Jump to content
  • Advertisement
  • entries
  • comments
  • views


Sign in to follow this  
Mike Bossy


This weekend while I was without cable I resorted to going to the video store to rent a DVD. I brought home a movie called Two for the Money, a flick about sports gambling. Being a red blooded North American male I know and love sports betting. You name the sport and I've probably bet on it at one point or another. The good thing is that I learned long ago that sports betting is hard and not at all profitable so when I do make it to Vegas and gamble on sports my average bet size is under $10. I still have the fun without the expense. :)

Anyway this movie got me thinking about using statistical analysis to do some handicapping. Basically using trends and simulations to figure out an expected outcome for a game and then compare it against a betting line to see if it's worth betting. This weekend I wasted a bunch of hours coming up with a simple C# app that takes some basic stats on teams around points for and points against. The final results for this past weekend was 8 out of 16 picks right. Exactly the same result you'd expect from flipping a coin. :) Not amazing but at least it wasn't worse than the coin. Once I get around to adding some more variables into the simulation formula as well it might even get better.

But for this season I've already been distracted enough by this. In the offseason though I want to flesh out a more detailed system taking everything from yardage to temperature into account. I think letting a neural net go wild on something like that and comparing against the last decade or so of NFL results and maybe I can come up with a system that can pick 9 out of 16. :)
Sign in to follow this  


Recommended Comments

See if you can make it better than my default system of just picking the home team [smile].

Then see if you can train your system for the stock market and make millions!

(Actually I am interested in what variables actually matter for a sports team performance - the trick would be being able to model all the possible variables from previous performance through to injuries through to intangibles such as good/bad press).

Share this comment

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Advertisement

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!