Skip to content

Who exactly are those silly academics who aren’t as smart as a Vegas bookie?

I get suspicious when I hear unsourced claims that unnamed experts somewhere are making foolish statements.

For example, I recently came across this, from a Super Bowl-themed article from 2006 by Stephen Dubner and Steven Levitt:

As it happens, there is one betting strategy that will routinely beat a bookie, and you don’t even have to be smart to use it. One of the most undervalued N.F.L. bets is the home underdog — a team favored to lose but playing in its home stadium. If you had bet $5,000 on the home underdog in every N.F.L. game over the past two decades, you would be up about $150,000 by now (a winning rate of roughly 53 percent).

So far, so good. I wonder if this pattern still holds. But then Dubner and Levitt continue:

This fact has led some academics to conclude that bookmakers simply aren’t very smart. If an academic researcher can find this loophole, shouldn’t a professional bookie be able to?

But the fact is most bookies are doing just fine. So could it be that bookies have a good reason to allow that loophole to dangle? Could it be that a seemingly dumb bookie is actually dumb like a fox?

So here we seem to be on familiar Freakonomics territory. An unnamed academic says something stupid, only to be shot down by regular-guy “Chuck Esposito, a genial, quick-witted and thoroughly sports-fixated man who runs the race and sports book at Caesars Palace in Las Vegas.”

I’m reminded of the punchline of that joke from grade school: “Hey, man, the smartest guy in the world just jumped out of the plane wearing my backpack.” Once again, the pointy-headed academics get their comeuppance.

What I want to know is, who exactly are those academics who “conclude that bookmakers simply aren’t very smart”? Are these respected academics such as John List, Emily Oster, and Casey Mulligan? Or are Dubner and Levitt talking about some professors of communication studies at some college I’ve never heard of?

I’m reminded of this bit by that great populist David Denby of the New Yorker:

Academics have told me [Denby] that “Oz” is a mythic structure, a descendant of the Odyssey or the Aeneid, but they look at me blankly when I say that the movie is also a summa of nineteen-thirties show business.

What kinds of academics was he talking with?? I’m no expert, but even I know that Dorothy’s sidekicks were old vaudevillians.

As a professor myself, I don’t like the implicit anti-intellectualism under which the Alvin H. Baum Professor of Economics at the University of Chicago, along with various comfortably situated journalists, forms an implicit alliance with regular-guy readers against those silly unworldly academics. But maybe in these particular examples, Dubner, Levitt, and Denby have particular examples in mind. I’m certainly willing to be informed on the matter.


  1. Prison Rodeo says:

    Noam Scheiber still said it best, more than five years ago:

    • Andrew says:


      Not to get all contrarian on you, but I disagree with some of Scheiber’s article. He writes of economists’ search for instrumental variables and other identification strategies and says, “economics had a cleverness problem. How was it that these students, who had arrived at the country’s premier economics department intending to solve the world’s most intractable problems–poverty, inequality, unemployment–had ended up facing off in what sometimes felt like an academic parlor game?”

      This is not cleverness for its own sake or anything like it. Rather, I think it reflects economists’ fundamental humility: aware of identification problems (“threats to validity”), they are willing to work hard to find clean identification. I think this stuff is fine. I do agree there is a problem, but it arises in the next phase, what statistician Kaiser Fung calls “story time,” when the economist steps beyond the clean identification to make general statements about causation and policy recommendations. Story time is fine too, but we have to be careful not to assume that identification in step 1 assures anything about the story in step 2.

      Also, I disagree with this statement in Scheiber’s article: “While still a student, Levitt wondered whether money drives election results or if the better candidate simply raises more money. He ingeniously demonstrated the latter . . .” Scheiber’s right that better (in some senses of the word “better”) candidates raise more money, but it’s also true that money affects election results. Scheiber is making the mistake of assuming that the existence of causal effect A precludes causal effect B. Both can be happening.

      • Jacob says:

        I also think it’s odd that Scheiber treats Card, Kreuger, and Angrist as so close to Levitt rather than Heckman.

      • Matt says:

        Just to add my dog to the pile: Why would anyone suppose that most economics PhD students at an elite university were there to “solve… poverty, inequality, unemployment”? Not that I disagree with Andrew’s point above — sometimes the only way to expose the patterns of causality at work in the world is to do something that seems too clever by half. But it also doesn’t strike me as a mystery that Ivy League graduate students in the liberal arts would treat their work as an intellectual game.

