An Economist's View of NFL draft "value"

[MoLewisrocks]

Some excellent points and likely an insight into what our own economics major sees when he contemplates trading up in the first round of a draft. He points out that not only is player value impacted by the current financial imbalance at the top of the first, it is also impacted by the compensatory imbalance in the draft charts and this is due in part to the random liklihood that a player selected 1 pick earlier will outperform one selected one pick later (52% by their analysis or roughly a coin toss...) or the liklihood that any one player selected in the top ten will outperform a combination of 2 players selected later in the first and in the second round (the projected cost to trade up just 6 spots basically from outside the top 15 into the top 10...)

This Times piece, which proposes corporations could also learn a lesson from the analysis, wraps up with a good business sense nod in our direction...

So maybe companies shouldn’t pay big bucks in the desperate hope of getting the equivalent of a Peyton Manning, who was the first overall pick in 1998 and, of course, has proved his superstar value. Instead, maybe they should dig around for a replica of Tom Brady, who was the 199th pick in the 2000 draft and has gone on to play in four Super Bowls, winning three.

Economic View - When a First Choice May Not Be the Best Choice - NYTimes.com

 

[chevss454]

The article is good but the actual paper is great! It is way too statistically complicated for me but I see his argument and accept his conclusions as truth.

One paragraph stands out and says it all:

"although the value of players declines throughout the draft, quality declines more slowly than compensation — players picked early are very highly paid. As a result, the first pick in the draft has often provided less value to his team, in performance per dollar, than the last pick in the first round (the one awarded to the Super Bowl winner). In other words, in the world of the N.F.L. draft, the rich get richer."

 

[the wrothbroughterer]

I thought it was really interesting. I majored in Economics in college so it was a fun read for me. Basing all this on dollar value brings up some good points, but I wonder if there is any statistical measure to apply to what a player actually brings on the field in terms of dollars.

For example, if anyone reads Baseball Prospectus and some of the other advanced statistical baseball web sites, there has been some research on how many wins an individual player actually provides his team with. Furthermore they've figured out in given years based on aggregate salary and some other factors, how much a team pays for a win and use that to figure out the value that players adds to the franchise. I think Joe Mauer was worth 30+M last season. Anyway, if you use that theory and apply it to football players and then compare it to where they were taken in the draft, I bet you'd find the ultimate sweet spot for value-production in the NFL draft.

End rant.

 

[patchick]

But...but...isn't this what we've all said for years? Isn't this why a rookie salary scale has been a hot topic for so long, and why so many teams at the top of the draft are eager to trade down, and why plenty of posters here say they wouldn't want the #1 overall pick if they could get it, and why a bunch of us have to revise the "draft value chart" to reflect the imbalance of salaries at the top, etc., etc.?

Economists have a funny way of acting like a field doesn't exist until the economists look at it. Just ask anybody who had years of expertise in any of the subjects Steven Leavitt "discovered" in Freakonomics.

 

[jeffbiologist]

Yes, but then it would make more sense for the top teams to forgo their rights and pass on their pick until the player value meets the pick value...and would make a mockery of the whole draft system. The team picking first for example should let the next 10-12 teams pick before them if they feel the money paid at that spot is even relatively close to #1 correct?? How many teams have you seen do this? And this is why the higher players DO get paid, because teams WANT to draft them that high.....

 

[IcyPatriot]

Yes, but then it would make more sense for the top teams to forgo their rights and pass on their pick until the player value meets the pick value...and would make a mockery of the whole draft system. The team picking first for example should let the next 10-12 teams pick before them if they feel the money paid at that spot is even relatively close to #1 correct?? How many teams have you seen do this? And this is why the higher players DO get paid, because teams WANT to draft them that high.....

Are you serious?

What team have you been following the last 10 years or so?

 

[jeffbiologist]

Are you serious?

What team have you been following the last 10 years or so?

Ah-whats YOUR point? This thread isnt about following a team it was about the "value" of the top picks. Many of the same teams draft high year after year, throw out those and blame that on the organization. But now and then a historically "good" or "smart" team will have a bad season and draft high. Do they forgo their pick to a more cost-effective spot in the draft?? Never. This is a poker game where the teams are playing each other not the deck. Ultimately its not the pick or the money that makes the "value" but the player! Generally speaking the more lottery tickets you have the better chance you have at winning.....

 

[IcyPatriot]

Ah-whats YOUR point? This thread isnt about following a team it was about the "value" of the top picks. Many of the same teams draft high year after year, throw out those and blame that on the organization. But now and then a historically "good" or "smart" team will have a bad season and draft high. Do they forgo their pick to a more cost-effective spot in the draft?? Never. This is a poker game where the teams are playing each other not the deck. Ultimately its not the pick or the money that makes the "value" but the player! Generally speaking the more lottery tickets you have the better chance you have at winning.....

That'[s not at all what you said above. Above you implied drafting the highly skilled player high and paying him. your post was not at all about drafting for value.

