Sports is something the whole world tunes into, but the tools around it are surprisingly shallow. As a fan who also bets and plays fantasy, I was tired of bouncing between a stats site, an odds aggregator, a fantasy ranker, and Google. None of which actually talked to each other or to me. Reputation-based ratings told me who used to be good; nothing told me who was in form right now purely based on data. So I set out to solve three problems: ratings that reflect current form instead of name recognition, predictions and mispriced markets grounded in real models instead of vibes, and insight that's conversational, something you can just ask, rather than dig through tables for. The approach evolved a lot, and it was entirely feedback-driven. I launched step by step in public on X and let the responses dictate what came next. It started as a pure ratings-and-prediction site: rate every player from real match data, then run statisical models (Dixon-Coles, Monte Carlo simulations) for match and tournament odds. But the replies kept reshaping it. People wanted to ask questions, not read charts. So I built "Ask 12thman," an AI analyst grounded in the live database. They wanted to act on the data, so I pulled in Kalshi prediction-market prices to flag where the market disagrees with the model. And they wanted to share, so I built a bracket simulator that lets anyone run their own tournament and download a card to post. That became the engine behind the launch, and almost every major feature traces back to a reply or comment on X.

12thman

Your analyst for fantasy sports & sports prediction markets

Your analyst for fantasy sports & sports prediction markets