Getting started with betbetter
In betbetter: Access the Bet Better Open Sports Model API

knitr::opts_chunk$set(collapse = TRUE, comment = "#>", eval = FALSE)

What this package retrieves

betbetter is a thin client for a free, open sports model API. For each upcoming fixture the service publishes the model's estimated probability that a selection occurs, and the decimal odds that probability implies.

Chunks in this vignette are not evaluated, because they require network access.

library(betbetter)

Available leagues

bb_leagues()

Fifteen competitions are covered, spanning Australian rules football, baseball, basketball, American football, ice hockey, association football, tennis and mixed martial arts.

Retrieving estimates

afl <- bb_picks("afl")
head(afl[, c("game", "market", "selection", "model_probability", "fair_odds")])

model_probability is a proportion between 0 and 1. fair_odds is its reciprocal: the decimal price at which a bet on that selection would break even if the model were exactly right.

Narrowing the results

strong <- bb_picks("afl", min_probability = 0.7, limit = 10)

Match-level and player-level markets can be requested separately:

lines <- bb_picks("afl", feed = "games")
props <- bb_picks("afl", feed = "props")

Out of season

A league with no upcoming fixtures returns a zero-row data frame:

nrow(bb_picks("nfl"))  # 0 during the off-season

This is expected behaviour, not an error. If the API cannot be reached, bb_picks() emits a message and returns NULL, so scripts can test for it:

result <- bb_picks("afl")
if (is.null(result)) {
  message("API unavailable, skipping")
}

Interpreting the numbers

These are model estimates, not predictions of certainty. A selection estimated at 70% is expected to fail roughly three times in ten; a run of losing selections at that probability is ordinary variance rather than evidence the model is broken.

The API publishes no bookmaker prices. There is no bookmaker identity, market price or market-implied probability in the response, because the upstream odds licence permits publishing derived work but not redistribution of the price feed. Comparing these estimates against a market price requires sourcing that price yourself.