How to develop a betting model

How To Create A Sports Betting Strategy

By in
No comments

When most people think about profiting in sports betting they think about the knowledge they need to have about the sport or some people will try to work out a very simple strategy around value. The truth is this just doesn’t work. If you want to build a statistical based theory (which I recommend you do) then you will need all statistical data as well as elements that bookies do not look at too much (e.g. weather, slight injuries, coaches comments etc), these are the easiest ways to generate an edge for your betting strategy.

How to develop a betting model

Stage 1 – Initial theories

I should note that I’m in the process of writing a 5,000 word guide which shows live how I built a strategy for a specific sport and league which now generates me around 20-30 units per year (that’s only betting 1 unit per game too, none of this ridiculous 5 units on every other game rubbish!) It’s NOT impossible to build your own strategies and although this isn’t the beast guide I’m writing I wanted to outline how to build a very very simple theory of your own.

So the initial theory can come from anything, generally I like to “consciously” be thinking about generating a theory when I’m watching sports. This is probably the easiest way to get the creative brain juices flowing. But if you’re not sure just make up something related to a TRACKED STAT in a sport. Let’s take baseball for example and the stat of HITS. A very (too much so) simple initial theory could be: A team that gets the most HITS in game one goes on to win game 2 and 3.

Stage 2 – Evolve the Theory and Add Variations

 

Adding variations to a theory is so important. It gives you room for error and also allows you to find a higher probability theory that not only makes you more money, but also gives you ideas for other theories too. So going back to our baseball hits example. Our variations could be something like: If team A generates more hits but LOSES the game then team A goes on to win game 2 (and/or) game 3. Can you see the number of variations we’ve created with a very simple theory?

Taking this one step further we can add the following: Weather – This is a great one I like to use, I personally do not bet baseball but tracking wind and temperature would be a good start. Pitchers History – Under certain weather conditions, against certain hitters. And a whole host of other elements.

Another element to add would be the difference in hits. For example if a team gets over 10 hits and holds a team to less than 4.5 hits. Then we can build a 2 part theory around over 9.5 hits and conceeding under 4.5 hits in a game. You can create a theory that talks about reversion to the mean, or you can build one around current form. Either may or may not work but we don’t really need to believe in the theory just yet.

 

Stage 3 – Back-testing The Theory

The single most important element to any theory.

Sticking with our baseball example, the following is how we should back test a theory:

  • Get the data for the past 3 seasons, including odds, hits, final score & any other elements you are including in a theory.
  • Simply test the theory as if you had placed 1 unit (or placed 1.X units to profit 1 unit) and see what your figures are.

That’s it.

This is extremely time consuming and something I’ve hired assistants to help with. But it gives you 2 things. a.) Data that actually tells you if your theory is any good or not and b.) Confidence in your theory so you can bet larger unit amounts on.

 

If your theory isn’t profitable for ALL 3 SEASONS. Then it should be discarded. Remember that if you have a break-even theory that’s usually a good sign that a bit of tweaking can lead to a profitable one. I like to ensure the data shows over 30 units profit in each previous season. Anything less and you might be in the realm of cause does not equal effect. A high amount of variance can lead to a negative season which is something we of course do not want.

Stage 4 – Bet It

Stage 4 is to actually test it for a season. Remember to keep your unit size the same during the season and follow the bankroll management rules we have written about previously. Once the season is complete remember to add this data into your tracking system. If year 1 has produced at least a 30+ unit profit season (or 60%+, as sometimes there simply aren’t enough plays to get to 30+ units) then I would double my unit size on this particular theory. For example we’ve built theories for a number of sports testing the 2012/13, 13/14 and 14/15 season. We then bet the 15/16 season at £100 units for a profit of around 45 units. Now in the 16/17 season we will be using £200 units, and so on with the doubling of units until our theory gets less than 30 units profit in a season.

 

Stage 5 – Scale it, Find More, Increase Unit Size

So 30 units a year doesn’t sound like much and honestly its not. It’s not a lot of money with a unit size at 2% of your bankroll which is what I recommend. So what is very important is building more and more of these theories and scaling the technique. For example if you find 5 theories in 5 sports over the course of a year/season, then suddenly your 30 units profit per year it’s timed by 5 per sport = 150 units and then times by 5 again = 750 units. All with only risking 1 unit or 1.x units. That’s a very good Risk to Reward Ratio!

 

This is how you scale your sports betting. Don’t raise unit sizes first as this is very risky. A poor season where you only make 10-15 units profit is now seen as a loss. But when you have 25 theories across 5 sports you MITIGATE YOUR RISK and you also do something called reduce your variance. Which is a posh way of saying you allow yourself to even out luck as you have more plays.

Once you do have over 500+ units of profit per year (risking only 1 unit per play) then you can increase your unit sizes or increase to 2 units per play (which is essentially the same thing.) And now (assuming it takes you 3 years to find 25 back-tested profitable theories) you can make $100,000 per year profit with only $200 units. That’s some good risk to reward for you. Although personally I’d scale that to $2k per unit and generate $1M a year. But we do not have $100,000 in the bankroll just yet. Join us for the journey though! Sign up on the sidebar (emailed when new posts go live)

Thanks,

T.

Leave a Reply

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