Brad Blank apparently thinks Brady left too much money on the table in his recent extension:

http://www.bloomberg.com/news/2013-02-26/tom-brady-s-redone-patriots-contract-might-upset-some-nfl-agents.html

Most pertinent quote:

“He gave up a lot,” Blank said in a telephone interview. “The top quarterbacks are getting about $20 million a year now, and they’re getting as much guaranteed as Tom got.”

But did he give anything up? He essentially signed a 5 year contract worth about $60 million, guaranteed, for his age 36-40 seasons. Drew Brees signed last year for a comparable guarantee over 5 years, and he is significantly younger.

Blank’s objection seems to be that the contract doesn’t have a nominal value well in excess of the guarantee. But do long-term, high-value contracts in the NFL ever approach the max value in practice? My impression is that they are almost always either renegotiated or terminated before the later high-value years kick in. I have a hard time seeing how guaranteeing Brady $60 million at this point in his career can be anything but good for elite QB’s as they approach their late 30’s.

My main question: Are agents paid up-front on the value of NFL contracts, or are they paid over the life of the contract as the player is paid? I.e., does Blank have a significant financial incentive to push his clients into contracts with high nominal values over contracts with high guaranteed values? Is he just talking his own book here, or does he have a legitimate point that I am missing?

Didn’t he ACTUALLY sign a 3/27 extension?

You can’t just bundle that with his existing 2/33 (or whatever he has) and call it a 5/60 deal that he signed.

This is the same problem people made with Felix.  Felix signed a 5/135 extension.  That they “tear up” the previous contract, or they give him a signing bonus payable in 2013 and 2014 is irrelevant.  (Other than the time value of money.)

Anyway, signing an extension with still two years to go for a guy in his mid-30s… well, you have to tell me what that should be worth.  It’s obviously not going to be the same 20MM$ standard rate that a 30yr old QB is going to get as a free agent.

We need less blustering from journalists and more analysis.

Fair enough (though the NFL is slightly different from baseball in that the tear-up itself is usually a material part of the motivation for signing an extension). For the value of the added 3 years, you can start with the following data since 1970:

1. 34 players have started 8 games in their age 36 season (excluding Peyton Manning in 2012, since he could ultimately fall in or out of the groups below)
2. 24 have started 8 games in their age 37 season
3. 16 @ age 38
4. 6 @ age 39
5. 2 @ age 40 (Favre and Moon)
6. 3 @ age 41 (Testaverde joined Favre and Moon after falling off in his age 40 season)
7. 1 @ age 42 (Moon)
8. 0 @ age 43+

Obviously the game and medical treatments have evolved, but prior history gives little evidence that Brady is likely to be a meaningful contributor beyond the end of his current contract (which ends after his age 37 season).

For lack of a better handy metric for comparing across eras, of the 16 players who played at age 38, 11 posted a TD/int ratio of 1 or better. Of the 6 who played at age 39, 4 posted a TD/int ratio of 1 or better.

So to first order, assuming Brady makes it through his age 36 season, prior history would give him a sub-50% chance of starting regularly in the first season of his new contract, and call it a 2/3 chance of being “productive” if he does so. If we give some credit for improvements in medical treatment and reduced injury risk (due to both rule changes and offensive evolution) maybe he has a 3/4 chance of starting 8+ games, and a 3/4 chance of being “productive” if he does? So call it a 9/16 chance of being worth X at age 38, where X is the fair market value of a productive QB (whatever that is), and a 7/16 chance of being worth 0 (for simplicity, but I think it’s defensible that a 38-year-old below-average QB has no material value in the NFL given roster constraints). Let’s be generous and say X is 20, since Blank cited that figure for star QBs. So I’d cap Brady’s expected value in the first year of his new contract at 11.25 million, and declining steeply from there.

In other words, I’m not betting on him playing up to this contract.

