Why you should build a statistical model for betting baseball

(1 votes, average: 4.00 out of 5)

Loading ...

Jeffrey | Poker Articles, Sports and Fantasy betting

Submitted by Alex, this article belongs to the Sports and Fantasy betting series.

Alex is a fantasy betting expert (compared to me at least) who has recently joined forces with one of the top fantasy betting sites, namely FanDuel. Alex has been kind enough to write some general sports and fantasy betting articles.

In fantasy betting your opponents are other sports bettors and the format resembles the SNG (Sit and Go) and MTT (Multi Table Tournament) formats found on poker sites. Briefly, the object of fantasy betting competitions is to pick a team of players within a certain sport given certain restrictions (usually an overall salary cap). Depending on your player’s performances during the games played, you score points counting towards your overall placement in the competition.

Lots of people will try to sell you on the idea of using systems or angles for betting baseball. Some will be simple: for example, always bet on the home underdog. Others will be far more complex, introducing a variety of criteria or filters for determining whether a team is worth betting on. For example, bet against a team with a losing road record in the final game of a series. While there can be some validity to some of these patterns, they can rarely be used for long term profitable betting.

The problem is that in most cases, the patterns used to ‘prove’ that the system works have not existed over a long enough sample size to provide statistically meaningful results. Or they’ve been discovered by data mining or ‘cherry picking’ what has happened to work best in the past. Using that approach does a much better job of predicting the past than predicting the future.

In the rare instances when there truly was enough of a bias in the betting ‘market’ to yield a valid angle, once the statistical data accumulates enough to be meaningful, other bettors will discover it, exploit it, and remove the inefficiency from the market…often making it a ‘play against’ in the future.

So if angles and systems don’t work, how should you approach sports betting? By building a statistical model. Typically created in Excel or a dedicated statistical analysis software package, models are basically your attempt to simulate or calculate the likelihood of various outcomes in a sporting event. The advantage of using them is that once you model what you think the odds of various outcomes are, you can compare them to what the market (in the form of the odds) says will happen and limit your bets to situations where there’s value in the form of a positive expectation bet.

In addition, if your statistical model is based on the players in each team’s lineup (as it should be), then you can use it to attack less efficient markets, such as daily fantasy sports. Your opponents won’t be using such sophisticated techniques, and over the long haul you’ll be able to grind out some serious profits.

Check out fantasy betting at FanDuel

Special promotions

Poker Pro Bankroll Series

Latest Poker Strategy Articles

Contact PokerBankrollBlog

Questions? Concerns? Comments? Contact us! We'll return your email within 12 hours.