Just How Bad Are Football Pundits at Picking Winners?

Answer: pretty bad! From a 1999 Journal of Business paper by Chris Avery and Judy Chevalier:

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



View All Comments »
  1. Fancycwabs says:

    “Beating the spread” is not the same thing as “winning.” All this shows is that The Boston Globe is pretty good at setting the spread.

    Well-loved. Like or Dislike: Thumb up 13 Thumb down 4
  2. Kevin says:

    Shouldn’t this post be titled “Professional Spread Makers are Pretty Good”? There are only 2 pundits which are outside of the profitability range of a standard 10% Las Vegas spread (Freeman and Eisenberg), and Freeman has a very small sample.

    Well-loved. Like or Dislike: Thumb up 12 Thumb down 2
  3. Nick Ryder says:

    Does a 50% success rate vs the spread make them bad, or does it make them equally good as the betting house?

    Thumb up 0 Thumb down 0
    • Levi says:

      Folks taking bets (either professional sports books or bookies) include a vig in the line, which means that you need to win above a 50% clip to maintain profitability.

      Thumb up 2 Thumb down 0
  4. Nylund says:

    In general, sports suffers from an East Coast media bias. East Coast teams often get more coverage (and love) from sports experts. Star players also sometimes warp expert judgements. Single players make for good articles and stories, and I think their influence on the probability of winning can often get overstated. Team sports are about teams, not superstars. Together, I’d imagine they could routinely lead to some bad picks.

    Thumb up 5 Thumb down 7
  5. Paul says:

    The table appears to be incomplete. Surely some of the experts’ picks pushed (that is, the result of the game was identical to the spread, in which case the pick neither won nor lost).

    Thumb up 0 Thumb down 2
    • Robert says:

      A push is considered a loss since “the house” still collects its premium. Pushes make sportsbooks happy since, on the pushed game anyway, they collect from everyone.

      Thumb up 1 Thumb down 2
      • Brandon says:

        The “House” will collect on a push ONLY if it is mentioned in the bet. For example some parlay cards. A push typically results in a “void” bet. If you have a 3 team parlay off the board a push will reduce the bet to a 2 team parlay.

        Thumb up 0 Thumb down 1
  6. Jams says:

    I’d be more interested in seeing how accurate the spreads are to real game results.

    Thumb up 4 Thumb down 0
    • Ned says:

      The spreads may well be lousy predictions but still be good spreads, since they are trying to attract an equal number of bets on both sides, insuring a profit for the sports book, rather than predict the score of the game, which is what sportswriters, fans, and bettors do. The book’s spread is a prediction of the betting public’s behavior, rather than the game itself.

      Thumb up 2 Thumb down 0
    • Matt says:

      That would be sort of interesting, but the idea behind a point spread isn’t to predict the outcome of a game. It’s simply the amount of points that the underdog must be “spotted” that leads to an even amount of money being bet on both sides.

      Thumb up 0 Thumb down 0
  7. Allen says:

    Not only are so-called experts bad, but so are most computer models. In graduate school, I proposed a variation of the semi-strong efficient market hypothesis: that no outside can beat the spread on a risk-adjusted basis using publicly available information.

    Thumb up 2 Thumb down 0
  8. Eric Williams says:

    What makes this even more interesting, is that Frank DeFord was right with his prediction based on time zones for ALL THREE of the games this season,


    Well done Frank, you’re quirky method actually really did beat the professionals all season long.

    Thumb up 0 Thumb down 0