Quarterbacks and Salary

Quarterbacks and Salary
Quarterbacks and Salary
Photo: USA Today Sports Images

Guest column by Kevin Kolbe

Every year, multiple teams in the league have to answer the question of "what is a player's value to our organization?" It's not an easy question for any position, but when it comes to quarterbacks, there doesn't seem to be a rational approach to determining that amount. Consider this: eight quarterbacks make between $18 and $22 million a year, an amount one might think is still a bargain for elite or even just very good players at the most important position in football, and I am not going to argue otherwise. After all, a reliable quarterback who can carry his offense against any defense does guarantee a certain level of success.

One might expect that between the best starters and the merely above average, however, there would be a significant drop-off, but this is not the case. After the top eight, there are another six quarterbacks making between $14 and $18 million a year.

Between $5.6 and $14.6 million a year, there is just one quarterback: Tom Brady. This is ridiculous for many reasons. Only one quarterback in a $9 million dollar range? This means that, aside from Brady -- who can "sacrifice" to give the Patriots cap room in part because he doesn't need to be the primary breadwinner in his family -- all quarterbacks fall into four categories:

1) A "top of the league" talent
2) An above average talent without whom the team would still suffer
3) A player on a rookie contract
4) A completely replaceable journeyman/borderline backup.

I find this logic to be flawed. After all, some players have to be "average." Is the gap between the "good" and the "just OK" really worth $9-plus million a year? Let's find out. Note that the following points do not take into account strength of schedule, quality of offensive "weapons," protection, or even coaching. This will be about production versus cost. Initially, I did this without taking rookie contract quarterbacks into account, as their pay is intended to be lower than their potential value. However, I knew that fans of teams with young passers would want to see how their young quarterbacks compared, and it would be interesting to see an objective assessment of those players' value before they sign their first big contracts (especially with many of them in talks for said contract).

[ad placeholder 4]

Before we begin, I would like to apologize to the Titans, Rams, and Redskins. None of your quarterbacks made it into this article, as I arbitrarily decided to look at only those quarterbacks who started more than eight games in 2014.

I began by using Adjusted Net Yards/Attempt, a reliable metric for passing efficiency, but pass attempts do not show a quarterback's entire role in the offense. For example, on a failed passing play, sometimes a quarterback will take the sack, sometimes he will throw out of bounds, and sometimes he will run for it. In addition to that, read-options and designed runs are a significant part of the game plan for some teams. Lastly, a few gimmick plays feature the quarterback as a receiver, and it would be unfair to ignore them given how exciting they are to watch. TANY/A (Total Adjusted Net Yards/Attempt) adds in rushing as well, and the formula is written out here:

TANY/A =      {passing yards + rushing yards - sack yards + 20 * (TD passes + TD rushes) - 45 * (INT + 0.5 * Fumbles) }
(passing attempts + sacks + rush attempts)

As you can imagine, most quarterbacks will see their average yards decrease with rushing numbers added in. A lot of this is due to fumbles, but rushing attempts tend to yield lower average gains than pass plays. For simplicity's sake, in the following table, I added interceptions to half of each quarterback's fumble total and listed them in one column called "TO." "Attempts/Game" includes all passes, rushes, sacks, and (if applicable) receptions. In the first column, "games," I rounded to the nearest half-game, since some quarterbacks entered or left games due to injury.

I followed this up by looking up the average cap hit for each quarterback displayed. After all, the point of this exercise was not to determine which quarterback was best, but to figure out what each quarterback's value was. The following scatterplot displays each quarterback's average cap hit by TANY/A.

As you can see, there are cap hits that seem to have no relation to performance. Let's see what things would look like if all of them did.

[ad placeholder 3]

Aaron Rodgers had both the highest TANY/A as well as the highest average cap hit. This is wonderfully convenient for me because it gives me a clear and objective starting point to the exercise. The difference between the highest cap hit (Rodgers, $22 million) and the lowest (Russell Wilson, $0.75 million) is $21.25 million. The difference between the highest efficiency (Rodgers, 8.17 TANY/A) and the lowest (Blake Bortles, 3.90 TANY/A) is 4.27 TANY/A. I'll be using those high and low values to determine quarterback value by efficiency. The better a quarterback's TANY/A, the higher his base value.

The problem with only using TANY/A to determine base pay is that it does not account for the demands placed upon a quarterback. It is easier to move on from a quarterback who rarely has to throw thanks to a strong running game or stout defense than it is from a high-volume quarterback who is forced to carry the offense and keep the defense rested, and I wanted the formula to reflect that.

To account for this, I created a new metric, which I have named PAYD (Pay by Average Yards and Demand). PAYD modifies base value by twice the percentage difference of plays per game (using Aaron Rodgers' 38.1 plays per game as the standard). For example, if a quarterback with a base value of $10 million per year averaged 41.9 plays per game (10 percent more than Rodgers' mark), his PAYD would be $12 million per year (20 percent more than his base value).

Player Games Plays/G ANY/A TANY/A Cap Hit Base Value PAYD PAYD - Cap
A.Rodgers 15.5 38.1 8.65 8.17 $22.0 $22.0 $22.0 $0.0
B.Roethlisberger 16 41.8 7.82 7.26 $14.7 $17.5 $20.8 $6.1
A.Luck 16 44.2 7.28 6.68 $5.5 $14.6 $19.2 $13.7
P.Manning 16 39.9 7.68 7.10 $19.2 $16.7 $18.2 ($1.0)
D.Brees 16 44.7 6.77 6.42 $20.0 $13.4 $18.0 ($2.0)
M.Ryan 16 43.0 6.71 6.47 $20.8 $13.6 $17.0 ($3.8)
T.Brady 16 39.9 7.01 6.50 $11.4 $13.7 $15.0 $3.6
E.Manning 16 41.1 6.67 6.43 $16.3 $12.4 $14.4 ($1.9)
R.Wilson 16 38.3 6.72 6.56 $0.7 $14.0 $14.1 $13.4
T.Romo 15 32.7 8.11 7.49 $18.0 $18.6 $13.3 ($4.7)
J.Flacco 16 38.3 6.66 6.20 $20.1 $12.2 $12.3 ($7.8)
P.Rivers 16 40.2 6.45 5.96 $15.3 $11.0 $12.2 ($3.1)
M.Stafford 16 42.4 6.03 5.74 $17.7 $9.9 $12.1 ($5.6)
R.Tannehill 16 43.3 5.83 5.50 $3.2 $8.7 $11.1 $7.9
M.Sanchez 8.5 43.1 6.18 5.47 $2.3 $8.6 $10.8 $8.6
Player Games Plays/G ANY/A TANY/A Cap Hit Base Value PAYD PAYD - Cap
A.Smith 15 37.0 6.14 5.94 $17.0 $10.9 $10.3 ($6.7)
R.Fitzpatrick 11.5 33.3 7.15 6.51 $3.6 $13.7 $10.3 $6.7
K.Orton 12 41.3 5.69 5.41 $5.5 $8.3 $9.6 $4.1
B.Hoyer 13 37.4 6.11 5.70 $1.0 $9.7 $9.3 $8.3
D.Stanton 8 34.5 6.22 5.80 $2.7 $10.2 $9.2 $6.5
C.Kaepernick 16 39.6 5.58 5.42 $19.0 $8.3 $9.0 ($10.0)
J.Cutler 15 42.5 5.57 5.17 $18.1 $7.1 $8.7 ($9.4)
C.Newton 14 42.1 5.45 5.12 $5.5 $6.8 $8.2 $2.7
T.Bridgewater 12.5 39.0 5.46 5.27 $1.7 $7.6 $7.9 $6.2
A.Dalson 16 32.4 5.75 5.95 $16.0 $10.9 $7.7 ($8.3)
D.Carr 15.5 42.1 4.82 4.40 $1.3 $3.2 $3.9 $2.6
G.Smith 13 32.4 5.13 4.63 $1.3 $4.4 $3.1 $1.8
B.Bortles 13.5 43.4 3.81 3.90 $5.2 $0.7 $1.0 ($4.2)
J.McCown 10 38.8 4.30 3.93 $5.0 $0.9 $0.9 ($4.1)

As I mentioned earlier, 14 quarterbacks made between $14 million and $22 million last year. How did that cap hit compare to production? Poorly. Only six of those 14 quarterbacks saw production worth within $4 million of their average cap hit. Nineteen of 29 quarterbacks had a PAYD within the range of $7 to $15 million a year, whereas only two quarterbacks (Tom Brady and Ben Roethlisberger) see average cap hits in that range in real life.

Deciding whether or not to let a starting quarterback go is a difficult decision, one that I would compare to the decisions teams make on fourth-and-short. Coaches know that going for it and failing will earn them a lot of blame. Punting is just safer, and doesn't receive the same level of scrutiny, and so punts are called far more frequently. NFL general managers are doing the same thing. Look at the four quarterbacks under Andy Dalton, and you'll see exactly what I am referring to. The Bengals didn't pay Andy Dalton $16 million a year because he outplayed Brady, Philip Rivers, and Roethlisberger. They paid him because keeping him meant not risking his departure and getting stuck with someone much worse; it was the GM equivalent of a punt on fourth-and-short.

The best teams will always be the ones that know when to go for it and take a wise risk.

Kevin Kolbe is originally from Houston, Texas, and has been in the Air Force since December of 2001. This is his first article for Football Outsiders. If you are interested in writing a guest column, something that takes a new angle on the NFL, please email us your idea at Contact Us.


173 comments, Last at 12 Aug 2015, 9:21pm

1 Re: Quarterbacks and Salary

If I'm reading this correctly -- Mark Sanchez provides the best value of any QB who isn't on a rookie deal?

