Skip to contents

PWHL and College Hockey Data in R: Women's Pro and NCAA Hockey with fastRhockey

Saiem Gilani, Ben Howell

Source: vignettes/articles/pwhl-and-college-hockey.Rmd
pwhl-and-college-hockey.Rmd

fastRhockey covers the Professional Women’s Hockey League (PWHL) end to end, and NCAA men’s and women’s college hockey through ESPN. For the PWHL you can load whole seasons of pre-built data in seconds, or call the league’s live HockeyTech feed for today’s games, shifts and on-ice Corsi. For college hockey, the espn_mch_*() and espn_wch_*() families mirror ESPN’s NHL wrappers.

What you’ll learn

  • Which PWHL loaders exist and what each returns.
  • How to build a standings table and a scoring leaderboard from the loaders.
  • Which live PWHL functions to use for schedules, play-by-play, shifts, time on ice and Corsi.
  • How to get NCAA men’s and women’s college hockey scoreboards, play-by-play and box scores.

New to the package? Start with Getting started with fastRhockey.

What is the PWHL?

The PWHL began play in January 2024 with six teams: Boston, Minnesota, Montréal, New York, Ottawa and Toronto. Seattle and Vancouver joined for 2025-26. In fastRhockey, PWHL seasons are end years, so season = 2026 is 2025-26 and the inaugural season is 2024.

Which PWHL loaders are there?

The load_pwhl_*() functions read one file per season from the sportsdataverse-data releases. They are the fastest way to get PWHL data, and they don’t touch the league’s own servers.

Loader One row per
load_pwhl_pbp() Play-by-play event, with coordinates, on-ice players and strength state
load_pwhl_xg_pbp() Shot on goal, with expected goals (xG)
load_pwhl_shifts() Player shift
load_pwhl_schedule() Game
load_pwhl_player_box(), load_pwhl_skater_box(), load_pwhl_goalie_box() Player per game
load_pwhl_team_box() Team per game
load_pwhl_rosters(), load_pwhl_game_rosters() Player per season, or per game
load_pwhl_game_info() Game metadata
load_pwhl_scoring_summary(), load_pwhl_penalty_summary() Goal or penalty
load_pwhl_three_stars(), load_pwhl_officials() Star or official per game
load_pwhl_shots_by_period(), load_pwhl_shootout() Team period, or shootout attempt

Aliases named to match sportsdataverse-py (load_pwhl_schedules(), load_pwhl_skater_boxscores(), …) return the same data. update_pwhl_db() keeps a local database of play-by-play up to date. The full list is in the PWHL Data Loaders section of the reference.

Build a standings table from the schedule 

library(fastRhockey)
library(dplyr)
library(ggplot2)

schedule <- load_pwhl_schedule(seasons = 2026)

regular <- schedule %>%
  filter(game_type == "regular", game_status != "") %>%
  mutate(home_score = as.integer(home_score), away_score = as.integer(away_score)) %>%
  filter(!is.na(home_score), !is.na(away_score))

standings <- bind_rows(
  regular %>% transmute(team = home_team, gf = home_score, ga = away_score),
  regular %>% transmute(team = away_team, gf = away_score, ga = home_score)
) %>%
  group_by(team) %>%
  summarise(
    games = n(),
    goals_for = sum(gf),
    goals_against = sum(ga),
    goal_diff = goals_for - goals_against,
    .groups = "drop"
  ) %>%
  arrange(desc(goal_diff))

standings
#> # A tibble: 8 × 5
#>   team      games goals_for goals_against goal_diff
#>   <chr>     <int>     <int>         <int>     <int>
#> 1 Montréal     30        78            41        37
#> 2 Boston       30        74            45        29
#> 3 Minnesota    30        91            73        18
#> 4 Ottawa       30        71            73        -2
#> 5 Vancouver    30        68            81       -13
#> 6 New York     30        63            83       -20
#> 7 Toronto      30        51            72       -21
#> 8 Seattle      30        64            92       -28

The PWHL awards three points for a regulation win and two for an overtime or shootout win, so for official points use pwhl_standings(), which calls the live feed.

ggplot(standings, aes(x = goal_diff, y = reorder(team, goal_diff), fill = goal_diff > 0)) +
  geom_col(show.legend = FALSE) +
  scale_fill_manual(values = c(`TRUE` = "#531896", `FALSE` = "#A6E5F3")) +
  labs(
    title = "PWHL goal differential, 2025-26 regular season",
    x = "Goals for minus goals against", y = NULL,
    caption = "Data: fastRhockey::load_pwhl_schedule()"
  ) +
  theme_minimal()

Bar chart of PWHL team goal differential in the 2025-26 regular season.

