Monte Carlo probabilities for NFL, MLB, NHL, NBA and soccer

Seven probability models by Elio Liberatore, published on Apify as APIs and as an MCP server for AI agents. Public data goes in, the rest of the season or the next match is simulated thousands of times, and fair probabilities come out as JSON, CSV or an API response.

Try the live demo Model probabilities for the coming week of matches, computed in your browser. No sign-up, no API key.

Soccer Predictions API

Title, top-four, playoff and relegation odds

Replays every remaining league fixture thousands of times and returns fair probabilities for the title, top-four and European places, promotion playoffs and relegation. Covers the Premier League, Championship, LaLiga, Serie A, Bundesliga, Ligue 1, Eredivisie, Primeira Liga, Super Lig, Brasileirao, Liga MX, MLS and more, with an optional value-bet comparison against bookmaker odds.

Open on Apify How it works and FAQ API reference Input schema

16 ready-made examples

Soccer Match Predictions API

1X2, over/under, BTTS and correct score

One probability grid per upcoming fixture: home win, draw and away win, over/under 2.5 goals, both teams to score and the most likely scorelines. Attack and defence ratings are fitted per team with a Dixon-Coles bivariate Poisson model on real match history, weighted towards recent games.

Open on Apify Matches this week How it works and FAQ API reference Input schema

9 ready-made examples

MLB Playoff Odds API

Division, wild card and postseason odds for 30 teams

Simulates the rest of the MLB schedule for division, wild card, postseason and top-seed probabilities for all 30 teams, with projected records, model-versus-market edge against live Kalshi prices and quarter-Kelly stake sizing.

Open on Apify How it works and FAQ API reference Input schema

8 ready-made examples

NFL Playoff Odds API

Division, wild card and No. 1 seed odds for 32 teams

Simulates the remaining NFL schedule for division, wild card, playoff and No. 1 seed probabilities for all 32 teams, with projected win totals, market edge and Kelly stake sizing.

Open on Apify How it works and FAQ API reference Input schema

10 ready-made examples

NHL Playoff Odds API

Playoff, wild card and Presidents’ Trophy odds for 32 teams

Simulates the NHL regular season with overtime, shootouts and the loser point modeled explicitly, and returns division, wild card, playoff and Presidents’ Trophy probabilities for all 32 teams, priced against live Kalshi contracts with fee-adjusted expected value.

Open on Apify How it works and FAQ API reference Input schema

7 ready-made examples

NBA Playoff Odds API

Playoff, play-in and championship odds for 30 teams

Simulates the NBA season game by game, including the play-in tournament and the full best-of-seven playoff bracket, for playoff, play-in, division, conference and championship probabilities for all 30 teams, priced against live Kalshi contracts.

Open on Apify How it works and FAQ API reference Input schema

7 ready-made examples

Sports Probabilities MCP Server

The models as tools for AI agents

Exposes the season models as Model Context Protocol tools, so Claude, ChatGPT, Cursor or any MCP client can ask for team odds, a full league table ranked by probability, model-versus-market edge against Kalshi, what-if scenarios and archived probability history, and get a computed number back instead of a guess.

Open on Apify How it works and FAQ API reference Input schema

4 ready-made examples

Articles

How the models work

The season models blend current form with a baseline from previous seasons, regress each team towards average according to how much of the season is left, and redraw every team's underlying strength in each simulated season so the output does not become overconfident. The results are compared with live prediction market prices where a market exists, after fees.

The match model fits attack and defence ratings for each team with a Dixon-Coles bivariate Poisson model, weighted towards recent games, and turns them into a full scoreline grid for every upcoming fixture.

What they are used for

Comparing fair probabilities with sportsbook or prediction market prices, feeding dashboards and spreadsheets on a schedule, building a season-long history of how the odds moved, backtesting a past season, and giving AI agents a tool that computes a probability instead of inventing one.

Using them

Every example above is a saved configuration you can run in one click from its page. You need a free Apify account; no third-party API key is required. The source of the NHL and NBA models is public on GitHub.