nwslR is an R ecosystem that contains datasets and analysis
functionality for the National Women’s Soccer League (NWSL). Founded in
2013, the NWSL is the United States’ top professional women’s soccer
league, featuring players from all over the world. In an effort to
create a reproducible and friendly ecosystem, the setup is heavily
inspired by the nflverse. Many thanks to
Tan Ho for all his help and advice. An
earlier version of the package, created in collaboration with Sophia
Tannir, exists
here. It is no longer maintained.
Accessibility of data in women’s sports (and particularly soccer) lags
behind its male counterparts. This package is a contribution to a
growing ecosystem of open data sources that aims to make data available
to fans and analysts of the league. The main package, nwslR, provides
functions to easily access data that is loaded in the nwsldata
repo. Data is updated several times
a week during the season and more sporadically in the offseason.
If you see anything you’d like added, changed, or updated, please open up a new issue of your own. If you are interested in contributing, please contact us directly. If you use this data in any work, please cite us.
This package is currently in the process of changing and updating significantly, so it is suggested that you frequently update it so that you have access to the most recent version. CRAN submission will come when it reaches a stable state.
You can install the development version of nwslR from GitHub with:
knitr::opts_chunk$set(
fig.path = "man/figures/README-"
)# install.packages("devtools")
devtools::install_github("nwslR/nwslR")load_player_match_stats(): Loads player level stats for a given matchload_player_season_stats(): Loads player level stats for a team/seasonload_team_match_stats(): Loads team level stats for a given matchload_team_season_stats(): Loads team level stats for a team/season
load_matches(): All matches from 2016-present with information and match IDsload_players(): All players rostered from 2016-present with information and player IDsload_teams(): All teams active from 2016-present with information and team IDsload_metrics()All metrics available from scrapers with definitions. Not all metrics are available for all players/matches/teams/etc.
First, we want to load the data. While the match stat load functions
currently only take single inputs, we can use purrr to pull several
matches at once.
library(nwslR)
library(magrittr)
library(ggplot2)matches_2022 <- load_matches() %>%
dplyr::filter(season == 2022)
players <- load_players()
# * A small number of matches do not have data, so we wrap this in a safe call to ensure this runs without error
safe_load_pms <- purrr::possibly(load_player_match_stats, otherwise = data.frame())
player_stats <- purrr::map_df(matches_2022$match_id, safe_load_pms, .progress = TRUE)We then calculate percentage of shots outside of the box for each player and format labels for plotting.
player_shots <- player_stats %>%
dplyr::select(player_id, team_id, shots_total, shots_outside_box) %>%
dplyr::group_by(player_id) %>%
dplyr::summarise(shots_total = sum(shots_total),
shots_outside_box = sum(shots_outside_box),
percent_outside_box = shots_outside_box/shots_total) %>%
dplyr::filter(shots_total >= 15) %>%
dplyr::left_join(players, by = "player_id") %>%
dplyr::mutate(name_label = dplyr::coalesce(player_known_name, paste(player_short_first_name, player_short_last_name, sep = " "))) %>%
dplyr::arrange(percent_outside_box) %>%
dplyr::slice_max(percent_outside_box, n = 10)Finally, plot!
ggplot(player_shots) +
geom_bar(aes(x = reorder(name_label, percent_outside_box), y = percent_outside_box),
stat = "identity",
fill = "#3765b0") +
scale_y_continuous(labels = scales::percent, limits = c(0, 1)) +
labs(x = "",
y = "% of Shots Outside Box",
title = "NWSL 2022 Regular Seasons: Players with Highest % of\nShots Outside of Box",
subtitle = "Minimum 15 Shots",
caption = "Source: nwslR") +
coord_flip() +
theme_minimal()
teams_2022 <- c("CHI", "HOU", "NJY", "RGN", "ORL", "POR", "WAS", "NC", "KCC", "LOU", "LA", "SD")
teams <- load_teams()
team_stats <- purrr::map_df(teams_2022, ~load_team_season_stats(team_id = .x, season = "2022"), .progress = TRUE)wrangle_team_pts <- team_stats %>%
dplyr::select(team_id, pts_gained_from_losing_positions, pts_dropped_from_win_position) %>%
dplyr::mutate(pts_dropped_from_win_position = 0 - pts_dropped_from_win_position) %>%
tidyr::pivot_longer(!team_id, names_to = "pts_type", values_to = "points") %>%
dplyr::left_join(teams, by = "team_id")ggplot(wrangle_team_pts) +
geom_bar(aes(x = points, y = team_abbreviation, group = team_id, fill = pts_type),
stat = "identity") +
scale_y_discrete(limits = rev) +
scale_x_continuous(limits = c(-16, 16),
breaks = seq(-16, 16, 4),
labels = c(16, 12, 8, 4, 0, 4, 8, 12, 16)) +
scale_fill_manual(labels=c('Points Lost by Conceding After Leading', 'Points Earned by Winning from Behind'),
values = c("#1f3b5e", "#a53924")) +
labs(x = "Points",
y = "",
fill = "",
title = "How Did Teams Deal with In-Game Adversity?: Points Earned/Lost By Teams ",
subtitle = "2022 NWSL Regular Season",
caption = "Source: nwslR") +
theme_minimal() +
theme(legend.position = "bottom")