Why I built Play Ai Odds: showing the math behind football betting and the illusion of big wins
I’m Adrien Ditta, a math graduate and indie builder.
Over the years, I’ve been fascinated by how numbers can reveal the truth behind what we think we understand — and few things illustrate this better than betting.
Whether we like it or not, millions of people bet on sports every week. For some, it’s entertainment; for others, it becomes an obsession. What almost no one sees clearly, though, is the math — the real probabilities hidden behind those seductive odds.
I created Play Ai Odds to make those probabilities visible and easy to understand.
The goal isn’t to predict winners perfectly, but to show how mathematical models interpret the game, and how that often differs from what the bookmaker wants you to believe.
For example, if the model says Team A has a 63% chance to win, but the bookmaker’s implied odds suggest only 55%, that gap tells you something — about margin, perception, or public bias.
It’s not about gambling — it’s about understanding.
Information and transparency are the best antidotes to illusion.Play Ai Odds uses models like Poisson, Dixon–Coles, and Glicko to simulate football matches and compute realistic probabilities for every outcome.
If you’re curious about the math behind this, I wrote a short article explaining how bookmakers’ odds differ from real probabilistic models — and why Poisson-based approaches can reveal that gap:
👉 Why bookmakers’ odds don’t always reflect real probabilitiesWhether we like it or not, there are millions of people who bet — some even treat it like a habit.
I don’t want to encourage betting; I want to bring realism to it.If more people understood that big winnings come from extremely small probabilities, they might stop chasing illusions — or at least make more rational decisions.
If just a few people walk away from the app thinking,
“wow, that 20× bet isn’t as promising as it looks,”
I’ll consider that a success.I’d love your thoughts on one question in particular:
📊 Applicability: Do you see this as something people could use to learn, analyze, or simply explore sports data more rationally?
Thanks for reading — and I’d really appreciate your honest feedback, both on the concept and the product itself.
