My journey to creating this blog actually began over a decade ago, after reading this article in Sports Illustrated. To summarize, the article basically explains what goes into setting lines in Vegas, and the skills and knowledge required to be a quality oddsmaker. What I was surprised to discover when reading the article was that the oddsmakers chronicled within were not necessarily concerned with making accurate predictions about the outcome of games; rather, their main aim was to set a line that would ensure even action on both sides of a given competition. This makes sense from a financial standpoint: If there is equal action on both teams involved in a given game, the casino will simply collect their commission (or juice/vig), and not have to risk exposure to losses based on the outcome (Sidenote: There is some research that suggests bookies actually do not set a line to ensure equal action on both sides, and instead attempt to set up a scenario in which one team will have both more action and a less likely chance to cover). This was interesting to me, in that it meant bookies would essentially have to predict how the betting public would allocate their money in order to determine the correct line; it meant that bookies are just as much social psychologists and sociologists as they are sports experts.

In order to explain a little more deeply what I’m talking about, consider the following scenario between two fictional basketball teams: The Brighton Bombers are scheduled to play a Monday night home game against their rivals, the Revere Redcoats. The Bombers are having a terrible season, and their record sits at 6-21. On top of that, they are playing their second game in consecutive days, and their star player has a wonky ankle. The Redcoats, on the other hand, are having a fantastic season. Not only are they off to a 19-8 start, but they are well-rested and healthy. Without any influence from the betting public, the line would stand at Redcoats -9.0. However, the Bombers have a rabid fanbase who will typically place bets on them out of sheer loyalty alone. Not only that, but the majority of bettors are unaware of the fact that so far this season the Redcoats have performed significantly better on the road. Due to these factors, an intelligent bookie has correctly set the line at Redcoats -7.5 in order to encourage more action on them.

Another facet of this discovery that intrigued me was that in order to consistently win bets, an informed gambler does not actually have to outsmart Vegas (a daunting task), but simply has to outsmart the aggregate masses of people placing money on a given contest.

The next question was this: Could I discover a consistent flaw in the betting habits of sports bettors? Was there some trend that the general population almost always overlooked, and that was taken into account by Vegas when setting their lines? Possibly due to the fact that I was in the mindset of envisioning oddsmakers as social psychologists, or more likely due to the fact that I already had some knowledge of it as a result of my time as a college Psychology major, I decided to use SPSS (or Statistical Package for the Social Sciences) software to analyze historical betting data in an attempt to spot statistically significant trends.

Sidenote: Why did I choose to analyze NBA games you might ask? In all honesty I am actually not a fan of the NBA, but I wanted to pick a sport that people frequently bet on and that had a large enough schedule to provide an adequate sample size.

  5 comments for “Story

  1. Marver
    June 3, 2016 at 4:50 pm

    I don’t know how I got here, but I was looking for a blog about betting on NBA games, since I do that.

    I liked your phrase: “it meant that bookies are just as much social psychologists and sociologists as they are sports experts.”.

    Have you ever started some software project in order to predict NBA games?

    • melitob
      June 5, 2016 at 5:15 am

      Thanks for the comments, glad you’re enjoying the blog! As for writing a program to predict games, its something I’ve thought about but I feel that at the moment my programming skills aren’t exactly up to the task. If you have any experience/advice on the matter, I’d love to hear about it

      • Marver
        June 6, 2016 at 1:16 pm

        I asked you that, because I see you have the knowledge of statistics and you know what to look for.

        Concerning programming skills, I think the difficult is learning how to “connect” the programming language to the database (all in all, it’s not that hard). You only have to learn the basics, that is, how to sum, multiply, using lists and so on (and of course, install “speak” to the database).

        About my experience on the matter, I developed a system to bet on NBA games. Almost two years ago, I started to think how I could devise an algorithm to bet. In the beginning it was not working, but improvements were made since then. And today, I think the system is ready to use. I want to start a blog to write about it (I wouldn’t show the logic behind it, only its story and the results), but I don’t know if it’s a good idea. I don’t want to lure the bookmakers’ attention. I think they could try to track my account and it could get limited or closed.

  2. LB
    December 12, 2016 at 6:00 pm

    I’m curious to know if you’re still doing NBA analysis, and if you’re an active bettor.
    I have built about 5 different math models and wager daily.
    I am interested in the idea of working with other like minded people as long as those individuals are active and committed. If interested, please contact me at the email listed.

Leave a Reply

Your email address will not be published. Required fields are marked *