4 Re: Quarterbacks and Salary

You could say that, but bear in mind that Sanchez was signed as a backup to Foles and became a starter through injury. Hoyer actually began this last season as a starter, though began his contract in Cleveland as a backup as well (to Weeden).

In terms of QBs who signed contracts with the expectation of being a starter, Ryan Fitzpatrick would have been the best value not on a rookie deal.

59 Re: Quarterbacks and Salary

Hoyer is second, and was playing for the Browns. And for being on the Browns, he played quite well, I might add.

So in all their wisdom, the Browns hand the keys of their Pontiac Aztek to Johnny Manziel and... you know.

61 Re: Quarterbacks and Salary

In fairness to the Browns, they were on a slump and effectively out of the playoff hunt. They'd just lost a game despite getting TWO defensive touchdowns. Fun fact: only the Browns have lost a game despite two defensive TDs. They've now done it twice.

Manziel was an all or nothing gamble.

I'm actually really looking forward to him as a starter in Houston. Bill O'Brien works really well with known quantities at QB. Look at what he did with Fitz.

91 Re: Quarterbacks and Salary

Hoyer accounted for the bulk of that slump, as well as he played early on in the season. Not only that Colts game, but the other games down the stretch (Houston incidentally, and he had 3 INTs against Atlanta's terrible defense despite lucking into that win.) The run D holds a lot of blame, too.

It seemed like DVOA and DYAR also liked Hoyer unusually well. Most other stats/analysis seems to show Hoyer/McCown as basically a wash.

93 Re: Quarterbacks and Salary

McCown was dogshit, though bear in mind that he had no OC that year. Hoyer had an OC, but he didn't have quality receivers. Again, I am not bashing the Browns, as their situation was worth trying something bold.

104 Re: Quarterbacks and Salary

PFR turns up 29 games since 1940 where a team lost despite scoring 2 TDs on INT or fumble returns:


These may include some fumble return TDs by the offense, though.

2 Re: Quarterbacks and Salary

This is an interesting analysis. I have one suggestion for you for a meaningful way to extend it. Your TANY/A and PAYD metrics are both linear. Very often in a market where you pay for incremental value you are faced with diminishing returns. I would expect the marginal cost to go from a QB with TANY/A 7 to 8 to be greater than to go from 3 to 4, for example. I think if you looked at a few more years of data and threw out rookie contracts you could measure that marginal cost curve. That would in itself be an interesting result, but you could go further and compare actual salaries against this curve to see who is outplaying or under-performing when measured against a curve that depicts how GMs think about pay vs. performance.

6 Re: Quarterbacks and Salary

Thanks! I did strongly consider a non-linear approach, but I felt that doing it that way required me to make too many assumptions. I could have used actual salaries from past years, but I actually wanted this metric to be completely separate from contract trends, as the point of it was to argue against the blind following of them.

I *hope* to put something together for next offseason that ties pay to an expected loss of production without the QB in question, since the basis of value is the result of not having said player. Obviously, this site has some great metrics to use as a starting point for that, like DVOA and DYAR.

3 Re: Quarterbacks and Salary

Rodgers is underpaid, by a lot.

If you cash things out in terms of a quarterback's impact on winning, I'd guess that each point of TANY/A is worth about $10 million (assuming a full workload), not the $5ish million that is used here.

26 Re: Quarterbacks and Salary

He very well might be, as 7 of the 20 other non-rookie-contract QBs come out with less surplus value than Rodgers's 0.0 (as he's the baseline).

Though some of the ones with less - notably Brees and both Mannings - are within two million of Rodgers, so he might not be underpaid "by a lot".

Also, the point was not to put a value on each point of TANY/A, but to figure out relative differences.

52 Re: Quarterbacks and Salary

Roethlisberger was the only QB in 2014 with a PAYD comparable to Rodgers'. One could argue that Romo, having the second highest TANY/A, would have been closer had he been healthy at the start of the year.

Also, spot on regarding the point of the article.

63 Re: Quarterbacks and Salary

Putting a value on each point of TANY/A is a step on the way to comparing expensive elite QBs with cheaper pretty good QBs. Would you rather have Aaron Rodgers at his salary, or Ryan Tannehill at his salary? I say Rodgers by a lot, this "PAYD - Cap" formula says Tannehill by a lot. Because the formula counts the $19M difference in cap hit as being worth more than the 2.7 point difference in TANY/A, and I think the opposite.

5 Re: Quarterbacks and Salary

Where did you get the data for the cap hits? They don't match up with what I'm seeing from other sources. For example, both OverTheCap and Spotrac have Rodgers with a cap hit of $17.5 million for 2014, and Brady's cap hit is $14.8 million.

I definitely think this analysis is interesting, but I also think that comparing QBs contracts of veteran starters to rookies and backups is an apples-to-oranges comparison. I think two separate sets of analysis, one that looks at veteran starters on their second contract or later and the other that looks at rookies and backups. While including their data is interesting from a fan standpoint, comparing the contracts of Wilson and Luck to Rodgers and Manning really doesn't make any sense, and that's what this analysis is doing. It would be better, in my opinion, to do a separate analysis on both categories.

8 Re: Quarterbacks and Salary

"While including their data is interesting from a fan standpoint"

And that's kind of why I did it.

As for the cap hits, I averaged out the salary from the entire contract, since some are structured quite oddly.

11 Re: Quarterbacks and Salary

I'm still not entirely sure how you got $22.0 million for Rodgers then. The highest cap hit over the course of his current contract is $21.1 million in 2019, which is still less than what you have down for him.

13 Re: Quarterbacks and Salary

The 22m average was from overthecap.com, as were all cap numbers. I see what you are referring to and have to assume that the reason was due to an extension that boosted his pay for the previous contract on top of the new numbers.

7 Re: Quarterbacks and Salary

The Bengals didn't pay Andy Dalton $16 million a year because he outplayed Brady, Philip Rivers, and Roethlisberger. They paid him because keeping him meant not risking his departure and getting stuck with someone much worse; it was the GM equivalent of a punt on fourth-and-short.

I disagree with this analogy in one very important respect: paying Dalton does not preclude the possibility of still searching for and finding a better QB and cutting him later. I don't think it's necessarily a bad idea to overpay for a mediocre QB, unless the contract locks you in or puts you in cap hell (neither of which is true in the Bengals' case).

9 Re: Quarterbacks and Salary

"[P]aying Dalton does not preclude the possibility of still searching for and finding a better QB and cutting him later."

Depends on what you mean by later. If they cut him this year, the hit is 13.6m, which does "lock them in" at that point. Next year it drops to 7.2m, which is still fairly significant, though acceptable if he really drops off in 2015 (which I think is very unlikely. If anything, I'm expecting improvement due to healthier receivers). 2017 could be the do or die year, as the Bengals would save 9.9m by releasing him. In 2018 the savings jump to 13.9m, also pretty tempting.

12 Re: Quarterbacks and Salary

The Bengals owe Dalton $10.7M in new money in 2016. They'll cut him after this season if they think he isn't worth that (which does seem unlikely, but would be easy to do if they don't think he's a capable starter). If they cut him at the end of this season, then they will have paid him $24.2M for a one-year extension (which is a steep price, but not backbreaking).

In 2017 they owe him $13.3M in new money. If they cut him after the 2016 season, they will have paid him $34.9M for a two-year extension.

25 Re: Quarterbacks and Salary

And yet in theory if you drafted a rookie QB Dalton wouldn't have to be cut, you could carry him as an expensive backup for a while without altering your cap situation at all, and eventually trade him. The real issue with this is that NFL teams feel having a potential starting QB on the bench creates too much of a distraction. Hence, if you draft a new QB with potential starter pedigree, that implies whoever was there will immediately go to the bench or out the door.

Who, me?

14 Re: Quarterbacks and Salary

The reason even average starters get insane money once they've proven themselves is that they have virtually unlimited leverage in negotiations, because the best QB available in free agency is typically a barely adequate journeyman, or coming off a major injury, or both. No team would voluntarily turn the QB position into a serious liability when they have a chance at making it a strength (or at least not a weakness), even if it costs a bunch of cap space. The cap space you save on a cheap journeyman QB would never get you enough talent at other positions to be worth turning the most important position into a major weakness.

If you don't already have an average/good QB, you try to draft one and maybe use a cheap journeyman as a stopgap while they develop. In which case, they're either cheap because of the rookie contract, or cheap because they're easily replaceable. But once you have a QB who is a proven commodity, and is average or better, they are getting paid top dollar as soon as their rookie contract is up. Otherwise, you go from [average/good QB] to [whatever you can scrounge off of free agency] - and nobody in their right mind is going to do that over $9 million in cap space.

28 Re: Quarterbacks and Salary

The only way out if for these journeymen, say like Sanchez, to start leading successful teams and playing adequately. That would break the stalemate and then maybe teams would be willing to lowball the Dalton's of the league.

Who, me?

35 Re: Quarterbacks and Salary

Well, if you look at Sanchez, Orton, and Fitz last year, then we are on our way there. In addition, Charlie Whitehurst didn't look bad at all either, but he didn't meet the 8 game limit I imposed. So I think that there is evidence of that, but the narrative still exists.

This year, look at the Hoyer/Mallett fight. Neither QB is great. Hoyer's numbers last year with the Browns are pretty bad, and Mallett was #3 on NE's depth chart when he got traded. However, I am confident that if Hoyer starts he'll finish top 12 in TANY/A, and Mallett will be top 15. I'm not saying that Hoyer is that much better (though I do prefer him), but Mallett is harder to gauge.

