Skip to contents

Play-by-play of a finished game, newest action first, with the running score and lead.

Endpoint: GET https://fibalivestats.dcd.shared.geniussports.com/data/{match_id}/data.json

Rows in the feed's order: newest first by game time, so row_index 0 is the last action (normally the game end row); sort on action_number for entry order. score_1 / score_2 are always integer. qualifier joins the qualifier codes with commas. clock, action_number and previous_action are NA in the 2014-15 feed. Overtime restarts period at 1: read it with period_type.

Usage

genius_pbp(match_id, return_parsed = TRUE)

Arguments

match_id

Genius match id (match_id of genius_schedule(), genius_widget_games() or genius_competition_games()), e.g. 2902690.

return_parsed

Parse into a tibble (default TRUE); FALSE returns the parsed body or page record (a list): data.json as served plus _request (the ids asked for), or the record parsed from an HTML page.

Value

A hoopR_data tibble with one row per play-by-play event. Its columns are listed in return_columns (subset(return_columns, fn == "genius_pbp")) and in the searchable return columns article.

Details

Access. The Genius hosts publish no robots.txt and no terms. CloudFront refuses datacenter IPs on the LiveStats and hosted hosts: from a server set options(hoopR.proxy = ...) or the GENIUS_PROXY environment variable to an http(s) proxy URL (never printed or logged). Requests are spaced at least one second apart per host; keep crawls slow and incremental. Some league sites (the NZ NBL) forbid AI crawlers in their own robots.txt, which does not cover these hosts. Unofficial use of a public source.

Org and season. Every competition is an ORG code plus a competition id (comp_id): genius_competitions() lists an org's competitions with the sdv END-year season (a START-year org lists 2025-26 as season 2026), genius_match_competition() maps a game to its competition, and data.json itself names no date, competition or season. The hosted ORG can differ from the LiveStats ORG of /u/ (NZNBL vs NZN). Match ids are one global sequence across every Genius customer. data.json carries no global team or player id: tno (1 or 2) and pno are in-game numbers; only finished games are read.

Errors (classed conditions, all inheriting hoopR_error): hoopR_no_data when Genius has nothing for the request (HTTP 404, the S3 AccessDenied 403, an empty 200 from the hosted host, an empty recent-games window, a schedule or widget with no matches); hoopR_bad_request for HTTP 400 / 422 or a URL that is not a Genius route (e.g. a non-numeric id); hoopR_asset_fetch when the answer is unknown: the CloudFront "Request blocked" 403, a rate limit or 5xx, the intermittent HTTP 500 "A website error has occurred" page (transient = TRUE: retry later), a dead connection, an error envelope, a cut-off page, an unfinished game, or a body for another match, competition or ORG.

Examples

# \donttest{
  try(genius_pbp(match_id = 2902690))
#> ── Genius LiveStats play-by-play ──────────────────────────────── hoopR 3.2.0 ──
#> ℹ Data updated: 2026-11-13 21:13:47 UTC
#> # A tibble: 663 × 25
#>    match_id row_index action_number previous_action period period_type gt   
#>       <int>     <int>         <int>           <int>  <int> <chr>       <chr>
#>  1  2902690         0          1052              NA      1 OVERTIME    00:00
#>  2  2902690         1          1051              NA      1 OVERTIME    00:00
#>  3  2902690         2          1048              NA      1 OVERTIME    00:00
#>  4  2902690         3          1044              NA      1 OVERTIME    00:02
#>  5  2902690         4          1045              NA      1 OVERTIME    00:02
#>  6  2902690         5          1043              NA      1 OVERTIME    00:02
#>  7  2902690         6          1042              NA      1 OVERTIME    00:02
#>  8  2902690         7          1040              NA      1 OVERTIME    00:02
#>  9  2902690         8          1041              NA      1 OVERTIME    00:02
#> 10  2902690         9          1039              NA      1 OVERTIME    00:02
#> # ℹ 653 more rows
#> # ℹ 18 more variables: clock <chr>, tno <int>, pno <int>, player <chr>,
#> #   shirt_number <chr>, first_name <chr>, family_name <chr>,
#> #   international_first_name <chr>, international_family_name <chr>,
#> #   scoreboard_name <chr>, action_type <chr>, sub_type <chr>, qualifier <chr>,
#> #   success <int>, scoring <int>, score_1 <int>, score_2 <int>, lead <int>
# }