Simply put, it’s an automated prediction platform born out of a desire to meld together my love for soccer and full-stack software engineering.  The core idea had been rattling around in my mind for many years and manifested itself in various forms, all eventually unsuccessful. The many iterations though did build upon themselves providing an overall framework for something that might be truly viable.

The accelerant turned out to be the shuttering of my last startup attempt with SuperDraft. It was a long, arduous haul not without merit or success but eventually succumbing to a brutally tough market requiring much deeper investment in user acquisition than we were able to provide.

The huge takeaway I had, outside of how much fun an individual can have melding sports and engineering together, is the absolute advantage the house has over the consumer and that with a little tribal knowledge and effort that disadvantage can easily be turned on it’s head.

So needing to take a step back and recharge, I took the time to start researching and developing prototypes utilizing some of the ML models available from the python stack, and scikit-learn specifically. This took many iterations and finally over several beers with my good friend Matt we happened upon a strategy. One that focused efforts on a simple outcome, a binary classification problem on whether a soccer match would end in a draw or not…

From there the idea quickly grew and expanded to a point where I felt comfortable sharing with others.

Is it a cash machine, multi-year fantasy or simply a great exercise in full-stack, blue-sky development?  Time will tell…