But in any case, I was actually more interested in Blanks’ incentives here. Maybe he’s right that Brady signed for below market value (certainly other people perceive it that way). But my cursory read of the economics says he’s not right, and I’m curious if he has an identifiable incentive to misinterpret the contract in that way, or if he is just fooled by big non-guaranteed numbers in the same way that the general public usually is.

If you figure he’s worth 11MM$ in the first year of the extension, and then say 9MM$ in the second year and then say 7MM$ in the third year, that’s a 3/27 extension.  Which is what he signed, right?

If that’s a reasonable set of expectation, then that’s what he’s being paid, and there’s no money being left on the table.

Except that if the expectation is 11 million in the first year, it would be more like 5.6 in the second and 2.8 in the third, since the historical attrition rate of QB’s over 38 is so high. That would be 19.4 million total. And I’d call that an upper bound on expected value rather than an unbiased estimate, since I’m giving him “superstar” credit for anything above “bad”.

Tangotiger - 27 February 2013 03:15 PM

If you figure he’s worth 11MM$ in the first year of the extension, and then say 9MM$ in the second year and then say 7MM$ in the third year, that’s a 3/27 extension.  Which is what he signed, right?

If that’s a reasonable set of expectation, then that’s what he’s being paid, and there’s no money being left on the table.

Just to be clear, I think he signed for more than he is likely to be worth, not less. I think Blanks is wrong and that Brady left nothing on the table - quite the opposite, in fact.

Right, so if anything, by your view, his extension is an overpay, and there was no money left on the table.

In order to justify the deal, and given the 50% year-over-year decline at that point, he’d need to be worth say 16MM$ in the first year of his extension, then 8MM$ in the second year and 4MM$ in the third year.

Given NFL infllation, is that possible?

I don’t think so, according to the NFL’s salary cap projections:

http://www.bigcatcountry.com/2012/6/18/3094750/nfl-flat-salary-cap

(I don’t know if those figures are the most up to date.)

Taken at face value, the cap will be about 4% higher than today in the first year of Brady’s extension and 8% higher in the second year. QB salaries could grow as a percentage of the total pie, of course.

Just because the cap is fairly flat doesn’t mean the mean will be.  You could make the cap even drop, and the mean can still rise.

How does revenue sharing work in NFL?  Is it a flat rate like NHL’s 50%?

This thread started with a comparison to Drew Brees.  Certainly he is younger than Brady, but also not nearly as good.  Here’s a 3-yr average DYAR (basically YORP, Yards-Over-Replacement-Player) for some top QBs:

T.Brady   1,983
P.Manning   1,600
D.Brees   1,599
A.Rodgers   1,581
M.Ryan   1,151
T.Romo   955
P.Rivers   895
B.Roethlisberger   895
R.Wilson   867
E.Manning   809
M.Stafford   803

(zeros were removed, so Russel Wilson has a sample size of 1 year, Peyton has 2 years, etc.)

Brady may be older than Brees, but he’s also 25% more valuable at the moment.  In a world where a (young) Eli Manning or Matthew Stafford is worth $20M/year, Brady has recently provided 2.5x as much value.  So if you think Brady is worth $20M/year today, then sure, it’s unlikely he’ll be worth $16M in the first year of his extension.  But what if he’s worth $30-40M/year today?

Actually the $20 million figure was cited by the agent (Blanks) as being Brady’s benchmark, but it’s possible he could be underestimating as well. In any case, it’s a valid point.

But isn’t the more apples-to-apples comparison the amount of guaranteed comp, given the customary backloading nonsense? In which case e.g. Brees would be valued at $12 million/year rather than 20, and Brady at a 25% premium to Brees would be $15 million.

Things get really messy when you delve into the intricacies of guaranteed-vs-non-guaranteed nature of NFL contracts.  I’d hazard a guess that QBs at this upper echelon are rarely cut, especially these days with the way the QBs are protected on the field via the rulebook and with modern medicine both ensuring that the risk of performance-crushing injury is low.  And especially considering the value they bring off the field - I think it’s safe to say Brady will sell more jerseys and more commercials in his extension than any other players that $27M of salary can buy.  But I do agree with your assessment regarding the actual guaranteed portion.

