Monte Carlo simulation
Replaying a season game by game tens of thousands of times to turn per-game probabilities into season-long odds.
Monte Carlo is not a prediction model in itself. It is the engine that converts single-game probabilities into playoff odds, seeding chances and championship probability by simulating every remaining game repeatedly and counting how often each outcome occurs.
The quality of the output depends entirely on the quality of the per-game inputs and on modelling the right correlations. Ignoring correlation, for example treating divisional results as independent when they are not, produces confident but wrong season odds.
Run counts matter for stability. A few thousand simulations give noisy tails; tens of thousands settle the low-probability outcomes that people care most about.
Related terms
See it applied
Every published call on AI4Gameday carries a stated confidence and is graded in public once the game is final.