So if the Texans put up two decent offenses in a row with different 3-5m dollar QBs, then why shouldn't Dalton and Cutler and Kaepernick get lowballed?

92 Re: Quarterbacks and Salary

The interesting thing is that Mallett looked quite a bit better head-to-head last year, and is expected to have much more upside being younger and having more pure athleticism/size/arm strength.

70 Re: Quarterbacks and Salaryb

Agree, but that outcome (journeymen QBs becomign viable starters with which you can produce a winning team) seems more dependent on coaches and GMs than on the journeymen QBs themselves. For the most part, a guy like Fitzpatrick is what he is-- his competency level is not going to shoot up dramatically. However, it is theoretically possible for a GM and coach to create a winning roster with only an average or even slightly-below-average QB (e.g. 2002 Tampa Bay, the constantly-cited 2000 Ravens, Rex Ryan's playoff Jets teams) via strong defense, carefully-constructed scheme, etc. If enough teams do this, then it diminishes the value of the quarterback position and the leverage for guys like Dalton goes down.

Of course, everything in recent NFL history tells us that the trend is going in the OPPOSITE direction-- increased reliance on the QB for success, rather than deemphasis of the position.

I guess the other possibility for decreased QB leverage is the Chip Kelly Truthers argument-- a wave of brilliant coaches create schemes that are so user-friendly for QBs and so fullproof in their ability to get offensive playmakers open that QB talent becomes less important and more interchangeable.

78 Re: Quarterbacks and Salaryb

For the most part, a guy like Fitzpatrick is what he is-- his competency level is not going to shoot up dramatically. However, it is theoretically possible for a GM and coach to create a winning roster with only an average or even slightly-below-average QB (e.g. 2002 Tampa Bay, the constantly-cited 2000 Ravens, Rex Ryan's playoff Jets teams) via strong defense, carefully-constructed scheme, etc. If enough teams do this, then it diminishes the value of the quarterback position and the leverage for guys like Dalton goes down.

Trouble is, it is extremely rare for a defense to be so good that $9 million spent in free agency would make them as good as the 2000 Ravens. For the vast majority of NFL teams, that’s just not a viable strategy. And the Daltons of the world know it.

Of course, everything in recent NFL history tells us that the trend is going in the OPPOSITE direction-- increased reliance on the QB for success, rather than deemphasis of the position.

Exactly. That’s why lowballing a solid QB is a terrible idea.

I guess the other possibility for decreased QB leverage is the Chip Kelly Truthers argument-- a wave of brilliant coaches create schemes that are so user-friendly for QBs and so fullproof in their ability to get offensive playmakers open that QB talent becomes less important and more interchangeable.

Fun fact: During Chip Kelly’s tenure as coach, only one of his QBs has had a positive DVOA. And that would be the one he just traded away. So much for all the talk about Chip Kelly’s “user-friendly” system.

16 Re: Quarterbacks and Salary

The real problem stems from the complexity of the contracts, making comparison even among the same position difficult. The cap also keeps changing and so deals like the one Rodgers signed may in fact have him underpaid within 3 years in. Add in the fact that teams structure the contracts in such a way to enable them to have flexibility. Take the dolphins and Suh. I believe he makes pretty low money in year 1, but his year 2 is enormous.

22 Re: Quarterbacks and Salary

You are right about Suh, and that isn't that uncommon. That was largely why I did average cap hit as opposed to yearly. As for Rodgers, one could argue he'll be underpaid by the start of 2016, once Luck and Wilson get their contracts.

34 Re: Quarterbacks and Salary

The best way to do it would be to analyze each contract individually, but I guess that might introduce bias. For example, not count Dalton's last two years but do count Rodgers'? By the way, there has to be some kind of mistake in overthecap's numbers regarding Rodgers. If you'll notice, the guaranteed base salary amounts don't seem to be counting towards the cap number, which seems wrong and would account for the higher per-year average.

Who, me?

38 Re: Quarterbacks and Salary

Yeah, someone else asked me about Rodgers as well. Short answer, I trusted overthecap. Speculative answer, extensions can be a bit weird.

I see your point in looking at "contract years" for context, but since I plan on running this metric every year, eventually those kinds of trends will become apparent.

18 Re: Quarterbacks and Salary

I'm not convinced that releasing a significantly underperforming quarterback and signing a replacement-level (ZERO) player at QB, then signing other players to fill the cap space improves the team. First, I doubt it is possible with a player like Flacco. Second, some competitor gets to sign Flacco at his nominal value and then have a correctly priced good QB. Is the 40% savings worth dropping the 60% performance at the QB position? Or does $1M buy more performance at the QB position than at guard? I suspect that performance of $1M at QB is not properly normalized against other positions, not that I can prove it.

20 Re: Quarterbacks and Salary

Funny you mention the replacement player risk, as I am trying to put together a formula to determine when a team would be better off doing just that.

Also, this is more about the decision to re-sign QBs to big contracts rather than cutting them. Cap hits for QBs can be pretty nasty, after all.

19 Re: Quarterbacks and Salary

I don't think front offices are making poor decisions paying the 'middle tier' of QBs - the likes of Dalton/Flacco are probably being paid the correct amount given their importance to their team and the scarcity of their skills. More, the top tier of QBs are drastically underpaid, which is simply a function of having an artificial construct like the salary cap in the market. Rodgers deserves a lot more money relative to his peers, but paying him a lot more would quickly begin to be sub-optimal as it damaged the rest of the roster. He isn't operating in a free labor market. (There's also the issue of diminishing value of additional salary. Brady quite obviously eschews a few $million a year in order to play for a good team in familiar surroundings, and others may well be doing likewise).

21 Re: Quarterbacks and Salary

Is Dalton a scarce talent, though? How many backup QBs showed potential but only got one bad team to try with? Did you notice Fitzpatrick's TANY/A for last year? It was the 7th highest of all QBs with 8 games or more, and he was paid a whopping 3.6m a year. On top of that, the Texans then traded him away for a late pick. And I'm not bashing Dalton or Flacco. Hell, I actually kinda like Flacco. I hate his contract, but that has more to do with a VERY odd structuring than the amount.

As for the top tier, I did say I wasn't going to argue about the value of the best at the position, just the other QBs compared to them. I do think Rodgers could be paid more on another team, and Brady could have earned a lot more as well, but...priorities.

23 Re: Quarterbacks and Salary

It's interesting. It looks like Orton, Sanchez, and Fitzpatrick were available for nearly nothing last year and outperformed quite a few QBs making way more money.

Just to throw grease on the fire, it's hard to say the Bears would have been worse last year starting Orton and spending more on defense than what they had.

29 Re: Quarterbacks and Salary

Well, they might not have gotten significantly better, but I don't think they would have been worse off either.

Also, the d-line was the only competent unit on the defense last year. $10 million in cap could have brought in potentially a competent linebacker and safety. Probably not, as I don't think this coaching staff could correctly assess linebackers at all and weren't good at safety either. You never know though.

72 Re: Quarterbacks and Salary

I think there's a bit of a cause-and-effect question on Sanchez, at least. Was Sanchez truly undervalued relative to his production potential, or does Chip Kelly's system make QB talent less important by creating an extremely user-friendly setting for a quarterback, effectively devaluing the position?

I kind of lean toward the latter, but you could certainly rebut me by citing Kelly's immediate decision to seek an expensive "upgrade" over Sanchez/Foles in Sam Bradford.

76 Re: Quarterbacks and Salary

Well, it could be that Rex's system and bad receivers made him look worse, with Chip's making him look better. His value is probably in the middle. Admittedly, though, I want to stay out of evaluating QB talent. There's too many folks who are a lot better at that (especially on this site). I'm just a guy who likes playing with numbers.

68 Re: Quarterbacks and Salary

Rodgers, Manning, Brady, etc. keep making the playoffs every year, which means that we are pretty far from the point where an elite QB's salary is damages the rest of the roster by enough to cancel out what the elite QB adds.

73 Re: Quarterbacks and Salary

I don't feel you are trying to learn anything here. For example, the most consistent in terms of playoff appearances is Brady, who does not have an elite QB salary, or even a good QB salary. Meanwhile, Manning and Rodgers actually have fairly poor post season records. Manning is 11-13, and Rodgers 6-5. Bear in mind that you can win up to 4 playoff games a year, but can only lose 1. Brady is 21-8.

So yes, I would say there is a visible effect from salary on team success.

74 Re: Quarterbacks and Salary

I don't really buy that argument. After all, Manning has also won a ton of games as has Rodgers. They are perpetually a lock to win their divisions over and over. The playoffs have much more randomness at work and Manning and Rodgers have both been victims of it(Hail Mary TDS and Onside Kick recoveries).

Brady being this successful could also be a testament to the Patriots franchise and their ability to be innovative and their coaching staffs being so top notch.

75 Re: Quarterbacks and Salary

I agree that the less you pay your elite QB, the better you can make the rest of the team.

The question I'm interested in is: how much do you have to pay your elite QB in order to cancel out the benefit of his eliteness and leave the team average?

As a ballpark estimate, I'd guess about $40 million.

It's pretty clear that the answer is significantly more than what Brady, Manning, or Rodgers have been getting. Excluding each of their first 5 NFL seasons (which roughly spans their rookie contracts), they have made the playoffs in 25 out of 25 seasons in which they played 8+ regular season games.

77 Re: Quarterbacks and Salary

That is one hell of a question, and I don't have a formula to determine that. Obviously, as you pointed out, their teams are all well above average, and all of them could probably drop another 10m and be above average, barring health concerns.

