Football Freakonomics: Incentives

I recently helped the Freakonomics folks and NFL Network with an installment of 'Football Freakonomics', the series of mini-documentaries on interesting wrinkles in fabric of football. This particular feature is on incentives, particularly big contracts and Super Bowl championships. They air periodically on NFL Network, I believe on GameDay Morning on Sundays.

Here is the feature.
 
After watching it, ANS readers will quickly say 'Wait! What? That's it? What about concepts like regression to the mean or the winner's curse? Wouldn't they explain what we see in the data? We did cover those considerations in detail on camera, but they just didn't make the cut. In defense of the producers, they only have a few minutes to tell an interesting story, and it has to appeal to a broad audience. It's just the nature of the medium.

For those curious, here's a brief discussion on why we see performance decline after big contracts:
  • Total performance actually tends to increase (total yds, total TDs).
  • But per play performance, which is a better measure for several reasons, tends to decrease.
  • This is likely due to 'regression to the mean', which is an inescapable tendency in any measure with a random component: Teams give big contracts to players who have 'breakout' seasons, expecting outlier performance to continue. In reality, big seasons are nearly impossible to repeat because so much depends on factors beyond a single player's control--offensive line, opponents, scheme, random luck.
  • Also, teams may feel compelled to give a newly signed star RB more carries than perhaps they should to justify their investment in him. This would explain why total cumulative stats increase, while per play stats decrease.
  • The Winner's Curse: This is a phenomenon of auctions, widely studied in behavioral economics. Signing (or re-signing) FA RBs is basically an auction for his services. In auctions, the bidder who is most likely to mistakenly overestimate the value of the prize is the one who ends up winning the auction. This is why teams (and their fans) are so often disappointed in big-name FA players.
And the decline in performance for skill players the year following their first Super Bowl victory can be at least partially explained by regression to the mean:
  • By definition, SB winning team naturally peaks in a championship season due to a confluence of factors--health, age, opponents, luck, etc. There's no reason to expect those factors to repeat themselves. If they did, we might see the same team win the SB year after year.
That's not to say that 100% of the decline is due to natural factors. Incentives certainly play a very large role in all aspects of human performance. It's just that we can't really tell exactly what's going on just with one look at the numbers.

I can't say enough good things about the producer I worked with and those behind the scenes. They genuinely worked hard to get all the numbers and ideas right. When I get my own series I promise to do a two-episode special on riveting concepts like regression to the mean. Can you say 'Ratings Bonanza'? Cha-ching...talk about incentives!

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

7 Responses to “Football Freakonomics: Incentives”

  1. Anonymous says:

    Are you sure the Alex Smith incentive is described correctly? Currently it says that he "earned an extra $2.2 million per win." I think you probably meant to say that he earned an extra $2.2 million total as a result of his 11 wins, or $200,000 per win. Otherwise he could earn $22 million in a season just for going 10-6, which seems silly.

  2. Anonymous says:

    I know this isn't your doing, but holy terrible web design. This takes a small amount of interesting information and packages it in an enormous, hard-to-use interface.

    For instance, the bit on icing kickers: can't you just say that the only difference between icing and not icing is from 50-60 yards? No, there has to be a slider so you can set various distances, and this "fact" gets its own page. Gah.

    But at least that page has a fact. The controversy page is even fluffier.

  3. Anonymous says:

    Further reasons I don't watch TV. Let's make it broad. Well, if you're 30 years old and don't want to learn basic math, you should not be catered to.

  4. Willy Hu says:

    Has this been looked at for newly drafted rookies? For example, what their draft position predicts their performance to be, how long they actually take to live up to that expectation, using that to measure a team's draft success for any given year, etc....Seems like it'd be interesting and perhaps more applicable to teams?

  5. MattieShoes says:

    I was really disappointed in their football stuff. I realize that we readers of football statistics blogs aren't who they're catering to, but...

    When they called MarTay Jenkins' kick return yards in a season record "coveted" and then compared it to Jerry Rice's receiving and Eric Dickerson's rushing numbers... Well, I died a little inside.

  6. Ian Simcox says:

    You're right about regression to the mean. It was a shame they used the previous season as a control group, rather than a group of players who performed similarly the previous year. The mathematician in me just screamed about regression. Ah well, a good piece anyway.

  7. bigmouth says:

    Brian, just a heads up the link takes you to a different Freakonomics feature (re coaching changes). Regarding the contract-year effect, I was surprised by your results finding such an effect in the NFL. That seems to contradict the findings of the only other study I've seen of this question:

    http://www.brown.edu/Departments/Economics/2009_undergrad_theses/Ben_Singer_thesis.pdf

    Any thoughts on how you reached a different result?

Leave a Reply

Note: Only a member of this blog may post a comment.