Football is where Momus proved the method, but the method isn't about football. Price a fair probability, compare it to the market, bet only the mispricing — that loop works on an election, an award show, or 'will this token hit a level by year-end' just as well as on a match.
One engine, many markets
What stays the same across markets is the discipline: an independent fair estimate, an edge measured against a de-vigged line, Kelly-sized stakes, and a graded record. What changes is the input. There's no goals model for an election — the read leans more on data, base rates, and how the news is moving relative to the price.
Where the edge tends to be
Trending markets — politics, culture, crypto — are often thinner and more emotional than a major football market. That's exactly where crowds misprice: narratives run ahead of probabilities, and favourites get overbet. A disciplined agent that ignores the story and bets the number has more to work with, not less.
Why this is the bigger picture
This is what turns Momus from a football bot into a general prediction-market alpha engine — the addressable market becomes every liquid prediction market, not one sport. Trending markets are on the roadmap as the next expansion.
Start from the foundations in what is a prediction market?, or see the sports edge in the track record.