30 Re: Quarterbacks and Salary

Can anyone think of teams that tried to build an offense with not-bad-but-not great QBs, and surrounding them with high level offensive talent?

I have two that sort of fit the mold.

The GSOT Rams, especially the Bulger version, but really I don't think anyone was expecting Trent Green or Kurt Warner to grow as they did.

The other would be the Denny Green and Mike Tice coached Vikings who seemingly plugged in different journeymen QBs every year and let Carter and Moss carry the offense (along with a pretty good offensive line).

41 Re: Quarterbacks and Salary

I would argue the early 2000's Chiefs fit that mold...but maybe I have a lower opinion of Trent Green than most people do.

Scott Mitchell had a great season in 1995 (#2 DYAR and #5 DVOA) primarily because he had a great supporting cast, but it wasn't sustainable, which, like the Rypien example, begs the question of whether teams should attempt to do that instead of drafting a QB with a relatively high pick.

44 Re: Quarterbacks and Salary

The chiefs to me will always be defined by that incredible O line and Gonzo. Just think - Roaf, Shields, Bryan Waters, and John Tait were all good to great to hall of fame level players. I probably could get three years running behind that line.

"There likes will not be seen again"

47 Re: Quarterbacks and Salary

Those Vikings teams had Brad Johnson (turned out pretty decent), Randall Cunningham (very good and surrounded by talent for the first time, just at the very tail of his career), Jeff George (possibly the NFL's most talented QB ever, and it's biggest head case), followed by a young Dante Culpepper, who was pretty decent until his body fell apart.

50 Re: Quarterbacks and Salary

Cunningham was well past his prime by the time he got to the Vikings and I think Brad Johnson's face is in the encyclopedia if you look up "journeyman quarterback".

Edit: the whole point is that you can't have a complete scrub at QB, but finding one that's just alright instead of great and trying to make that work.

152 Re: Quarterbacks and Salary

But that premise is built on the idea that it is possible to find non-scrub QBs in the FA market. I feel like this is a baseball idea where they have a deep pool of replacement level talent due to extensive developmental leagues and either relatively interchangeable positions (position players), or deep rosters (pitchers).

In football the only significant developmental leagues are the NCAA and the CFL. The NCAA being age restricted and the CFL playing with slightly different passer friendly rules. Finally, the QB is the most unique position on the field where only a few QBs, (Wilson, Kaepernick, and Newton) could play other positions and effectively zero other players could play QB.

The reason that Dalton gets the money he gets is that next best alternative is... Mark Sanchez and their are probably only three players at that level. Sanchez, Fitzpatrick, and Tavaris Jackson assuming Kyle Orton stays retired. So the choice is not actually between Dalton and capable it might be between Dalton and Sanchez/Fitzpatrick/Jackson, but is more likely between Hoyer, McCann, Kolb ... etc. and Dalton.

So, basically Dalton gets his paycheck because replacement level QBs in the NFL are in fact scrubs.


I remember when they were the Sea-chickens.

156 Re: Quarterbacks and Salary

It doesn't have to be free agency. You could trade for Derek Anderson, Mike Glennon, Nick Foles, or Drew Stanton. None of them could cost that much in trade value, and all of whom outperformed Dalton by DVOA last year.

Most of whom I would say played with worse offensive teammates.

By DVOA, Dalton was the 25th best QB last year. With what looks like a decent offensive line, running game, and one of the best receivers in the game.

32 Re: Quarterbacks and Salary

Isn't this the effect of the salary cap? Any qb who is even vaguely describable as a starter is too valuable to let hit free agency, as a result they are all either tagged or re-signed early in a negotiation based upon the value of the tag.

I think this is likely the result of there not being enough quarterbacks, if there was a surplus of starting quality qbs then a true market that responded to player quality might form because. However, while we have a league of quarterback scarcity teams will continue to do all they can to hold onto such vital assets.

36 Re: Quarterbacks and Salary

Totally agree with the scarcity argument. It's all well and good to argue for teams to pick up a theoretical replacement level QB instead of paying out the nose for Dalton or Smith or Flacco, but then you find yourself starting Josh McCown and your season is over before it even starts. At any given time there are maybe 25 QBs out there that you would actually be comfortable with as your starter. The teams that have one of those 25 see what the other seven QB-poor teams look like (the Cleveland Browns) and do whatever it takes to make sure they don't end up one of them. That's how Alex Smith brings home $15.6M in 2015 even though his name is synonymous with bland mediocrity.

54 Re: Quarterbacks and Salary

This is exactly it. The Browns, Jaguars, Bills, and Bucs continue to draft the Brady Quinns, Brandon Weedons, Blaine Gabberts, Trent Edwards, and Josh Freemans of the NFL. They might build an otherwise decent team only to start with someone like Colt McCoy, Chad Henne, Kyle Orton, or Bruce Gradkowski. It's not pretty.

56 Re: Quarterbacks and Salary

This isn't about good v bad. It's about value. Kyle Orton at 6 million is better for a team than most of these QBs at 20 million. Ryan Fitzpatrick was the 7th most efficient QB last year. Was it because he was an upper tier player? No, of course not. But at 3-4 million a year, Houston was able to have him throwing to Andre Johnson and DeAndre Hopkins, handing off to Arian Foster, and sitting behind an above average O-Line. And he didn't even cost us a draft pick.

Basically, QB decisions need to be made rationally, instead of emotionally.

60 Re: Quarterbacks and Salary

Actually, I am more confused by the Flacco signing than against it. The 6yr/120m gets thrown around a lot, but is that really what he got?

In 2013, his cap hit was 6.8 million.
In 2014, his cap hit was 14.8 million.
In 2015, it will be 14.5 million.

Does this look like a 20 mil a year contract to you? Now, after 2015, it gets WEIRD.

2016: 28.55m
2017: 31.15m
2018: 24.75m

As bad as that seems, it is actually worse. The dead money in 2016 is 25.85m. In 2017, the worst year, it is still 15.3m. Baltimore will have close to no leverage. Despite that, I feel a deal will be worked.

66 Re: Quarterbacks and Salary

I agree, too, but these QBs simply don't stick around. Fitzpatrick is moving on, Sanchez won't start, Hoyer was let go. Do you think there might be a non-football reason for not going with these QBs? Commercial, I mean. Or maybe if they actually stuck with them they'd ask for their 16 million and you'd have to start over? It's certainly a very strange state of affairs.

Who, me?

67 Re: Quarterbacks and Salary

Well, I won't attempt to explain much of what Cleveland does, except to say that it has to difficult consistently coming up with new and exciting ways to fail.

As for Fitz, it is because Houston was able to acquire Hoyer for a price comparable to Fitz and then trade Fitz away, effectively buying a late round pick without really losing anything. In addition, he may slightly prefer Hoyer due to their time together.

As for Sanchez, Bradford is not exactly a guarantee in performance or health. Sanchez might actually be the highest paid backup in the league this year, and may end up becoming a starter again down the road.

94 Re: Quarterbacks and Salary

Hoyer became a malcontent late in the season with struggling to hang on to and then losing his starting job. The expectation with McCown was that he would be more of a "good soldier" since he's been a mentor in the past.

37 Re: Quarterbacks and Salary

Actually, the cap does more to prevent this. It forces teams to decide how much a player is worth relative to other players of different positions, because there is a limit on how much they can spend. Otherwise, Dallas would be making the playoffs every year.

Good players at other positions get lost all the time. Revis just hit free agency. It isn't about lack of talent, either, as teams frequently start sub-par players over paying an overpriced starter. The thing is that bad QB play gets noticed by fans more, so is more likely to result in coaches and GMs getting fired.

Also, if you look at the chart, are those QBs really vital? Look at the comments. There is a serious argument that the Bears would have been better without Cutler, and I agree with it.

39 Re: Quarterbacks and Salary

I feel like Cutler is still a matt stafford level qb. I think last year he had a down year for some weird reason, but his true value is better than an Orton. Honestly, people seem to have a warped view of Orton. He was successful in Denver, but pretty mediocre everywhere else.

I know hes used as the quintessential average starter, but players like Dalton and Smith to me are better than Orton. And Cutler is better than those 2.

46 Re: Quarterbacks and Salary

I agree that Cutler is essentially on par with Stafford. The difference is that the Bears had a choice in how much to pay Cutler and the Lions were stuck with their draft slot.

The other difference is that Stafford is 27 and should be just entering is peak while Cutler is 32 and should be just leaving his peak.

Again, though. The question isn't whether or Orton or Cutler is the better QB. The question is whether Orton + $12 million dollars of empty cap is better than Cutler.

86 Re: Quarterbacks and Salary

Because you mentioned Stafford only being 27—and I think he got his extension at 25—I think it's worth pointing out that two of the least valuable QBs on the list, Dalton and Kaepernick, were also extended at 26. Part of the misidentification of their value/talent on the part their teams (or the market as a whole) was the assumption that they were going to get better, and both had pretty rough 2014s.

This isn't an excuse for the money paid to Cutler, Smith, and maybe even Flacco, though.

87 Re: Quarterbacks and Salary

Well, Stafford's extension was largely due to a bad cap situation in Detroit, so they kind of just postponed a problem. Kaep's was really odd in that they didn't know much about him, and he had a year left on his contract. It's worth noting that Kaep can be let go pretty easily, making it less Cutler-y.

As for Flacco, Ravens fucked with the wrong guy. I really don't think he's in the same boat as Cutler and Kaep, or even Stafford. To date, Flacco has yet to have a cap hit above 15m.

88 Re: Quarterbacks and Salary

We can also semi extend this to people like Dalton and Tannehill. We don't know how much of their extensions are also based on the view that they are likely to improve.

