Probabilities
A calibrated win probability for any matchup, from Elo ratings rebuilt over every match since 2023.
Context
Head to head, recent form, pair synergy and career stats on the same player and pair IDs.
History
Results, draws and rankings from past seasons to build and backtest your own models.
Pre-match, not in-play
This guide is about pre-match data, and the line is worth drawing clearly.
The Padel API is not an official betting data feed and is not a source for settling in-play markets. If you operate live markets, the official data comes through the tour’s official betting partner. The API sits next to it: it gives you the context to understand a match before it starts, which is what most of the content and models around betting actually run on.
Who this is for
Prediction websites
A preview page for every match on the order of play: probability, head to head, form, ranking. Generated automatically, every day of every tournament.
Affiliates
Padel content that ranks and converts: match previews, tournament outright odds, “who will win” pages, updated as the draw moves.
Tipsters
A calibrated baseline to compare against. When your pick disagrees with the model, you know it is a call worth explaining.
Models and bots
Clean historical results and features with stable IDs, ready to train, backtest and run every morning against the next day’s matches.
Win probabilities: the core signal
Every professional player carries an Elo rating, updated after each completed match. The prediction model turns the rating gap between two pairs into a win probability, calibrated so that when it says 75%, the favourite really wins about 75% of the time. It has been validated against more than 7,500 professional matches.POST /api/matches/simulate takes any two pairs and returns the probability, whether or not they have ever met:
Calibration is what makes a probability usable in a betting context. Comparing a model’s number with a market’s implied probability only makes sense if the model’s 70% means 70%. The prediction guide publishes the calibration table, the accuracy figures and the band where the model is weakest.
Head to head
GET /api/matches/{id}/headtohead returns every previous meeting between the two pairs of a match. POST /api/matches/headtohead does the same for any combination of pairs and players, which matters in padel: partnerships change every season, so the rivalry between two players usually runs across several partners.
Recent form and pair synergy
The rating says how strong a player is. Form says whether they are winning more or less than that strength predicts lately.GET /api/players/{id}/form: expected against actual win rate over a recent window, with 7, 30 and 90 day zooms, the current streak and deciding sets won and lost.GET /api/pairs/{player1}-{player2}/synergy: whether two players together win more than their individual ratings predict.
delta and when it is too early to trust it.
Pair and player statistics
For markets beyond the winner, such as sets, tiebreaks or a match going the distance, the stats endpoints aggregate the full match history:/players/{id}/stats: career win rate, titles, finals and best round, filterable by date and round./pairs/{p1}-{p2}/stats: the same for a specific partnership./matches/{id}/stats: serve, return, break point and streak metrics for a finished match, set by set.
Historical rankings and ratings
A model needs to know where a player stood at the time of each match, not just today. The player rankings endpoint returns the weekly history of the official ranking, the race and the Elo rating:A daily preview pipeline
Most prediction sites run the same job every day: take tomorrow’s matches, enrich each one, publish. With the API it is a short loop:Use it responsibly
A calibrated probability describes many matches, not one. Even a 90% favourite loses one match in ten, and the prediction guide lists what the model cannot see: injuries, conditions, motivation, a long match the night before. Present numbers as probabilities, not certainties, and follow the gambling advertising rules of every market you publish in.Which plan you need
A prediction site publishing previews for every match usually fits in Pro. Models and bots that need form and synergy, run large backfills or want changes pushed instead of polled are a Business use case.
Choose a plan
Compare plans and subscribe. Upgrade or cancel at any time.
Talk to us about Business
Building a betting product or a model at scale? Tell us what you need.
Frequently asked questions
Is the Padel API an official betting data provider?
Is the Padel API an official betting data provider?
No. The API provides pre-match and historical data: ratings, probabilities, head to head, form, stats, results and rankings. Official in-play data for settling live betting markets is distributed by the tour’s official betting partner.
Can I use the probabilities as odds?
Can I use the probabilities as odds?
You can compare them with odds. A probability converts to fair decimal odds as
100 / probability, without any margin. Whether the gap between that and a market price is an edge is a question the model alone cannot answer: it does not see injuries, conditions or team news.How far back does the data go?
How far back does the data go?
Complete draws from 2023 onwards, and finals from 2006. Ratings are built over the complete draws, so every player active since 2023 has one.
Does it cover women's padel?
Does it cover women's padel?
Yes. Ratings, probabilities, head to head, form and stats are computed the same way for both categories.
How quickly do ratings update after a match?
How quickly do ratings update after a match?
New results are applied daily, so a simulation run the morning after a match already includes it.
Next steps
Match prediction model
How the probabilities are built, calibrated and validated.
Head to head
Every previous meeting between two pairs or two players.
Form and synergy
Who arrives in form and which pairs beat the sum of their parts.
Your first API call
Get a token and make your first request in a couple of minutes.