About DartsPredict

DartsPredict estimates the probability that one professional darts player beats another. The production model path is intentionally simple first: ingest completed matches, update ELO and Glicko ratings chronologically, then display calibrated head-to-head probabilities.

Model shape

Darts is a clean one-on-one sport, so the first version does not need team, doubles, or surface logic. Each player gets a global ELO and Glicko rating. Glicko rating deviation keeps newer or inactive players from being treated as certain too quickly.

Data plan

The first adapter should optimize for historical completed matches, not live scores. Once a backfill exists, we can add a smaller incremental adapter for current PDC events and use PDC Order of Merit as a public benchmark.

  • PDC public GC APIs: Official rankings, completed fixtures, upcoming fixtures, tournament metadata.
  • Darts Orakel / darts stats databases: Deeper historical/statistical backfill once PDC official coverage is exhausted.
  • Mastercaller: Event pages, match histories, player aliases, tournament structure.

These are model probabilities for entertainment and analysis, not betting advice.