Cutler and Smith are two examples of teams in a situation of overpaying for known quantities.

89 Re: Quarterbacks and Salary

Completely agreed on Cutler and Smith. In Smith's case, it was coming off a solid year and amazing playoff performance. Public opinion was well on his side.

I won't speak to Dalton, but I can say that there is clear expectation for Tannehill to improve, and based on his first 3 years, it is a reasonable expectation. I don't know if he will improve enough to warrant his contract, but it does seem likely that he will at least come close. I don't see Miami regretting this decision, barring injury.

90 Re: Quarterbacks and Salary

This is just food for thought. The argument that high salaries will limit talent seems reasonable in theory. Actually, more than reasonable. That said, and this is something no one as of yet seems to understand, but why is parity much worse today than it was back when Qbs were making less than they were today?

In fact, as qbs have been paid more, passing offenses have become more and more prolific. We'd expect the reverse.

102 Re: Quarterbacks and Salary

Flip the correlation into a causation (that QBs are being paid more because passing offenses are more prolific) and the two make sense. Rules changes have made passing offenses more dominant, a statement I don't think needs any defense here, to explain the original cause.

98 Re: Quarterbacks and Salary

I'm with you on Tannehill, not unreasonable at all to bet on improvement, but I also don't feel it was unreasonable for San Francisco and Cincinnati to feel the same way about Kaepernick and Dalton last offseason. In fact, I thought Kaepernick signed a really bad deal at the time. (In hindsight Kaep was actually still a bit of an unknown quantity, as you pointed out.)

Also just wanted to point out, since you mentioned that SF can get rid of Kaepernick fairly easily after this season—Dalton and Tannehill have virtually the same dead money hit if they were to be cut two seasons into their extensions. (Lower, actually.)

99 Re: Quarterbacks and Salary

With Kaepernick, though, they had another year, not including a tag, so they could have been more patient.

Yeah, the Tannehill contract is surprisingly team friendly, but I don't suspect they'll strongly consider cutting him, though it would be relatively painless. As for Dalton, I worry that the Bengals will want to cut him after 2016 (again, expect better numbers from him in 2015), but will keep him to avoid even the slight cap hit, unless a young QB really impresses.

153 Re: Quarterbacks and Salary

Reading this question I have to ask myself, if the defense of last years Buffalo team gets paired with Dalton due they make the playoffs and conversely if last years Cincy team fall from playoff contention with Orton?

What are the marginal improvements that Cincy could have made with Orton that would have been superior to Dalton?

Also, I believe we're using Orton because of the tier 3 non-starting QBs he or Sanchez or Fitz IS the next best alternative.

I remember when they were the Sea-chickens.

163 Re: Quarterbacks and Salary

Which question are you referring to?

I can say that Cincy would have made the playoffs with almost ANY QB last year. Look at their schedule and you'll see what I mean. The NFCS and AFCS were the divisions with the two worst records last year, and AFCN played both. As far as improvements, not much of note if we are only talking 2013, as Dalton wasn't much more expensive for that year. I think Michael Johnson would have stayed, though.

As for Buffalo, harder to say. They had a lot of games that could have gone either way.

111 Re: Quarterbacks and Salary

I should have been more clear, I meant the structure of the salary cap ie the franchise tag and the machinations of the CBA. And mostly the tag.

Can anyone explain why the tag and compensatory picks exist? Comp picks just make life easier for teams that are already good and the tag allows teams to hold onto players that they otherwise wouldn't be able to hold onto.

125 Re: Quarterbacks and Salary

But which management? I can see how it's handy for a team with good young players to limit their access to the market but how has it been helping the Bucs, Raiders or Jags in the past few years as they have tons of cap room but very few decent, young players ever hit free agency?

I don't see how it works to increase parity and from the perspective of the league as a whole it doesn't increase profitability, just reduces player movement.

As for comp picks, all they do is reward teams that lost free agents but why do they need to be rewarded? Every team has the same cap, why give picks to teams that can't keep all their players?

135 Re: Quarterbacks and Salary

Tagging is only used when a team wants to keep a player with whom they can't agree to a contract. If the cost is too high, they won't use the tag. Even if it's a nice salary for the player, he clearly could do better on the open market.

In ownership's perfect world, the team that signed a free agent would give up their pick to the player's old team. This would be a serious constraint on signing free agents. The current system mostly guarantees that teams that can't/won't sign a departing free agent get something for their troubles, with the advantage that teams who do sign players aren't directly penalized.

118 Re: Quarterbacks and Salary

I feel like compensatory picks are a vestige from before management really understood what revenue sharing and a salary cap meant. They must have been worried about big market teams buying up players and that the smaller teams needed something to compete.

The tag is for fans, team stability, and the crazed panic about free agency in the late 80s. Making sure a team can hold onto one player they really want promotes keeping teams together.

49 Re: Quarterbacks and Salary

I love this analysis. I would marry it, if the Supreme Court said it was ok.

I have come to the conclusion that most NFL owners don't have it as their number one goal to win the Superbowl. They just want to put a good enough team on the field to keep the fans in the stands. It is increasingly difficult to win the Superbowl, or even make it to the Superbowl, without a great quarterback. The last one was maybe Flacco in 2013, I would consider him to be pretty average, but he played way above his level in the playoffs.

I have slowly come to the conclusion that the most valuable thing to a franchise trying to find a great NFL quarterback isn't money or first round draft picks, it's NFL starts and first team practice reps. A team without a great quarterback only has 16 trials to try to figure out what they have. A team that starts Alex Smith or Matt Cassel for 16 games is not trying to win the Superbowl. As you say, they care more about downside risk than upside potential.

If I were trying to win the Superbowl and I didn't have a great quarterback, the first thing I would do would be to trade my average quarterback for whatever I could get.

With how important the position is and how bad teams are at evaluating QB talent, I don't think it makes any sense to take any other position in the draft while you still need a QB. Take a QB in the first round. Take a QB in the second round. Take a QB in the third round. Take a safety in the fourth round. Take a QB in the fifth round. Bring them all in for camp and let the best one start on opening day. If they all tank, repeat next year.

This is somewhat extreme maybe, but everyone says that QB is the most important position on the field, by far. Well, why draft a backup guard when you need a QB? Get more guys to try out. Bring in a bunch of UDFAs, fill out your practice roster with nothing but QBs.

55 Re: Quarterbacks and Salary

Thanks, glad you enjoyed it!

I would like to point out that I don't believe that you need a great QB to win. I believe you need great value at QB to win. Rex Grossman was far from a great QB, but his VERY low cap hit gave the team a lot of options.

Flacco, who is above average, was great for the team on his rookie contract, but now that he is paid rather highly, isn't looking as great. The same could be said for Dalton as well, whose playoff streak is impressive. That said, I don't think it will last much longer now that his cap hit is increasing (only 9.6 this year, though).

Alex Smith could easily win a team a Super Bowl...just not while being paid 17 million. Drop him to 10 where he belongs, and the Chiefs go back to the playoffs, where most anything is possible.

The logic of needing a great QB to win the Super Bowl is largely due to average QBs being paid like great ones.

58 Re: Quarterbacks and Salary

Well...I think a good qb keeps you in the hunt and is probably the most consistent thing in a sport that's incredibly volatile. The corner chart showed the perpetual swings in year to year performance.

That's one reason you pay these qbs...consistency. Criticize Alex Smith all you want, but you know he's going to give you a baseline level of mediocrity that is much better than volatile swings of mediocrity to horrid play like a Fitz.

62 Re: Quarterbacks and Salary

OK, nobody can argue that Alex Smith at $10M is better than Alex Smith at $17M (though based on OTC his cap hit is $15.6M this year and only $4.6M last year, not sure where you get your cap numbers). The issue is, can Smith be had for $10M? If the Chiefs had offered him $10M, it's pretty clear he ends up in Oakland or Cleveland or Jacksonville or Tampa for the $15M that the Chiefs aren't giving him. Would KC really be better starting Chase Daniel (of career 57% completion percentage) and buying $15M worth of free agents at other positions? I don't think they would be.

64 Re: Quarterbacks and Salary


I stated in the article that I used average cap hits. Also, I fully agree that the Chiefs would not have kept Smith with a 10m/yr offer. You are right that too many teams would have picked him up. Now, Chase Daniel plus 15m is where I might disagree. In the last two seasons he has played two games. Both were against the Chargers, borderline playoff team. He went 41/66 with 405 yds, 1 TD, and 1 INT. Not impressive, but not bad either. And I should point out that that was without the extra 15m dollars. Can you imagine last years Chiefs with Maclin AND 5m left over? I think Daniel would look okay.

65 Re: Quarterbacks and Salary

Maclin is a pretty good receiver and $5M buys you a basic competent starter. I don't think I agree that an upgrade at one WR position and a minor upgrade at another position is worth more than a pretty serious downgrade at QB. I think you're underestimating the difference between an average QB and a bad QB.

71 Re: Quarterbacks and Salary

I think the way to go with cap hit is to use a QB's average salary per year, assuming that the team will cut him (or restructures) in order to minimize his amount per year. For example, suppose that a 4-year, $40M deal looks like this:

signing bonus: $10M guaranteed
year 1: $4M (not guaranteed)
year 2: $6M (not guaranteed)
year 3: $7M (not guaranteed)
year 4: $13M (not guaranteed)

The team gets to choose whether this deal is for 1 yr $14M, 2 yrs $20M, 3 yrs $27M, or 4 years $40M. 3 yrs $27M is the cheapest per year ($9M/yr), so this would count as a $9M/yr deal (during the first 3 years). Then during the 4th year it would count at $13M.