    • MVE says:

      Levitt is quoted in that TNR article: “Sometimes you write papers and they’re less about the actual result, more about your vision of how you think the profession should be. And so I think some of my most ridiculous papers actually fall in the high-fashion category.” That’s actually a little horrifying. I’m fine if he thinks he’s just being a clever guy playing with numbers, but this is his ideal of economics? I don’t care about whether the topics are serious or frivolous (not being an economist) but how about some Research Methods 101?

      I can’t be bothered to pay attention to a guy whose papers routinely fall apart due to completely basic, Stage 1, easily testable assumptions.

      “Election year -> mayors hire more police -> less crime.” Oh wait… turns out they don’t hire more police.

      “Roe v Wade -> fewer poor babies -> less crime.” Oh wait… the years don’t even line up. And, as anyone could find out, it isn’t only poor/”at-risk” women getting abortions.

      Or the one they promoted about “Hep B -> Fewer male babies -> Asian birth gender ratio.” Oh wait… there is no gender disparity for the first birth.

      I have no problem with doing “cute” or “clever” studies. But this is like a student who comes up with a great idea in class discussion… which was thoroughly disproved around page 5 of the reading. Yeah, good job, you’re a smart kid. But you didn’t do the basic prerequisite work, and so your cleverness doesn’t mean crap.

  2. Anonymous says:

    Classic “fooled by randomness” moment.

    Given that we can focus on underdogs at home / away, and favorites at home / away (in 4 major american sports) + choose a given time frame –> some form of result is bound to come up in one of these scenarios. One could of course add more fine grained variables if needed. 30 units up (over how many plays?) is interesting, but not something I as a gambler would follow blindly.

    • Tom says:

      This is over a 2 decade time period – so I’m betting $5000 on every home underdog in that time in order to make $150 000. Maybe working for a living would be a better way to make this amount of money over 20 years…

      There’s something in the scale of the numbers as well – being up by $150 000 sounds great, but by the same maths, if you bet $10 on each home underdog, you would be up $300. I’m guessing that doesn’t sell quite as many articles.

      • Anonymous says:

        a) Gamblers would usually have far more than 1 angle. The very best sports punters make far more than anyone on this blog.
        b) I wrote about units, to avoid the misleading numbers.

  3. Ed says:

    People who actually have studied their specialty and been recognized for their studies by receiving degrees and academic positions are easy “Straw men” (and women). Especially to people who haven’t done these things and want to feel superior. Academics are too busy paying attention to what other academics are saying to waste time refuting the Gladwells and Dubners of the world. Of course there are exceptions (e.g., Pinker).

  4. Zubin says:

    why you gotta knock communication studies? I agree anti-intellectualism sucks, but so does elitism.

  5. As with any system based on the law of large numbers, the question is how much your are prepared to lose in the short run. If had bet $5000 on every home dog this season, I figure you’d be down $34,000. By my calculations, home dogs are 35-38-2. One Website I checked has it as 31-48-2. Different books have different spreads, and they may change the line up until kickoff, so with careful timing, you might have done slightly better. The $34,000 is based on 11-10 vig, but books now also adjust the vig sometimes rather than moving the line, so the $34K is an estimate.

    My impression is also that many of those home-dog winners won on the field, so if you had bet the money line instead of ATS, you’d be ahead.

    • zbicyclist says:

      Reminds me of this: in my forecasting class (I teach part time), I cover investment strategies based on moving averages because students are interested in getting rich enough to pay off their student loans and I try to discourage amateur stock market forecasting if they are using their own money. Since this is Chicago, it’s normal for a couple of students to have been worked at financial firms or done internships there.

      Their reaction: “This used to work until everybody found out about it. Now it doesn’t any more.”

      So this home underdog thing might not work any more. Or, it might just be the case that since there are millions of possible betting strategies, some will look like they work in a retrospective study, even if they didn’t really work. Or, to exactly your point, it may very well be that if you actually did these bets with a reasonable bankroll, the bankroll would go to $0 at some point, so you’d lose everything even if your expected value was positive.