Who led the league in scoring? 

skaters <- load_pwhl_skater_box(seasons = 2026)

skaters %>%
  group_by(player_id, first_name, last_name) %>%
  summarise(
    games = n_distinct(game_id),
    goals = sum(as.integer(goals), na.rm = TRUE),
    assists = sum(as.integer(assists), na.rm = TRUE),
    points = sum(as.integer(points), na.rm = TRUE),
    shots = sum(as.integer(shots), na.rm = TRUE),
    .groups = "drop"
  ) %>%
  arrange(desc(points), desc(goals)) %>%
  head(10)
#> # A tibble: 10 × 8
#>    player_id first_name   last_name    games goals assists points shots
#>    <chr>     <chr>        <chr>        <int> <int>   <int>  <int> <int>
#>  1 23        Kelly        Pannek          35    17      20     37    72
#>  2 21        Taylor       Heise           35    13      19     32   107
#>  3 189       Britta       Curl-Salemme    34    11      20     31    64
#>  4 42        Abby         Roque           38    12      18     30    84
#>  5 72        Rebecca      Leslie          38    18      11     29   120
#>  6 32        Laura        Stacey          39    10      19     29   141
#>  7 58        Brianne      Jenner          38    12      15     27    98
#>  8 36        Jessie       Eldridge        34    14      12     26    95
#>  9 31        Marie-Philip Poulin          28    11      15     26    86
#> 10 25        Grace        Zumwinkle       34    14      11     25    99

These totals include playoff games; filter game_id against the schedule’s game_type for regular season only.

Play-by-play, shifts and xG

load_pwhl_pbp() has one row per event. Coordinates are in feet from centre ice in several frames (see the rink plots article), and strength_state gives the skater count.

pwhl_pbp <- load_pwhl_pbp(seasons = 2026)

pwhl_pbp %>%
  filter(event %in% c("shot", "goal")) %>%
  count(strength_state, sort = TRUE) %>%
  head(8)
#> # A tibble: 8 × 2
#>   strength_state     n
#>   <chr>          <int>
#> 1 5v5             5208
#> 2 5v4              796
#> 3 4v5              725
#> 4 5v3              144
#> 5 4v4              139
#> 6 5v6              126
#> 7 6v5              124
#> 8 3v5              105

A republish of the PWHL release files with corrected x_coord columns is pending, so the derived coordinate columns in files loaded today may predate the fix in fastRhockey 1.1.0.

For shot quality, load_pwhl_xg_pbp() adds expected goals to each shot on goal. How that model works, and its known issues, are in the PWHL expected goals article.

Which live PWHL functions should I use?

The pwhl_*() functions call the PWHL’s HockeyTech feed directly. Use them for games in progress, for data the loaders don’t have, or for the newest season before the loaders catch up. They are not run on this page.

# league and schedule
pwhl_teams()
pwhl_schedule(season = 2026)
pwhl_scorebar(days_back = 3, days_ahead = 3)
pwhl_standings(season = 2026)
pwhl_playoff_bracket(season = 2026)

# one game
pwhl_pbp(game_id = 27)
pwhl_player_box(game_id = 27)
pwhl_game_summary(game_id = 27)

# on-ice analytics for one game
pwhl_game_shifts(game_id = 42)
pwhl_player_toi(game_id = 42)
pwhl_game_corsi(game_id = 42)

# players and leaders
pwhl_leaders(position = "skaters", season = 2026)
pwhl_stats(position = "goalie", season = 2026)
pwhl_player_game_log(player_id = 28, season = 2026)

pwhl_game_corsi() returns on-ice Corsi and Fenwick per player for one game. To compute the same metrics for a whole season yourself, follow Corsi, Fenwick and PDO in R using load_pwhl_pbp() and load_pwhl_shifts().

The AHL, OHL, WHL and QMJHL run on the same HockeyTech platform, so ahl_*(), ohl_*(), whl_*() and qmjhl_*() offer the same set of functions (ahl_schedule(), ahl_pbp(), ahl_game_corsi(), …). See the AHL, OHL, WHL and QMJHL reference sections.

How do I get NCAA college hockey data?

ESPN covers Division I men’s and women’s college hockey. espn_mch_*() is men’s college hockey and espn_wch_*() women’s; the two families have the same functions, mirroring the ESPN NHL family. These call ESPN’s API and are not run here.

Men’s college hockey 

# games on a date (YYYYMMDD)
scores <- espn_mch_scoreboard(dates = "20250110")

# one game
espn_mch_pbp(game_id = "401717648")
espn_mch_team_box(game_id = "401717648")
espn_mch_player_box(game_id = "401717648")
espn_mch_summary(event_id = "401717648")

# season context
espn_mch_standings(season = 2025)
espn_mch_rankings()
espn_mch_teams_site()

Women’s college hockey 

espn_wch_scoreboard(dates = "20250110")
espn_wch_pbp(game_id = "401762970")
espn_wch_player_box(game_id = "401762970")
espn_wch_standings(season = 2025)
espn_wch_rankings()

A typical workflow: pull a date’s scoreboard, take its game ids, then loop over espn_mch_pbp() or espn_mch_player_box(). Keep loops modest and pause between requests. Each family has 121 men’s and 121 women’s functions, covering athletes, coaches, rankings, odds and more; see ESPN College Hockey in the reference.

Where to go next