SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com » Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisWhat’s a draft pick worth? (part II)
A couple of posts ago I summarized the work of Cade Massey and Richard Thaler who, after studying the NFL draft rigorously, claim that it is not what it seems. Where I hope to add to the discussion is by seeing if the Massey and Thaler's theoretical results actually translate to the field.
I don't have that done yet, but it will happen next week. In the mean time, the following fact is worth pointing out. If you add up the Massey-Thaler surplus value of every draft slot, the team with the first pick (i.e. the worst team) still has the most surplus value, and the team with the last pick has the least. More interesting, though, is that the difference between the most and least amount of total value is almost nothing:
Total
surplus
TM value
==========
1 3.723
2 3.717
3 3.709
4 3.702
5 3.694
6 3.686
7 3.677
8 3.669
9 3.659
10 3.651
11 3.640
12 3.629
13 3.617
14 3.607
15 3.595
16 3.582
17 3.569
18 3.555
19 3.543
20 3.529
21 3.514
22 3.499
23 3.487
24 3.475
25 3.462
26 3.449
27 3.435
28 3.423
29 3.410
30 3.395
31 3.382
32 3.366
The units are millions of dollars of surplus value. So the Raiders endowment of draft slots is theoretically worth $3.72 million in surplus performance, while the Colts' is worth $3.37 million. So the logical conclusion of Thaler and Massey's paper isn't that the worst team gets the worst draft slot. It's that the worst team gets, all things considered, the best draft slot, but only by an insignificant amount.
If you think about it, it should be obvious that, even under the Massey-Thaler hypothesis, the Raiders' collection of picks is worth more than the Colts'. If you consider #33 equivalent to #32, #65 equivalent to #64, and so on, the only difference is that the Raiders have #1 while the Colts have #224. Massey and Thaler do not claim that the #1 pick is a liability. While not as valuable as the other picks in the first round, it's still more valuable than pick #224.
Back to the main question. If, as maurile suggested in the comments to Wednesday's post, having draft slots with extra theoretical surplus value is like having extra money added to your salary cap, then we're talking about a maximum possible difference between teams that is less than half a percent of the cap. I should have figured this out as I read the paper; it says that the surplus value of the #1 pick is about $600,000 and the that of the best pick (#43ish) is around $750,000. That's $150,000. Do you want every edge you can get? I guess, but having an extra .15 million added to your nine-figure salary cap is about as much of an advantage as playing a game needing 9.995 yards for every first down instead of 10.
So Massey and Thaler contradict themselves in their own title. There is no "Loser's Curse." There just isn't much of a Loser's Blessing, either. Assuming you buy into their methodology, what they've shown is that the market has organized itself in such a way that all 32 teams have a roughly equivalent endowment of theoretical draft value each year. That's pretty neat.
This entry was posted on Friday, March 30th, 2007 at 3:58 am and is filed under NFL Draft. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
Doug, I've done a lot of writing and thinking about The Losers Curse, and I have to agree with two of the central pillars of their theory:
1)You can just about always trade down a half a round and find a player just as good as the one you passed on earlier.
2)The penalty for missing on a high draft pick is greater than the penalty for missing on a lower one. Thus, The Loser's Curse.....you're already a bad team, and now you're put into a position to make a massive mistake.
Obviously, if you have the #1 pick and you get Peyton Manning, it's not much of a curse. But if you take Ryan Leaf...........
Assuming you buy into their methodology, what they’ve shown is that the market has organized itself in such a way that all 32 teams have a roughly equivalent endowment of theoretical draft value each year. That’s pretty neat.
Is it? Shouldn't teams which are finishing at the bottom of the league have significantly more draft value every year? By the standard Gil Brandt draft value chart, the team with the #1 pick has significantly more value than other teams, right? Isn't that what the draft system is supposed to do, give teams at the bottom of the league more draft value?
If the draft is essentially giving everyone the same value, it also seems like that implies that compensatory picks could be more significant than they appear. What's the surplus value of a pick at the end of the third round?
ABW, here is the same chart but using the valuations from the "official" chart:
Also, I didn't mean to imply that the situation was "good" (nor that it isn't). I just think it's pretty cool to observe that the market has worked itself out in this way.
ABW, I think the difference between Massey Thaler and Gil Brandt is that Massey Thaler takes contract of drafted player into account. Gil Brandt just takes player value into account.
I'm hoping for a draft when every team believes in Massey Thaler, and they all try to trade down, or simply "pass" on their picks and move down like Minnesota did a couple of years ago.
Also, I didn’t mean to imply that the situation was “good” (nor that it isn’t)
Oops, I misunderstood. I agree that it's interesting that the market has worked out this way. I still think there's a question here of what the draft system is supposed to do - if it's supposed to "encourage parity" or otherwise help teams at the top of the order, it's not doing that. Should it? I think yes, or at least more than this research suggests it is.
