I'm a football fan who spent years watching matches and wondering why certain teams dominated despite the scoreline not reflecting it. I'd see a team lose 1-0 but completely outplay their opponent, or watch a side rack up shots without ever threatening the goal. The numbers told a different story from the highlights, and I wanted to understand that story better. Most analytics tools available to everyday fans are either paywalled, overly complex, or built specifically for professional clubs with data science teams. The free options — SofaScore, FotMob, OneFootball — are great for live scores, but they don't give you predictions, expected goals context, or any sense of what's likely to happen next. There was a gap between "here are the stats" and "here's what the stats mean." That's what Scorina AI is trying to fill. I built a Poisson-based machine learning model that analyses team strength, recent form, xG data, head-to-head records, and Elo ratings to predict match outcomes before they happen. Not just win/draw/lose — but win probabilities, predicted scores, confidence ratings, and the key factors driving each prediction. The kind of analysis a data analyst at a football club would produce, but accessible to any fan for free. The platform covers the Premier League, La Liga, Bundesliga, Serie A, Ligue 1 e Primeira Liga. You get live scores with real-time xG tracking, a full match centre with H2H history and team form, player intelligence across 2,000+ players with radar charts and position ratings, and a season simulator that runs 1,000 Monte Carlo iterations to predict final standings, Champions League qualification probabilities and relegation risks. One of my favourite features is the prediction tracker. You make your own picks before matches, and Scorina AI automatically checks the results after full time and tells you your accuracy percentage. It turns watching football into something more engaging — you're not just A spectator, you have skin in the game. The technical stack is React on the frontend, Python/FastAPI on the backend, Supabase for authentication and data, all deployed on Vercel and Railway. I built everything from scratch — the ML model, the data pipelines, the UI — over several months of evenings and weekends. The Android app is also launching on Google Play soon, built with Capacitor so the same codebase works across web and mobile. I'm still early. There's a lot I want to add — more leagues, better predictions, subscription features for power users. But the core product works, it's free, and I genuinely use it every matchday. If you're a football fan who wants to understand the game beyond just the score, give it a try. I'd love to hear what you think.

Scorina AI

AI football predictions, live scores & xG analytics

AI football predictions, live scores & xG analytics