Mistakes to Avoid When Creating Bet Labs Systems - Overfitting

Mark Gallant
Mark Gallant

Bet Labs is a powerful tool bettors can use to boost their bankroll. However, new users need to know what to do and what to avoid when creating profitable systems.

Just because a system is profitable does not mean it will be profitable going forward. One of the most frequent mistakes we see is overfitting.


Users can make systems that appear insanely good if you look at the numbers, but they're built on cherry-picked data points. Overfitting can be done by selecting too many filters or too many individual data points within a list-style filter. 


In the example above, we see a system with over 100 units won and a 50% ROI. However, it got that way because only profitable data points were chosen from three different filter types. While it's true that bettors would have profited largely by making those bets in the past, the system has no rhyme or reason to it and will certainly not be that profitable going forward. In fact, there's a decent chance it won't be profitable at all in the future. 

Note that the system got a 'D' for a grade, which you can see next to the Money Won and ROI. If you click on the letter, it will explain the grade in further detail. 


As you can see, this system has too many individual items chosen within the filters themselves. Using three filters is fine, which we can see by the 'A' in the bottom section, but when each one has over ten items selected, you're going to run into trouble.

If you still think your system would be profitable going forward, you can create it and keep track of it on the 'My Systems' page. Just toggle the results to 'Since Built' and you can see how it's performed since you made it. 


Was this article helpful?



Article is closed for comments.