80 Re: Quarterbacks and Salary

Forgive me if this has been pointed out, but near as I can tell this analysis needs to incorporate the fact that the market for QBs exists only for a limited amount of time during the season: basically between free agency and the start of camp. Furthermore, there is a limited supply of proven QBs during any given market; typically 0-3. It's a mistake to think that a QBs cap hit reflects his production; it reflects his production relative to the cost and production of available alternatives in the year he's a free agent.

81 Re: Quarterbacks and Salary

Some of this was pointed out in the article itself. In my remarks on Dalton I specifically addressed the lack of proven QBs and the risk of letting one go. It is a risk. But, as has also been pointed out (this time in comments), would Chicago have really been worse off with a no name QB and an extra 17m dollars? How about San Francisco?

82 Re: Quarterbacks and Salary

It also has to be pointed out...you face tremendous media backlash if you let a qb go and the next season the team really struggles.

I think Chip Kelly will be a great example of this. He's had two successful seasons in Philly, but if the eagles tank(which is likely imo), I guarantee he will be fired by the constant press of the media.

83 Re: Quarterbacks and Salary

No, it doesn't have to be pointed out, as I already addressed the factor of blame for a risky decision (last paragraph).

Also,you guarantee that a coach the media talks about as the next Belichick would be fired after 1 bad season? Doubtful.

84 Re: Quarterbacks and Salary

Should have said, reminded.

And yes, if the eagles finish with a 4-12 season, I could see him getting fired. Especially since his fingerprints are all over the offseason moves and it will be seen as a kind of cause and effect if the eagles tank badly next year.

And honestly - coaches have been fired even when they have been really successful. Jim Harbaugh, Marty Schottenheimer got fired after a 14-2 season. Lovie Smith got fired after a 10-6 season with the bears where his qb got hurt.

85 Re: Quarterbacks and Salary

Sorry you are right, I Should have said "should be stressed"

And yes, if the eagles finish with a 4-12 season, I could see him getting fired. Especially since his fingerprints are all over the offseason moves and it will be seen as a kind of cause and effect if the eagles tank badly next year.

And honestly - coaches have been fired even when they have been really successful. Jim Harbaugh, Marty Schottenheimer got fired after a 14-2 season. Lovie Smith got fired after a 10-6 season with the bears where his qb got hurt.

95 Re: Quarterbacks and Salary

I can't believe this didn't come up. But if Wilson gets what he wants( highest paid qb meaning cap hit of at least 22 mil), by this model - he will be significantly overpaid. Should the seahawks really let him go and try to make it work with some journeyman?

96 Re: Quarterbacks and Salary

Well, highest paid could refer to salary or guaranteed pay or even bonus. The rumor is that he's turned down 21 million, but I doubt he is going to play an extra year for 1 million extra a year. I think it is more likely this is being played as a dispute, when they are really holding out for other reasons. Let's say he plays out his contract and wins another Super Bowl. He gets the contract he wants, which we'll be generous and say is 24 a year for 5 years.

2015: 1.7
2016-2020: 120

Total: 121.7

That's not bad. Now, let's say he caves and accepts a 20.5m a year contract for 6 years, less than what he's supposedly been offered. That comes out to 123 million total, MORE than he would make by holding out.

Now, to your actual question. Short answer: yes. If his current usage is any indication of his role in the offense, then they are better off playing someone else. What sucks (for them) is that no matter what they decide, the team will get worse.

97 Re: Quarterbacks and Salary

I am curious how most Seahawk fans feel about this. Wilson is beloved, but I'm curious to see what they will say once this deal effectively forces out a big chunk of their defensive talent.

154 Re: Quarterbacks and Salary

As a Seahawks fan I basically cannot believe any other QB could be effective behind their offensive and OFFENSIVE line. I do not even believe Lynch would be that effective without the threat of Wilson holding back side pursuit. Zone running requires a lot of patience from the RB for the hole, and Wilson creates time for that patience.

I've seen good teams with good Ds stack the box to try and make Wilson one dimensional. I've also seen him light those teams up for 300 yard passing days with 3-4 TDs like he did to the NO Saints in prime time during the 2013 season. I've seen teams decide that it was better to defend against the pass then the best running game in the NFL. I've never seen him appear overwhelmed by a defense, even in the NFC Championship the INTs were as often Kerse volleying the ball to the DB as bad play by Wilson.

I can remember off the top of my head maybe 2-4 multi INT days and many/many multi touchdown days.

I have also seen the Seahawks become effectively unstoppable on offense when they take the leash off of Wilson's running game and go from being an I form zone/boot team when winning to being a shotgun zone/read option team when trailing or tied in the fourth quarter or for the final drive before the half.

Wilson gives me pause because he can't hit the short crossing pass plays that are the staple of the modern NFL offense, I believe primarily because he is physically unable to see them and deliver the ball to those areas of the field. He gives me pause because towards the end of 2013 and through effectively all of 2014 it looked like his pass protection was beginning to effect first his footwork and mechanics and then his decision making. Finally, he keeps plays going in the pocket, where he takes his truly big hits, longer than other QBs and it makes me concerned for his durability.

I remember when they were the Sea-chickens.

157 Re: Quarterbacks and Salary

I have seen him overwhelmed by a defense. The 49ers have done it to him several times. And no, I thought his performance for 99 percent of the gb game was abysmal. So I consider that an overwhelmed performance. He was terrible and had a collection of crazy plays not gone the way they had, we'd have a very different impression of Wilson's performance.

I don't want to come across as a Wilson hater. I said if someone wanted to argue he was the 4th to 5th best qb in the league, it wasn't unreasonable. That said, I've noticed a number of people have brought up his o line as THE reason for him not otherwise putting up otherworldly numbers.

The reason I find this problematic is - you have to be watching specifically the o line and then being able to compare it to the league average. No one, not even ben Muth does that.

Furthermore - having done some analysis on pressure, its remarkable how much a qb can make your o line look good by virtue of having good pocket awareness, but especially, getting rid of the ball very quickly. In other words, maybe wilson's o line is hindering him, but I don't think hes doing anything to make it look better.

158 Re: Quarterbacks and Salary

When I site his o-line I'm basically siting a specific hitch in his mechanics that affects his accuracy and appears to have effected his decision making. During the 3rd qtr of the 2013 season, games 9-12 the Seattle o-line became a real dumpster fire due to injuries to both offensive tackles, center, and a guard and it became quite common on PAs and standard drops from center for the primary defensive pass rusher to beat Wilson to the top of his drop. Ever since that period on a large number of his drops he has developed a hitch where he immediately jumps forward off the spot where he is supposed to drop to and it throws off his accuracy within the design of the play. Once he began to develop his mechanical hitch he became hesitant in his decision making process, refusing to throw balls that he would throw before.

I would like to think that he started getting better in the tail end of '14 but that may have been the use of more shotgun by the hawks which has always helped him play better.

Don't get me wrong, the Green Bay game is correctly considered the worst game of his career, but two of his INTs in that game hit Kearse in the hands and on the deep INT that Ha-Ha caught if he had either thrown the ball earlier or waited for HaHa to clear his coverage (he was in man trail on a different player and the two WRs crossed briefly taking HaHa into position to make the play) the WR would have been open for a huge play.

Asside, PFF, rated Seattle's line in 2013 as the 30th best unit in the league, and I believe they were rated as the worst pass blocking unit that year as well by PFF. This year they had the worst adjusted line yards in rushing of a team with a top 10 rusher.


I remember when they were the Sea-chickens.

159 Re: Quarterbacks and Salary

Well, I just rechecked them. 2013 - Sea was 25th in pass blocking, 23rd in Rn blocking. In 2014 they were 18th in pass blocking, 17th in run blocking.

I'm not going to make them out to be the 2000's Chiefs, but they are clearly a below average group, not a tire fire in the way that say the bears of the late 2000s were. That being said, I do recognize Wilson has to compensate for this sub par bunch. He does it well, but he also exacerbates the problem. And for the gb game, I wasn't speaking strictly to his ints, it was his general ineffectiveness(just in case it sounds like I'm picking on wilson, I thought rodgers was terrible in that game also).

I will say, Wilson to me has some very obvious strengths and some pretty glaring weaknesses. I consider his broader accuracy, presnap ability to process reads and release the ball quickly, or his ability to compensate for his o line as all clear weaknesses.

On the other hand, he's an incredible scrambler, decision maker, and great touch on deep balls.

Just off my opinion, he's in the romo tier.

I guess much like Brady Manning, its become a Wilson Luck comparison.

I am probably harsher on Luck that most colts fans. To me its a conundrum. Luck is the more polished player, but Wilson is the more consistent player. Luck's accuracy can go from great to scattershot in the same game. He's a better pocket mover than Wilson in terms of processing reads and making decisions, but hes a worse decision maker when it comes to throwing the ball.

To me - at this moment - its a matter of taste. I'd rather have Wilson with the seahawks D and Luck with the colts D.

160 Re: Quarterbacks and Salary

I don't think you can fairly claim that Luck is a worse decision maker. His defense is bad and the Colts are asking him to win games with his arm. That puts a lot of pressure on a QB.

161 Re: Quarterbacks and Salary

Well I'm definitely not trying to take anything away from Luck, who's hampered by the talent around him and the coaching, though being a BIG fan of PAC-10 football I was surprised that Luck went gunslinger in the pros. I was expecting more of Aaron Rodgers/Peyton Manning mental control not Brett Farve grip it and rip it, which may be a reaction to what he's had to deal with.

