Use Machine Learning to Earn Money… And Make a Living From It?

Photo: Ryan McGuire, Pixabay

As of today, nobody would deny that machine learning or AI is the future. Most of us would also recognize that the revolution is already here. With it, a great volume of opportunities arise and, with every opportunity there’s always profitability.

I discovered this two to three years ago, when I was deeply interested and involved in the learning of artificial neural networks. I was — and still am — passionate about machine learning and taking advantage of it. Being a 20 year-old who loved money, loved data and loved the NBA, there wasn’t any other way: I had to build a predictive model to forecast the winner of a given game and make money from it.

I’m not going to go over the technical process today, I instead want to guide you through the internal rollercoaster this was, and the final results. I aim to encourage those who doubt whether to try something out, and to show the mindset I’d adopt if I had to go over it all again.

The beginnings

I started out knowing barely nothing about machine learning. In fact, I learned it all on my own (thanks Internet). I had become an NBA fan after multiple hours playing NBA 2k15, which was gifted to me after having been in the USA for the first time and having felt in love with the country. Apart from that, I had never played with a basketball before.

But as my love for the NBA grew, I started to imagine being able to predict the game-winners. This type of competition had something I had never seen before: games played every single night. This supposed a huge opportunity if I was able to predict enough games correctly to be able to break even (or even make money).

Knowing very few about basketball, I wasn’t in a position to make predictions myself and expect to make money… That’s why I turned to my new friend, machine learning and, more specifically, artificial neural networks.

The Process

The process was a complete rollercoaster. I absolutely enjoyed doing it but, when I focused on the results instead of the process, I sometimes stressed out — obviously, when results weren’t as desired. The slogan the Philadelphia 76ers had for a while explains it to the perfection:

Trust the Process

I’m sure you’ve heard this a lot, but it’s imperative we live by it if we want to get things done without negatively affecting our psychology. The path will be hard if you’re trying to do unrealistic things, but remaining confident and focused is key to success.

I discovered that later than what I’d have preferred. People around me were telling me that I wasn’t going to make it, because if it was possible someone would have already done it. All sorts of comments I received and they sometimes affected me, but I kept trying no matter what. At some point, I got great results and I was happy and proud during some months… Until I discovered it was all based on wrong data. Then, frustration appeared and it was really hard to keep trying for a while.

If you’re a productivity nerd like I am, chances are you’ll be aware of the Pareto Principle (also known as the 80/20 Rule). In case you’re not familiar with it, it simply states that you can get 80% of the results by doing 20% of the work. This is so true with projects like this one, that we need to internalize it as soon as possible, so we don’t feel like wasting our time when working on the other 80%.

Any project you choose — be it sports betting like I did, weather forecasting to know when to create marketing campaigns to sell rain gadgets, house pricing predictions to buy when a rental is cheap, or whatever idea you may have — will have difficult times. Knowing that in advance and getting ready for it will help you develop an unstoppable mindset.

Fears

We all have fears but I think there’s no worse feeling that letting fears control you. Don’t let your fear of failing or fear of wasting your time stop you from doing this project you have in mind. It may be improbable or unrealistic, but you never know if you don’t try, and you don’t want to regret not trying.

This idea could eventually make you earn money and you could make a living from it, just as I suggested in this post’s title. So many people have taken advantage of machine learning, why shouldn’t you do it too?

If you try to start your project, go all in. But take calculated risks, obviously — it will depend on the stage of your life you’re in. But start it anyway, even if it’s as a side hustle.

Conclusion

My project, NettyBets, ended up being profitable. I managed to get the model just two weeks before the NBA season started, the 2020–21 season, and I tested it out during the whole regular season. The results were better than what I had expected:

  • 51.17 units in profits (that is approximately $5,116 for someone with a unit of $100).
  • 9.14% ROI in just 4.5 months.
  • 280 games predicted.

Is this enough to make a living? Depends on how risk tolerant you are, of course, but this could perfectly be a salary if betting good amounts of money (just like investing).

It is not my intention to brag about these results. I want to prove everyone reading this that it is possible, anything is, if you remain focused on the goals, develop a sense of discipline and shape your mindset for success.

I am not any smarter than you, I assure you. I’m just a normal guy with a clear vision and incredible stubbornness, determined to achieve what he wanted to achieve. That’s all you need and I can promise that, if you don’t quit, you’ll get what you’re looking for.

Thanks for reading my first post on Medium! I hope this can be of great use for everyone reading.

Founder of NettyBets 🏀, lifelong learner and data passionate.