 

[lovsn]

I should probably read the whole paper first, but, as someone doing a PhD in behavioural economics (essentially the same field as Thaler), I can see some fairly immediate causes/effects which don't seem to have been covered

Firstly, there is a difference between what we call the "partial" and "general" equilibrium effects. The partial effect says "ok, are the high picks overvalued in isolation from everything else" whereas the general effect says "ok, what are the knock on effects of a high pick beyond the pick itself"

In this case, the obvious "general equilibrium" effect is that by paying a very high salary to a top pick that is preventing you from paying a high salary to someone else on your team (or through free agency).

Say you could draft the top QB at number 1 and pair him with a middling receiver, or you could draft the 3rd best QB at 32 and pair him with a top receiver (for the same total money). In the second case, the top receiver will make the number 32 pick look better than he would be, if considered in total isolation against the number 1 pick. Similarly, the positive effect of going to a winning team (i.e. the number 32 pick) could make that player look better than they maybe are. An interesting comparison would be the number 32 pick versus the number 33 pick. Essentially the same player talent, for the same player money, but going to a winning as opposed to a losing team.

The issue here is that the player is typically tied in to the team that drafts him for a minimum of 4 years, so you can never really isolate the "general" effect from the "partial" effect. The team and the player are intrinsically linked, so maybe what appears to be a problem with draft pick prices etc. is really an "externality" problem which is basically unavoidable. In which case, the draft pick prices could actually be efficient, but with respect to the limitations imposed by the contract/team situation.

 

[Urgent]

I had the opportunity to work with Mr. Massey many years ago - he was on my staff at one point. After he left, I corresponded with him on his analysis on this paper.

The central point, as I recall, is not so much the performance of players in the draft, but the comparison of relative performance and relative pay. The analysis led to a value comparison, demonstrating that picks 20 or so through 60 or so are the 'sweet spot' of the draft. Before 20, the salaries are very high relative to expected performance (expected value). Below 60, the expected performance (value) drops more rapidly. The relative value of players in the second round, in comparison to replacement players available at market rates in free agency, is at its peak.

So, to maximize value, you would want to maximize the number of picks in the 20-60 range. Clearly something the Patriots have done recently.

However, this is maximizing value, not performance. If you want to maximize performance, you want a bunch of players from the top 10. The value component is a constraint if you have cap concerns. Often (recently) teams at the top of the draft don't have a roster full of highly-paid All Pro's, and face less constraints related to value.

Good question on deferring the first pick in the draft, which has been debated before. But if you are a Detroit or St Louis with lots of cap space, and some hopes of attracting fans to the game, it's not an effective marketing strategy to communicate that you are 'cheap' and want to pass up the Sam Bradford or Matt Stafford for a less expensive Colt McCoy. Other than incompetence (that's you, Minnesota) no one has actively pursued a strategy of deferring in the top 10.

However, Thaler's contention that corporations should look for the next Tom Brady is poorly supported. If you compare the real risk and expected value framework to hiring CEO's, and suggest it is better to risk the billions of dollars of impact on taking a Tom Brady - the equivalent of a sixth-round pick with a 10% chance of making it as a starter - to save several hundred thousand dollars in base compensation is very short sighted.

 

[patchick]

In this case, the obvious "general equilibrium" effect is that by paying a very high salary to a top pick that is preventing you from paying a high salary to someone else on your team (or through free agency).

Say you could draft the top QB at number 1 and pair him with a middling receiver, or you could draft the 3rd best QB at 32 and pair him with a top receiver (for the same total money). In the second case, the top receiver will make the number 32 pick look better than he would be, if considered in total isolation against the number 1 pick. Similarly, the positive effect of going to a winning team (i.e. the number 32 pick) could make that player look better than they maybe are. An interesting comparison would be the number 32 pick versus the number 33 pick. Essentially the same player talent, for the same player money, but going to a winning as opposed to a losing team.

Beautiful stuff. I love the 32 vs. 33 comparison.

There are also positional confounds to consider. For instance, in the past 5 years, 7 quarterbacks have been selected in the top 10 picks, 3 of them at #1 overall. In contrast, only 2 QBs have been selected in the bottom 10 picks of the 1st round -- and one of those was a top-10 selecting team trading up from round 2.

And, of course, the kind of management that lands you at the top of the draft is the kind least like to select wisely. Of those 7 high-drafted QBs, there have been only 2 apparently successful picks (Ryan and Stafford). Both of them were selected by new management teams after a total housecleaning.

 

[cornhuskers_2002]

BB majored in economics..

 

[jeffbiologist]

That'[s not at all what you said above. Above you implied drafting the highly skilled player high and paying him. your post was not at all about drafting for value.

I commented about the process of finding the spot where talent equalled value and passing on picks til it was cost effective.....what about that is not drafting for value?? Geesh, maybe you meant to respond to someone else's thread and clicked mine by accident....:bricks:

 

[Don Kipines]

I read the paper. I think some people are missing what the author was trying to say.

Yes, it's true that you do have a better chance of getting a great player high in the draft. But those chances aren't that much better. They're not so much better that they're even close to worth passing up the chance to pick three of four times in the second round as opposed to once at the top of the first. Not only are you getting more value in the second, but you even have a better chance at getting a Pro Bowl player taking more players lower.