  6. Jonathan (a different one) says:

    Potential inefficiencies in sports betting markets is actually a fairly large economics literature. Less that five minutes on Google Scholar found:
    (1) Gray and Gray, Journal of Finance, 1997
    (2) Gandar, Zuber, O’Brien and Russo, Journal of Finance, 1988
    (3) Golec and Tamarkin, Journal of Financial Economics, 1991

    Of course, the man most known for this work, Raymond Sauer at Clemson see ( has argued (persuasively, IMO) that football betting markets are efficient, but I don’t think Levitt is out of line claiming that many economists diagree with him in work published in top journals.

    • Anonymous says:

      Inefficiences were easy to find in the 90s, and you could make a living doing that. Far more difficult now, where (most) bookmakers shut winning accounts down.

      • Jonathan (a different one) says:

        Note that some individuals making money is not a proof that the markets are inefficient. It simply means that arbitrage opportunities are not scalable. And your statement about bookmakers shutting winning accounts is just one example of that. Note that anyone can go into a sports book in Las Vegas or Reno and place a bet without telling anyone who they are, so there is no “account” to shut. And one could always dragoon an arbitrarily large number of people to make window limit bets. Market efficiency posits that the line would move too quickly for this strategy to succeed (especially given the heightened transaction costs of paying all these runners.

        • Anonymous says:

          Comments like that always make me smile :). I can assure you, that if you are a winning customer and keep coming back to the same Vegas sports shop that your bets will be limited (whether you are in the UK or in Vegas – Asian markets are far far bigger though). And yes, you could find helpers to place your bets – but that costs resources and makes this less interesting. I am not saying that inefficencies don’t exist – just that given that the papers you cited were from the 90s, one has to keep in mind that the game is totally different these days.

    • Andrew says:


      But, is finding potential inefficiencies in sports betting markets the same as “concluding that bookmakers simply aren’t very smart”?

  7. Jeremy Miles says:

    When you’re betting on sports, you’re not betting against the bookies, you’re betting against the other punters.

    The bookies don’t care (much) how likely an event is. They care how likely people think it is.

    The bookies will (almost) always adjust their odds to make sure that they are in profit whatever the outcome. And if they can’t, they’ll often close the betting on an event (at least in the UK, which I know more about). Recently there were a lot of large bets placed on the nomination for the next archbishop of Canterbury, presumably because they weren’t convinced they could get enough bets the other way to make a profit “After weeks of uncertainty surrounding the appointment, which has resulted in the Church of England being criticised for persisting with a secretive method of choosing its next head, a spokesman for Ladbrokes said it had stopped taking bets on Tuesday after a sudden run of bets on Welby – two of them from new accounts – had raised speculation that he had been chosen.”

    • Bill Jefferys says:

      You beat me to the punch. Basically, the bookie makes Dutch book against the punters (as a group). It doesn’t matter what the results of the contest are, the bookie is guaranteed to make money. Thus, it is completely irrelevant whether or not there is a strategy for making money by betting the home underdog.

      I wasn’t aware of the situation in Britain where the betting can be closed (as in the example given). That is interesting.

      • Jonathan (a different one) says:

        Read the article referenced, Bill and Jeremy: the bookie in question is discussing deviations from Dutch book to exploit the bias.

        • Bill Jefferys says:

          I read the article. There is a range of “lines” that will make Dutch book, not a single point spread. A bookie could use the strategy to bend the curve without breaking the Dutch book situation. The point is that in general bookies care about how much is being bet on each side, not on some underlying odds as in the underdog-home game situation. The most important thing for a bookie is not to be in a situation where he might lose money.

  8. Mark Palko says:

    In Denby’s defense, there is a large body of scholarly analysis that applies the standard critical tools to popular art, often with results that seem silly to those outside of the field (and sometimes in it). I once heard a media studies professor say that the key to understanding I Love Lucy was the class conflict between the Ricardos (workers) and the Mertzs (Capitalists) and I’ve lost count of the Hitchcock essays that read deep meaning into what appear to be skillfully executed entertainments.

    Oz has also inspired its share of pretentious sounding papers like this

    . “The Fable of the Allegory: The Wizard of Oz in Economics” Journal of Economic Education

    I’m not saying I agree with everything in Denby’s piece, but I do think it’s worth considering