My bare-bones analysis of this: I don’t think the Patriots have any reason to overpay Brady, but Brady does have a few reasons to be willing to be underpaid.  His wife is still the primary breadwinner in their household, and it wouldn’t surprise me at all if he’d be willing to forego a few million in salary in order to increase his chances of winning more super bowls.  His utility curve for money is probably showing some diminishing returns at this point; on the other hand, one or two more super bowl wins for him should end most arguments about who was the best QB ever.  That, as they say, is something that money can’t buy.

Mike L - 27 February 2013 04:11 PM

This thread started with a comparison to Drew Brees.  Certainly he is younger than Brady, but also not nearly as good.  Here’s a 3-yr average DYAR (basically YORP, Yards-Over-Replacement-Player) for some top QBs:

T.Brady   1,983
P.Manning   1,600
D.Brees   1,599
A.Rodgers   1,581
M.Ryan   1,151
T.Romo   955
P.Rivers   895
B.Roethlisberger   895
R.Wilson   867
E.Manning   809
M.Stafford   803

(zeros were removed, so Russel Wilson has a sample size of 1 year, Peyton has 2 years, etc.)

Brady may be older than Brees, but he’s also 25% more valuable at the moment.  In a world where a (young) Eli Manning or Matthew Stafford is worth $20M/year, Brady has recently provided 2.5x as much value.  So if you think Brady is worth $20M/year today, then sure, it’s unlikely he’ll be worth $16M in the first year of his extension.  But what if he’s worth $30-40M/year today?

Wow, I did not realize Brady was that far ahead of the competition. Still, the aging curve is very relevant for evaluating his deal.

Didn’t Manning miss a season or something?

Yes:

(zeros were removed, so Russel Wilson has a sample size of 1 year, Peyton has 2 years, etc.)

Peyton seems to have recovered from the neck surgery to his prior form, so my assumption is that excluding the missed season from the analysis would better reflect his “true talent” better than including it.  Also, I’d assume that this type of injury has a low chance of re-occurring, given that his neck bones are now literally fused together.  (I realize this is totally subjective).  Likewise there’s a lot more uncertainty around Wilson’s talent level due to the small sample size of only one season; I have no idea how to regress football performance within the context of this metric, so I just eyeballed it.

But in any case, it’s clear that the current performance of Brady is much higher than any other “elite” QB right now, and miles higher than the other old guys (Favre, Testaverde, Moon) who were no longer useful at age 40+.  So even if his usefulness follows a predictable aging pattern, he’s starting from a much higher baseline and thus I think he could still remain useful further into his 40’s than any of the comparables presented.

The fact that Brady is starting from a higher level of performance than some other “old” QB’s is perfectly valid. But the (very limited) evidence does not point to a smooth and predictable decline, but rather a sharp and unpredictable decline. It’s tough to draw any conclusions from a sample of 3 players, but that’s sort of the point - we can’t look at Favre, Moon and Testaverde and say “Brady should be even better at that age”, because that sample excludes all of the great QB’s that didn’t make it that far, like Marino. Really, would anyone have predicted ex ante that Vinny Testaverde would be more effective at an advanced age than Dan Marino? That’s why I think you need to look at both performance of the QB’s who survived as well as the number of QB’s that survived. The latter data doesn’t look great for projecting any QB past 36, albeit with the usual SSS caveats.

Also, one caution on using DYAR to compare players: like WAR, it is a counting stat, and usage/opportunity in football is even more variable from team to team than in baseball because of the influence of defensive performance and offensive strategy. If you are going to use Football Outsiders stats for prospective analysis, DVOA (comparable to wOBA, i.e. a rate state) is arguably more relevant, or at least equally relevant. And while Brady still looks superior on DVOA, he’s not as superior. For example, in 2012 Brady was 13% more productive than Manning by DYAR, but only 7% more productive by DVOA.