I also went and looked briefly at the paper, and looking at their chart(fig 8 in the paper) and a pick at the end of the third round appears to have almost the exact Massey-Thaler value as the first pick in the draft, which suggests to me that teams are potentially being seriously overcompensated for losing free agents.
A couple of questions:
--If the picks at 25-50 provide the most "bang for the buck" in the draft, should we expect that surplus value to decrease as agents for players drafted in those rounds realize that? If I were an agent of a player drafted in this range, I would be using this study to my advantage.
--Did the study take into account the differing lengths of contracts in the various rounds? First rounders get a lot more years than others, and more guaranteed money (thus making the length of the contract more certain). This is bad when the player is a bust (harder to cut), but good if they become a starter, because they are being compensated at least reasonably, will not become a free agent earlier, and will not have as much incentive to renegotiate or hold out as a 2nd rounder with similar production. Thus, they will be playing under the original contract longer.
I can think of several 2nd to 3rd rounders who outperformed their contract early, hit restricted free agency earlier, and thus the surplus value disappeared earlier. Briggs is the current example. Portis was a tremendous value for the Broncos for his first 2 years, but this disappeared when he demanded a bigger contract.
Re: 3, 5
Wow, Doug's list in #3 makes the Brandt value chart not look very reasonable. Even if every draft pick signed the exact same contract, I don't think that the set of picks the #1 team gets would be worth 3.8 times what the #32 would get. Even taking into account only player value, it looks to me like the "official" chart is overvaluing top picks.
Does anybody know if there is anyplace online that I can find a database of players and the number of games they have played each season? I know Doug has that here for the skill players, but I want games played data for all positions. I know I can get the info on nfl.com by looking at each player one by one, but I want a table listing all players, and their games played.
Does This Work?
I think if the NFL wants to help the worst teams more, then they need to do a better job regulating salaries. If every 1st round pick earns the same amount of money, then obviously the higher picks are more valuable.
Obviously, a system where the 1st player earns the same amount of money as the 32nd player isn't fair to the players, but there has to be a middle ground between that system and the current system in which there is an incentive to be higher drafted, yet the cost isn't so high that it penalizes the worst teams.
Isn't this whole "bang for your buck" what Moneyball is all about.
Why not look at it like this: mobility. Does is drafting early help you not draft-so-early next time around. The Saints went through the roof. The Texans, not so much.
I still think there’s a question here of what the draft system is supposed to do - if it’s supposed to “encourage parity” or otherwise help teams at the top of the order, it’s not doing that. Should it?
Economists have long argued that drafts do not promote balanced competition, but instead merely transfer league profits from the players to the owners by restricting rookie salaries far below their free market values. See this paper for details and a list of references.
I didn't read the paper, but I bet the draft increases the salary for most players. The top 2 or 3 players in the draft may have gotten more if there was an auction for their services, but I think the salary "slotting" system benefits everybody else.
Reggie Bush may have gotten more last year, but Mario Williams probably would have gotten less (yeah, that is different from what I just said, but last year was unusual).
Something to consider--
Either every GM in the NFL hasn't figured out that pretty much all 1st round picks are worth the same, or every one of them is hoping to trade their pick in the upcoming draft. Why else would crappy teams lose on purpose in the final weeks to improve their draft position?
Some responses to some posts in this thread:
JKL, the Massey-Thaler paper looks at player performance for the first 5 years of their career and then at compensation in their 6th year(when presumably they have had a chance to sign a free agent contract) to determine the "surplus value" that a draft pick provides. So, no, it is not really taking into account length of contract except that it expects that a 6th year player has signed a free agent contract. I agree this could be significant - 1st round contracts are generally longer than late round contracts - in fact, I know the Patriots insist on maximum length contracts for their 1st round picks, which presumably increases the surplus value that can be gotten from a 1st round pick.
Re: 12
Jim A, that's an interesting paper. I find the argument that there are other effects which are more economically imperative than promoting competitive balance from the draft interesting, especially the contention that
"The greater the freedom of teams to buy and sell players' contracts, the less balance will be created by a draft and reserve system, and the less damage it will cause to social welfare by shifting winning from places where consumers value it more to places where consumers care less about ``being number 1.''"
Has this happened? I can't recall a team ever losing on purpose, to improve their draft position. Do you have an example?
In fact, the 2005 49ers won their final 2 games. Had they lost them both they would have had the first pick.
Concrete evidence? No. And maybe teams are so disheartened by their crappy season that it just looks like they are trying to lose.
Notice who the Niners played in one of those two games they won? I'll give you a hint--they ended up with the 1st pick. The other team was the Rams, who finished a dismal 6-10. I wonder how many teams they jumped in front of with the loss.
Call me a conspiracy theorist--and it will never be proven, lest they be called for collusion--but I guarantee there are some coaches putting on a show the last game or two. This year was the first year I don't remember it clearly happening, although I haven't looked very closely at all the 6-10 teams.
But I can't recall any team being blamed for intentionally losing late in the season in order to improve their draft pick. I just don't think it happens. I think if it's a coach with his job on the line, he wants to win every game possible.