As for Wilson. Yes he's limited, but unless you're talking about Aaron Rodgers, every QB in the NFL is limited in someways. In his short over the middle passing he's not accurate most times he doesn't even try or cannot see. To all other areas of the field, he was accurate until a period starting at the end of 2013 through much of 2014 when he developed some accuracy issues. I believe that many of his accuracy problems are helped when he's in the gun and can see the pass rush rather than the predominantly PA game that they often have him playing under center that requires that he turn his back to the field during his drop.

Wilson in my opinion is clearly among the best of the QBs in the NFL, I would also include Romo in that group as for most of his career good numbers with terribly run team around him. Wilson's game is more similar to Romo & Roethlisberger, fellow pass rush escape artists, than it is Luck/Manning/Brees/Brady who are all some version of traditional drop back passer. Though, Luck with his 4.7-4.6 speed, in Kaepernick range, has the tools to go full Chip Kelly read/option if a coach ever asked him to.


I remember when they were the Sea-chickens.

162 Re: Quarterbacks and Salary

I think you are having some confirmation bias.

Andrew Luck's career INT rate is 2.4%, that's lower than Peyton's 2.6%. It's not Rodgers esque 1.6%, but then who is.

However, Rodgers didn't play until he was 25. Luck was 25 this past year. Rodger's INT% when he was 25? 2.4%.

Peyton's INT% through age 25 was 3.6%, but it was a different era.

Luck has never had a year with a significantly worse than league average INT rate.

Wilson's INT% is 2.1% for comparison.

165 Re: Quarterbacks and Salary

Aaron Rodgers' int rate is simply ridiculous. I do think part of it is a well constructed offense by Mccarthy, but Rodgers deserves a boatload of credit.

I am really surprised Luck's int percentage is so low. Having watched every single game of Luck's career - his accuracy is my biggest problem. It's just not consistent enough.

From the start of his career, I've basically sat and compared PM to Luck. They aren't similar players, not really at all.

166 Re: Quarterbacks and Salary

Yeah, Luck isn't all that accurate in the stratosphere of good NFL QBs.

I'd say if you're looking for a historical comparison, go with John Elway. He was another big athletic player who wasn't completing passes as frequently as his peers, but was still damn good.

167 Re: Quarterbacks and Salary

Stylistically, I prefer Peyton. I like everything Peyton does except his inflexibility when it comes to offensive style. In that regard, I prefer Brady.

I admit they are still awesome qbs, but I was never a fan of the roethlisburger, Favre, Andrew Luck style of qb. Of course I root hard for Luck, but Im not as partial to that style. Yes I realize I sound like a guy who's sending back filet mignon because its not been dry aged.

168 Re: Quarterbacks and Salary

As someone with living memory of the Seahawks missing out on Drew Bledsoe and getting Rick Mier instead... yes, you are totally sending back filet mignon because its not been dry aged. Not quite as bad as the SF fans disappointed in Jeff Garcia because he didn't measure up to Montana and Young, but come one.

Luck is a guy that you can say with comfort, baring injury I expect him to have a hall of fame career.

I remember when they were the Sea-chickens.

169 Re: Quarterbacks and Salary

I think a large part of it is that many Luck interceptions are not so much "The QB was off on his pass," or "The play was a bit obvious and the QB didn't execute it fast enough, so the defender jumped the ball," but "Please check for a concussion, because no professional QB should have decided to attempt THAT."

Just unfathomably bad and ugly interceptions, the kind you really remember, and they establish the narrative.

170 Re: Quarterbacks and Salary

Unfortunately - that became part of the favre narrative. I usually don't play qb psychologist, but I really believed many of favre's bonehead ints were the result of unbelievable hubris. That no risk was too great for him.

Luck on the other hand...who knows.

114 Re: Quarterbacks and Salary

it has to be possible to massively front load Wilsons contract- like 50% of the cap , and extend t the others players contracts with more guaranteed money later while taking a massive hit this year that way wilson gets his cash and the Hawks keep their core, when you have a chance at 5-7 years of dominance on the line you gotta go for it

107 Re: Quarterbacks and Salary

I'm not sure how I feel about this. If I am reading this correctly, Cam Newton just got the more-or-less going rate for a decent-to-good QB but will need to triple his output to make the deal worth it? ... and DangeRuss will need to be half again as good after two Super Bowl appearances.

In a lot of ways this reminds me of what we termed "Salary Compression" during the Engineering Hey-days of the 1980's. An Engineer with a degree and 5 years of experience was getting the same or even less as a freshly-minted college graduate.

I have the feeling that linear will not be your best fit, and something like a yield curve will be the best fit (which may make sense, in that you are looking at cost/performance.)

108 Re: Quarterbacks and Salary

The problem is...there's no theory. We are just guessing at what the relationship is. In principle, it should be linear as you are payed for performance. In actuality, its a nasty mix of risk, public pressure, strange contracts, and salary caps. A non linear relationship may "fit" better, but the inference will still be opaque imo.

109 Re: Quarterbacks and Salary

Great questions, let me explain.

Cam not only had nagging injuries all throughout the season (ankle), he also only missed one week for a brutal car wreck. His leading receiver was a rookie...a raw one, and his O-line took a step backward. He's also young and likely to show improvement. None of these things were factored into the metric, but they were factored into his real contract.

Russ really has had a limited role in the Seattle offense, which wouldn't be described as explosive. In addition, we are talking about another QB with a sub-par line and receivers. The metric isn't measuring talent...just production.

You aren't the first to suggest a non-linear formula, and I only avoided one for the sake of simplicity. Later editions will address some of the issues mentioned above. This is really just the starting point for more advanced formulas, and to get people discussing value.

Thank you for putting some thought into this.

110 Re: Quarterbacks and Salary

This is an interesting analysis, but any analysis that thinks that both Mannings, Philip Rivers, Matt Ryan, and Drew Brees were overpaid is pretty obviously seriously wrong in some respect.

I would add Romo, but the formula more obviously is a problem there for his 2014.

Any NFL team would gladly shell out $20M a year for one of the above. They would not think twice. I think there must be something wrong with the best-fit line, or else the use of ANY/A or PYAD as a proxy for value to the team. I suspect the former; somewhere in the middle of the rankings there is an inflection point or something.

116 Re: Quarterbacks and Salary

Although, Flacco was pretty much as good as all of the above cited examples (except P. Manning and Romo had higher DVOA) last season, and also is younger (except same age as Ryan), durable, and going into his prime.

127 Re: Quarterbacks and Salary

Again, it's less about which QBs are "good" or "bad" or what their absolute value is, only relative to other top QBs.

I chose Flacco mostly randomly; he was the first QB I saw that was worth between -5M and -10M, relative to Rodgers.

112 Re: Quarterbacks and Salary

I have quite a few issues with this analysis.

1. You should have used a multi-year average for TANY/A or ANY/A. since players are based on their performance from over more than one year. If you do that, the linear regression for $CAPHIT:ANYA goes from .22 to .33. That's a lot less noise.

2. plays per game appears to have no real impact on salaries...why did you use it? The p-value is abysmal

3. consider a non-linear approach (others have pointed this out in the comments).
3a. With a 3-year ANY/A vs Cap hit, All QBs making under $6m per year (except Bortles) are UNDERpaid, while all QBs making more than $14m are overpaid. With a power approach, a QB with a three-year ANY/A of 10 would make about $35m a year

4. If you look at advanced metric -- Expected Points Added is what I used (but you can use DYAR) -- the regression gives r^2s of over 0.5. QBR is above 0.5 too.

119 Re: Quarterbacks and Salary

"2. plays per game appears to have no real impact on salaries...why did you use it? The p-value is abysmal"

Because he's not aiming to explain why players make as much as they do so much as identify if they should be or not. It looks like NFL executives might be underpaying usage.

Flacco and Ryan had approximately the same DVOA, but Ryan was throwing 5 more passes per game. That's significant, like an entire extra drive. You have them trade defenses (and head coaches) and the Ravens probably get better while the Falcon's get worse.

122 Re: Quarterbacks and Salary

I'm not a fan of DVOA as a measure of qb play. Also in Ryan's defense, Flacco's line was much better than Atlantas. That said, by qbr(a statistic I like better), flacco did indeed have a pretty good season last year. I guess that kind of surprises me since I've had a pretty static view of Flacco for a while now.

Could this change my view of Flacco or was this an outlier/kubiak thing? Maybe. He did have a similar effect on Schaub(for a while anyways).

I guess now have to reassess where I put flacco in terms of tier rankings. He's not in the upper echelon, but last season was very much in keeping with Big Ben's avg qbr over his career.

123 Re: Quarterbacks and Salary

Matt Ryan is a lot better than Flacco when it comes to expected points created, which is the ideal measure, really. They being said, Flacco is in that 10-15 range and is paid accordingly.

However, Dalton is an average to below average QB based on expected points - and that's with one of the best three WRs in the NFL, an expensive O line. I can't believe he's paid like that.


148 Re: Quarterbacks and Salary

Well, I have quite a few issues with your feedback. ~grins~

Actually, I don't. I want you to know that I saved this entire comment for things to keep in mind for the more advanced formula later on. It's good feedback, even though a couple of your points were things I decided against for what I felt were valid reasons.

1) I wanted to do 1 year at a time because way too much changes on a year to year basis. I plan on doing this every year, so eventually there will be multi-year comparisons available.

2) Tuluse actually nailed this one. PAYD is meant to determine a worth independent of current trends.

3) Yup.
3a) Again, too much change year to year.

4) The next formula, in which I hope to actually be able to determine which teams would improve from taking a dive on QB quality, will use DYAR.

Not sure how to coordinate this, but I wouldn't at all mind working with you on the next formula. I'll get with someone from staff and see if I can somewhat safely pass some contact information to you.

