and Comp%, Brees and Brady are tied in Y/A. Manning has the best Y/A of the three, Brady the best Int%, and Brees the best Y/G (and Comp%). This is one of the situations were common perception is actually borne out in facts.
6 - Ben Roethlisberger (58). Barely got beat out by Drew Brees. As was evident in the past Superbowl, sometimes you need your QB to be more than a pure pocket passer, and you could argue Ben's ability to break down a defense by prolonging the play is worth more than the 4 extra places over three categories that Brees holds over him.
7 - Cam Newton (71). Definitely a rising star. Newton is the highest ranking of active QB's in Yds/Comp (85), which vaults him to a tie with Peyton Manning for Yds/Att. Newton ties with Big Ben in Int% and also joins Roethlisberger as the only QB's on the list who rank higher in Yds/Att than Comp% (although Rodgers and Rivers have the same ranks in each). He benefits from a strong running game, but he also provides some of that running himself!
8 - Tony Romo (75). One of the most under-recognized QB's in the league, whose impressive skills are basically discounted because of about four poor performances in big games. He's good enough to draw favorable comparisons to Roger Staubach, and there certainly aren't many better UDFA's out there.
9 - Matt Ryan (92). Matty Ice is the last guy under 100, and just sneaks in. I'd hesitate to call him "elite" on that basis, but I have no problem calling a guy who ranks in the top 10 all-time in Comp%, Int%, and Yds/G a member of the top 10 in the league currently.
10 - Carson Palmer (130). Carson Palmer. Really? How soon they forget. Palmer was once picked ahead of Roethlisberger, Rivers, and Manning the Lesser. Dude has skills. Not, like, enough skills to be mentioned in the same breath as Matt Ryan, who is a whopping, 40 places higher than him in QBOR, but enough to be mentioned as the premier second tier quarterback. And the guy got his knee shredded by the Steelers, c'mon show some sympathy.
11 - Matthew Stafford(136). How much of this is Megatron? I don't know. A lot of it is Yds/G, in which he's number 1 all-time. He's only 95th in Yds/Att.
12 - Jay Cutler(159). Jay Cutler, good at nothing, not terrible at anything. His middle of the road stats across the board give him the lead over half the AFC North by the narrowest of margins.
13 - Joe Flacco(160), Dalton(160). Is this crazy or what? Dead tie. Both make their best showing in the Yds/G and Int% fields, with Flacco better in Int% (not bad at 15th all-time) and Dalton better in Yds/G. Both show up the weakest in Yds/Att with Flacco worse actually at 118th all-time (significantly the worst of all the quarterbacks examined) despite being 34th in Comp%. Maybe Ravens opponents shouldn't be quite so concerned about deep passes, and Ravens fans should be more concerned about losing Ray Rice.
15 - Eli Manning(176). Did I say something earlier about rings not making you a good quarterback? San Diego fans now rejoice in Manning's petulant refusal to play on the West Coast.
Well, that's all he wrote, folks. I know I left Russell Wilson and Colin Kaepernick off the list of franchise quarterbacks. Pro-football-reference doesn't have them listed on the all-time list, likely because they're a couple of young pups who haven't yet the thrown the ball enough to be fairly evaluated. Newton probably shouldn't be on here either, honestly, but you have to put the bar somewhere and he just made the cut. I look forward to hearing what y'all think, as this is not intended to be a definitive list. Rather, a valuable metric for creating such lists. As I indicated in some of the commentaries, arguments can be made to move a player up or down based on subjective evaluation like mobility or being in a hard/easy division. I think this does provide, though, a helpful standard.