If the coach doesn't have his job on the line, he wants to get good looks at his best players to see which guys he wants to keep for next year.
Plus, in the NFL, having the first pick is usually not going to turn your team around instantly. No matter how good a player is, he usually is not going to make much of a difference, by himself, to a 3-13 team. Unlike the NBA, where a top player can single-handedly turn around a bad franchise.
Jim A (12), surely in the context of the NFL the draft transfers profits from rookies to veterans, given the salary cap/floor?
Monkeytime (11), don't be fatuous. The Saints improved primarily because of the acquisition of Drew Brees, which had nothing to do with their draft position. If you were to list the changes on the Saints in 2006 in descending order of their importance to the team's turnaround, it would go something like
1. Drew Brees
2. Sean Payton
3. Emergence of Jammal Brown as a franchise left tackle
4. Marques Colston
5. Radical improvement in play of Deuce McAllister
6. Reggie Bush
The Texans will forever regret not spending a first round pick on a guy who instead went to the Saints . . . but that guy is Jammal Brown, not Reggie Bush.
If the coach's job is on the line, and the owner wants a better draft position, what do you think will happen?
Am I alone in this? I don't care enough to look it up. I've just seen teams lose games they should have won after they were eliminated from the playoffs. I'm starting my own company this month, so I don't really have time to look it up, but when things settle I wouldn't mind if the hosts revisited this topic.
Mr Shush, you clearly didn't watch many Saints games.
Yes, they made some remarkable acquisitions in free agency and had an OUTSTANDING draft, but to say Reggie Bush had minimal impact is insane.
First off, why do you think Deuce's play miraculously improved so much?
Yes, Bush was mediocre at best running the ball, but do you have any idea how many receptions he had? How opposing teams had to change there defenses to attempt to contain him? Bush, Drew Brees, and Colston (when healthy) should all be tied for 1st in the turnaround department.
Ben, I think you might be fishing for a conspiracy here. I looked at all teams finishing 5-11 or worse, 1997-2006.
Here are some numbers:
winning percentage, games 1-12; 0.249
winning percentage, games 13-16; 0.250
When we break it into two week periods, these bad teams had their second highest winning percentage in games 15/16 (41-105, 0.281). The lowest period? Weeks 3/4 (29-117, 0.199). If anything, these bad teams win slightly more in the last two weeks than they had all year. Yeah, 0.281 is still a bad percentage, but these are just bad teams.
Now, is it possible that some players quit on their coach, sure. Oakland in 2005 seemed to throw in the towel after some close losses mid-season. But they also had some injuries, so I wouldn't say that with confidence. But for every Oakland 2005, there is a team that had a long losing streak early, and went 3-2 down the stretch, though it cost them draft position. I dont see any evidence that NFL organizations try to lose to get a better draft position.
Jim A (12), surely in the context of the NFL the draft transfers profits from rookies to veterans, given the salary cap/floor?
I'd agree it's probably a little of both to veterans and to owners. The cap is still soft enough to give owners some flexibility to control how much they choose to budget toward player salaries, which can vary from franchise to franchise even over the long term. And of course, the cap itself also transfers profits from players to owners by placing a drag on salaries.
The point is that there is economic evidence that the purpose of the draft extends far beyond promoting competitive balance, and that the draft's effect on competitive balance is probably overstated.
First off, why do you think Deuce’s play miraculously improved so much?
The ACL injury, strangely. He wasn't able to move laterally as well, and so therefore tended to just attack a given hole. He's said as much in interviews, and really, you can see it in the games as well. He's a much more consistent runner now.
Pat, pound for pound, committees perform better than a runner carrying the load himself. That's a lot of what happened with Deuce.
Yes, his recovery was nothing short of miraculous. Yes, he was not nearly as shifty, but was very good at power running. But I'm glad I wasn't the defensive coordinator dealing with those two-back sets. Deuce was made much better by the Bush (and Brees, and Colston). It was the perfect storm.
I'd really like to see how he does two years removed from the injury. That offense could be downright scary.
Ben,
In two-back sets, Deuce averaged 4.0 YPC; in single-back sets, he averaged 4.9 YPC. http://sports.espn.go.com/nfl/players/splits?playerId=2571
Wow. I had no idea that information was available on ESPN.com.
I see he had a 57-yarder from single-back formation, so that run was worth 0.6 of his single-back average for the season. I'm sure down and distance had a lot to do with it. On short yardage plays is probably when they brought in Karney to block for Deuce, hence a lower average.
Also, how do they count plays where Reggie Bush came in the game and lined up at WR?
Deuce's real first name is Dulymus???
Hmmm...Reggie Bush was also better running from the single-back formation. (But still not as good as Deuce.)
Deuce averaged 4.2 when there were no players in motion and 4.6 when players were in motion. I'm guessing that Reggie Bush was often that player in motion.
Richie made my point for me. You want to woo me with stats, tell me how Deuce did when Reggie was OFF the field.