115 Re: Quarterbacks and Salary

although this list does a good job of ranking the qbs on a money efficiency graph its hard to say where else you can get the points from, meaning lets say you replaced some of an inefficient qbs salary number and slid that money over to a defensive linemen or receiver at the end of the day a receiver is not gonna be worth more than 2 pts a game that's where they max out but a qb has a higher max number wich is why its more efficient to pay for an above average qb as opposed to a receiver also a good qb can make a wr look good attempts at the opposite have been frustrating and fruitless. thats why you don't see many teams that build around great position players and live with a crappy qb because its the most inefficient way to build a roster. Enter the present day Seahawks who are gonna be forced to pay Wilson more than he's worth in the salary cap era or they become the first great team fungible qb lab experiment

120 Re: Quarterbacks and Salary

It's better to pay an actual above average QB. The problem is that many teams are paying that above average money to average QBs (Lions, Bears, Chiefs, 49ers, Bengals, and arguably the Ravens).

124 Re: Quarterbacks and Salary

The problem is that many teams are paying that above average money to average QBs

Yes, average QBs are getting paid above average money, but there’s a good reason for that. You can build a championship team around an average QB. Building a championship team around a bad QB is virtually impossible. And building a championship team is the ultimate goal, after all.

And to be more precise, I don’t mean that there’s some magical point that a QB’s DVOA dips below where they suddenly go from a valuable commodity to a replaceable part. A QB with a -1.7% DVOA could still count as “average” by my criteria. In fact, instead of “average”, let’s call it “solid”. The main requirements for being a solid QB are that they perform somewhat consistently, and that they have no devastating weaknesses.

In other words, no matter what the defense does, they have both the mental capacity to formulate a competent response and the physical ability to act on it. They don’t necessarily have to formulate the ideal response, and they don’t have to execute it perfectly, but they have to have an adequate answer to whatever the defense does.

So, if the defense leaves a CB (not named Revis) on an island with your best WR, they have to be able to recognize the opportunity and deliver a pass downfield. They don’t need to throw a laser 60 yards downfield that hits the receiver in stride, but they need to throw deep passes well enough to make the defense pay when it’s left there for the taking. Or, if the defense drops a bunch of people deep, they have to be able to go through a progression and get a short pass to the open man quickly and consistently enough that the defense can’t just shut your offense down by taking away the deep ball (I’m looking at you, Rex Grossman). Same goes for recognizing blitzes, knowing when to audible to a running play – all the basic stuff.

They don’t have to be good at all of those things, but they can’t be really bad at any one of them to the point that a good defense could wreck your offense just by taking advantage of that one vulnerability. Because at some point in the playoffs, you’re going to face a good defense.

Beyond just being competent in all the basic skills, they have to be consistent enough to be able to string together a handful of decent games in a row. If they can put together a couple 4-5 game stretches of decent performance each season, and they have a good team around them, you’ll average around 9 or 10 wins a year and make the playoffs fairly often. And when you do make the playoffs, you’ll have a decent chance to win a Super Bowl, as long as the rest of the team does well and your QB puts together a good stretch of games in the playoffs.

Of course, it’d be easier to win a Super Bowl with an elite QB, but those guys are usually already taken, and a solid QB is enough to give you a decent shot.

Now, compare a solid QB to a QB who is good at most things but has one devastating weakness. Something that only a good defense (or at least a defense well suited to take advantage of that weakness) could exploit. This is the kind of QB who is fine as a backup, or a stopgap to start for a year or two while you get a new QB up to speed. They’ll play well enough in most games, but suffer against quality opponents. That’s about all you can expect from a backup, and it’s all you really need from a stopgap. They can get you 6-8 wins and run your offense well enough to let you evaluate your players while you rebuild, but you’re not winning a Super Bowl with them. That’s why there’s the huge disparity in pay. The difference between a QB with one major vulnerability and a QB with none is all the difference in the world.

In any case, if you already have a solid QB, and he wants $10 million/year more than you think he’s worth, you’ve got 3 options:

1) Get rid of him and acquire another solid QB later for less money.

2) Get rid of him and try to win a Super Bowl without a solid QB.

3) Pay him.

(1) would typically mean using a 1st round pick, and giving up the access to talent and cheap rookie contract that comes with it. That would defeat the purpose of saving all that cap space, because you wouldn’t be able to acquire players worth spending it on (and you’d lose the cap space that you would’ve saved on another player’s rookie contract).

(2) requires a defense on the level of the 2000 Ravens or 85 Bears. Historically great. You’re not getting there from here with $10 million or so in free agent defenders. That’s enough cap room for DeMeco Ryans and about half of JJ Watt. Yeah, good luck with that. Have fun losing games 17-10 on your way to a 7-9 season.

(3) is the only viable option, which is why it’s the option taken by just about every team facing this choice.

126 Re: Quarterbacks and Salary

"Yes, average QBs are getting paid above average money, but there’s a good reason for that. You can build a championship team around an average QB. Building a championship team around a bad QB is virtually impossible. And building a championship team is the ultimate goal, after all."

Other than the tautology that it's easier to win with better players, I think there have been more teams in Superbowls with bad, but very cheap QBs than average, but well paid QBs. Not that there have been many of either.

"A QB with a -1.7% DVOA could still count as “average” by my criteria. In fact, instead of “average”, let’s call it “solid”"

-1.7% DVOA would have been the 24th ranked QB last year. It's basically Mark Sanchez. Which is the prime example above of a QB that's cheap. If Mark Sanchez wants 15 million a year (10 million more than he's making). I'm pretty sure you take option 4) laugh in his face.

128 Re: Quarterbacks and Salary

"I think there have been more teams in Superbowls with bad, but very cheap QBs than average, but well paid QBs. Not that there have been many of either."

There haven't really been expensive-but-average QBs for very long, though, so you can't really compare to forty-nine years' worth of Super Bowls. And the game in the 1970s to the early 1980s was much more run-and-defense-focused, so the QB position wasn't quite as important, anyway.

136 Re: Quarterbacks and Salary

I think there have been more teams in Superbowls with bad, but very cheap QBs than average, but well paid QBs. Not that there have been many of either.

I count one team that actually won a Super Bowl with a cheap journeyman, and their defense gave up 0 points in the Super Bowl. So, yeah, if ditching your starting QB for a cheap journeyman frees up enough cap space for you to make your defense rival the 2000 Ravens, then by all means, ignore my advice. But what team in the NFL has a defense that good right now?

As for average QBs, there’s Eli in 2007. He was actually well below average that year in DVOA (at least in the regular season), but his previous seasons had been just about exactly average, so I think it’s fair to say he was a solid QB at that point in his career. And his cap hit was among the highest of any QB in the NFL that year, so he wasn’t cheap. Flacco in 2012 was slightly below average in DVOA, though his cap hit was only average. As Eddo pointed out, expensive-but-average QBs are a relatively recent phenomenon.

As you mention, there aren’t a lot of examples of either situation, because most Super Bowls are won by teams with elite QBs. Really, the best solution is to have an elite QB. But elite QBs don’t just fall out of the sky, so the teams that don’t have one have to make do with what they can get. If you’ve got an average QB, and a good team around them, you’ll make the playoffs reasonably often, and you can always spend late round picks on project QBs in the hopes of finding the next Tom Brady.

-1.7% DVOA would have been the 24th ranked QB last year. It's basically Mark Sanchez. Which is the prime example above of a QB that's cheap. If Mark Sanchez wants 15 million a year (10 million more than he's making). I'm pretty sure you take option 4) laugh in his face.

I was thinking of 2012, where it would’ve ranked 18th, right between Joe Flacco and Carson Palmer. But fine, maybe -1.7% is too low. I didn’t give the number much thought because the exact number is beside the point. The point is that a QB could be, say, 18th in the league in DVOA (and therefore technically a below average QB) while still delivering decent production and having no major weaknesses in their game – a solid QB who is still technically a bit below average.

I did not mean to suggest that anyone should be paying Mark Sanchez more than $5 million a year. You are correct that that should only inspire laughter. And frankly, even $5 million seems a little high for a QB whose most notable play was a buttfumble, but I’ve given up trying to rationalize the personnel moves the Eagles have been making of late.

And speaking of Eagles QBs, I'd like to point out that as valuable as an average QB might be, you should not trade away a Pro Bowl QB and a 2nd round pick just to get one. That will only end in tears.

So, in conclusion, fire Chip Kelly.

137 Re: Quarterbacks and Salary

I was looking at all teams in the Superbowl, not just winners, since making means you were damn close. I was also counting all bad-to-clearly below average QBs, not just journey men.

Rex Grossman was a late first round pick who didn't cost much. I can't imagine Kerry Collins or Brad Johnson cost the Giants or Bucs very much, respectively. Though, Johnson would fit into the above average category.

Eli is a strange case in that he flipped a switch or something and went from below average for the regular season in 2007 to well above average as soon as game 16 of the season started. Seriously he went from 4.8 ANY/A to 6 the next year to 7 the year after that. So this is a case of separating "quality of qb" from "production of qb", which admittedly is really hard to do. Flacco could very well be another player who is actually just better than his stats are. In 2012 he went from 7.17 AY/A to 10.79 in the playoffs (pfr doesn't list ANY/A on the gamelog page unfortunately).

So the Giants and Ravens got the production they were paying for once the playoffs started. If those QBs had played like they did in the regular season, no way they win superbowls.

I will concede that "qb who is generally average, but has the potential to go on a hot streak" is worth more than a QB who is only average, or possibly one who is better on a regular basis but doesn't have that streak potential.