Monday, April 5, 2010

KEARSONOMICS: Richard Thaler, whom half the members of The University of Chicago Law School Class of 1998 will remember as the guy who came into Sunstein's Elements class and ran some sort of con involving auctions and predicting other students' behavior, tells us what Sam Bradford doesn't want us to hear: the higher a player is picked in the NFL draft, the lower his value generally will be relative to his salary.

In a fantasy league beloved by a few readers here, I proposed that rather than assigning draft positions in reverse order of a team's finish, we let teams pick their draft slot. The guy with last year's worst record would get to pick any draft position he wanted, then the second-to-last-place finisher would choose from the remaining eleven slots, etc. A person could choose the remaining pick that maximizes the expected value as he calculates it, and if people had differing views, this might make more than one person happy. It's not as efficient as an auction, but it's a step in that direction and a heck of a lot simpler.

The value of a lower pick in a fantasy draft comes from the higher pick in even-numbered draft rounds, but the existence of real-world salary issues (either hard-cap or just soft-budget) in the NFL creates similar value in trading down in the NFL draft. Instead of assigning the #1 pick to the worst team, thus forcing the team to commit significant money to a risky prospect, why not let the worst team pick where it wanted to draft? For example, if the Rams thought that Bradford vs. Suh were a coin flip, they could trade down to #2 and save some money and cap room.

5 comments:

  1. Thaler is an impressive economist and might have shared a piece of the nobel price if tversky hadn't passed away. However, translating ideas from one arena to another is often not smooth. There is no inconsistency between drafting and paying Manning big bucks and wanting to find a future hall-of-famer in the 6th round of the NFL draft; every team would want to have uncovered Brady for themselves. Whoever picks at the top of the list, whether because they earned that spot by finishing last the previous year or because they traded up to get that pick, has to make a pick, obviously based on the information available at that time. They have to rank the players, according to the players' perceived talents and the team's needs. If Peyton scored (much) higher than Tom, you'd take Peyton. The choice could subsequently be shown to be wrong, but that was not known at the time of the draft.

    As for the $$$'s, they're not set (see Crabtree's hold-out in an attempt to get 7-pick dough despite being the 10th pick) in stone, and as Thaler points out, on the rookie salary scale issue the owners and the union are in strong agreement, so it will probably come in the next cba.

    ReplyDelete
  2. As for the money, as Thaler points out the rookie salary scale is about the only thing the owners and the union agree on, so it will almost certainly be part of the new cba. The current increases are certainly not set in stone, though (see michael crabtree's attempt to get 7-pick money from the 10-pick spot last year). Once we take the dollars/value ratio out of the equation, we still have Tom Brady going late in the draft and Peyton going at the top.

    ReplyDelete
  3. isaac_spaceman9:20 PM

    I think you're nibbling at the margins and avoiding the meat of the hypothesis, though.  The point is not that a sixth-round draft choice is better than a first-round draft choice.  It is that drafting at the top of the first round increases cost more than it increases benefit, relative to drafting at the middle or bottom of the round. 

    Let's say you're a GM and you have 50 players on your team.  You have a $120MM cap (basically last year's number, rounded down for ease of calculation).  You have the first pick in the first round, and you have to spend $12MM a year on that player for five years (roughly Matt Stafford's annualized salary).  Your goal is basically to win 50 regular season games over the next five years.  This is a good deal if you think your pick is going to contribute 10% of the value of your goal, or the equivalent of 5 wins all by his lonesome, for the duration of the contract, enabling you to spend the other 90% of your cap on the other 49 players.  If it's back-loaded, all the better. 

    Now you're a different GM with the same roster size and the same cap, but you pick 12th.  You have to spend about $5MM on that player for the next five years (roughly the size of Knowshon Moreno's contract).  Your goal also is to win about 50 games over the next 5 years.  For you to get a good deal, you need your guy to contribute about 4% of the output of a good team, or about 2 wins, leaving you 96% of your cap to get other players. 

    Of course, you have no idea which players are going to pan out.  You do have an idea, based on draft slot, of what the expected value of each slot has been in the last 25 years.  I don't know how Thaler is doing is calculations, but essentially he's saying that in our hypothetical, the #1 players have produced an average of less than 5 wins and the #12 players have produced an average of less than 2 wins over the course of five years.  But say the #1 players have produced an average of 2.5 wins and the #12 players have produced an average of 1.5 wins.  Would you rather commit 10% of your salary to a player who will contribute an expected 5% of the output you need, or 4% of your salary to a player who will contribute an expected 3% of the output you need?  The cap and scale (assuming they exist) basically penalize the teams at the top of the draft by forcing them to commit greater resources to risky players, leaving less money to fill out the rest of a quality roster. 

    there are obvious questions about the analysis, but the premise seems plausible to me. 

    ReplyDelete
  4. it's certainly true that the dollars/value proposition has gotten out of whack at the top of the draft as top end salaries have sky-rocketed. that's why, as thaler points out, there will almost certainly be a rookie scale in the new cba. everyone knows this -- owners, players, gm's, fans. everyone also knows that finding a future hall-of-famer in the 6th round is a good thing. it seems to me that thaler is pointing out the obvious and something that is indeed actually being addressed. everything is working as it should, right?

    ReplyDelete
  5. isaac_spaceman2:18 PM

    I think one thing that he's pointing out that not everybody knows is that the going value for a trade-down is wrong.  And also that there is a de facto rookie scale. 

    ReplyDelete