The author's other point is that you can't separate value from performance. This isn't a sport like basketball where you need to have that LeBron James or Dwayne Wade. You're better off going for three or four guys who are good as opposed to that one star. Moreover, as already pointed out, you're still about as likely to find that one star using the second method.

And beyond that, you have to calculate in the fact that trading down mitigates the risk of the Big Mistake -- overpaying for a mediocre performer who eats up a roster spot and money when a lower pick would have outplayed him. Selecting Laurence Maroney instead of trading down for Maurice Jones-Drew, Leon Washington, Joseph Addai, Jerious Norwood, or even Jerome Harrison was a major blow to the Patriots. But passing up Percy Harvin to get Darius Butler, Julian Edelman, Brandon Tate and a #2 this year is more like a wash. The negative of the Big Miss usually far outweighs the negative of passing over a high-ranked talent in favor of a collection of lower-ranked players who together are likely to have at least some value.

Just take the example of Maroney. You can try this yourself; just take the value chart and divide the 800 points the 21st pick is supposedly worth any way you like, and look at which players you might have gotten instead. Pick two numbers that add up to 800 at random and you'll see that in almost every case you did immensely better trading down, with the best results coming when you trade for two seconds or a low first and a high third. Just randomly, going by the chart and not looking at the draft results at first, here are some of the combos I came up with:

1. Nick Mangold (640 points) and David Thomas (160)
2. Daniel Bullocks (500 points) and Maurice Jones-Drew (300)
3. Joe Klopfenstein (440) and Bernard Pollard (360)
4. Marcus McNeil (400) Daryl Tapp (276) and Freddy Keiaho (124)
5. Cedric Griffin (420) and Greg Jennings (380)
6. Daniel Manning (480) and Richard Marshall (320)

It's not a terribly deep insight, but basically low first round and second-round players are not much of a worse bet than first rounders. They are worse, but not as much as the draft chart indicates. So if you can get two of them for the price of one, you do it almost every time, try to avoid drafting Joe Klopfenstein or Ron Brace, and win that way. The key thing is that trading down mutes the misses. If you select a Laurence Maroney in the second, it's not so bad, because he is producing a little, giving you something of value -- and if you also got a Tony Scheffler or a Chris Gocong in the package, it's not a bad deal at all.

If you're the Patriots this year, you definitely want to trade down into the second from #22 -- especially since this is a deep draft. Would you really rather have Ryan Matthews than, say, Montario Hardesty and Kareem Jackson? Or Ricky Sapp and a 2011 #2? This just seems like a no-brainer to me, and I'm glad the Patriots seem to be alone in pursuing this strategy. The perception that they missed by not picking Percy Harvin and Clay Matthews will only deter other teams from doing the same thing.

 

[patchick]

It's not a terribly deep insight, but basically low first round and second-round players are not much of a worse bet than first rounders. They are worse, but not as much as the draft chart indicates. So if you can get two of them for the price of one, you do it almost every time

Agreed; my main complaint, as I said above, is simply that something routinely stated in football circles becomes headline news when announced by An Economist. (I also think the significance of BB's economics degree tends to be overstated. At a liberal arts college, "economics" is the standard pre-business major, generally the most common major on every school's football team. And nobody makes a big deal about a business major.)

If you're the Patriots this year, you definitely want to trade down into the second from #22 -- especially since this is a deep draft. Would you really rather have Ryan Matthews than, say, Montario Hardesty and Kareem Jackson? Or Ricky Sapp and a 2011 #2? This just seems like a no-brainer to me, and I'm glad the Patriots seem to be alone in pursuing this strategy. The perception that they missed by not picking Percy Harvin and Clay Matthews will only deter other teams from doing the same thing.

Heck, yeah! Let everybody keep thinking that the Pats blew it by getting ONLY Darius Butler, Brandon Tate, Julian Edelman and this year's #47 overall for their 1st last year.

 

[Nonentity]

I read the paper. I think some people are missing what the author was trying to say.

...

If you're the Patriots this year, you definitely want to trade down into the second from #22 -- especially since this is a deep draft. Would you really rather have Ryan Matthews than, say, Montario Hardesty and Kareem Jackson? Or Ricky Sapp and a 2011 #2? This just seems like a no-brainer to me, and I'm glad the Patriots seem to be alone in pursuing this strategy. The perception that they missed by not picking Percy Harvin and Clay Matthews will only deter other teams from doing the same thing.

Thanks for taking the trouble to read the paper, appreciated your post.

I'm with you most of the way but not all the way. There are many positions at which you can get away with being 'good' as opposed to 'great'. Linebacker is a good example, as is defensive back. It's better to have an entire unit of 3/4 star players than one 5 star player and the rest 2 star.


I think it's more important to have elite players on the offensive side of the ball - and you should never settle value at QB. In the right position, one player really can make all the difference.


Here's a dilemma for everyone: Tom Brady and Laurence Maroney or Matt Schuab and either Adrian Peterson or Chris Johnson?

 

[Box_O_Rocks]

Here's a dilemma for everyone: Tom Brady and Laurence Maroney or Matt Schuab and either Adrian Peterson or Chris Johnson?

Who's blocking for them?