wnblr  

Buyer Coder beware: As of Dec 2021, this package is undergoing major re-development. The next version will introduce many improvements but also breaking changes.

The wnblr package contains a range of game statistics from the Australian Women’s National Basketball League (WNBL):

• box_scores: Team-level box scores, with all your usual basketball box score metrics like points scored, field goals made / attempted, free throws made / attempted, rebounds, turnovers, and the like.
• box_scores_detailed: Player-level box scores.
• pbp: Play-by-play data.
• shots: Shots data, primarily containing XY shot locations.

Installation

Install the development version of wnblr from GitHub with:

# install.packages("remotes")
remotes::install_github("jacquietran/wnblr")

Examples

There’s lots to work with - for example, here’s a quick look at the shots data:

# Load libraries
library(wnblr)
library(dplyr)

# Frequency of shot types attempted in the 2020 season
shots %>%
  filter(season == 2020) %>%
  select(sub_type, shot_result) %>%
  group_by(sub_type) %>%
  summarise(n = n()) %>%
  ungroup() %>%
  arrange(desc(n))
#> # A tibble: 10 x 2
#>    sub_type               n
#>    <chr>              <int>
#>  1 jumpshot            3381
#>  2 layup               2003
#>  3 drivinglayup         779
#>  4 pullupjumpshot       648
#>  5 turnaroundjumpshot   414
#>  6 stepbackjumpshot     209
#>  7 hookshot             162
#>  8 floatingjumpshot     141
#>  9 fadeaway              42
#> 10 alleyoop               1

With a little bit of elbow grease, you can move neatly from data to plot, too:

# Load libraries
library(ggplot2)
library(showtext)

# Tidy data
magbegor <- box_scores_detailed %>%
  filter(scoreboard_name == "E. Magbegor") %>%
  select(season, contains("field_goals")) %>%
  group_by(season) %>%
  summarise(
    `Field goals made` = round(mean(field_goals_made), 1),
    `Field goals attempted` = round(mean(field_goals_attempted), 1),
    `Field goal %` = round(mean(field_goals_percentage), 1)) %>%
  ungroup() %>%
  tidyr::pivot_longer(
    cols = c(`Field goals made`,
             `Field goals attempted`,
             `Field goal %`),
    names_to = "key",
    values_to = "value") %>%
  mutate(
    key = factor(
      key, levels = c("Field goals made",
                      "Field goals attempted",
                      "Field goal %")))

# Import Google Fonts
font_add_google(name = "Patua One", family = "patua")
font_add_google(name = "Source Sans Pro", family = "source")

# Build plot
showtext_auto()
# After building the plot, run showtext_auto(FALSE) to close off
ggplot(
  magbegor,
  aes(x = season, y = value, group = key, fill = key)) +
  facet_wrap(~key, nrow = 1, scales = "free_y") +
  geom_bar(stat = "identity") +
  scale_fill_manual(values = c("#FFC53D", "#688d3a", "#0E4735")) +
  labs(
    title = "Ezi Magbegor: WNBL shooting averages from 2017-2020",
    x = "Season", y = NULL,
    caption = "Data source: {wnblr} & WNBL.com.au | Plot: @jacquietran") +
  ggdark::dark_mode() +
  theme(
    legend.position = "none",
    text = element_text(
      colour = "#FFFFFF", family = "source", size = 48),
    plot.title = element_text(
      family = "patua", margin=margin(0,0,15,0)),
    plot.caption = element_text(margin=margin(15,0,0,0)))

Known issues

Game data is missing or wildly incomplete for some games in most of the seasons from 2014/2015 to 2020, as documented here. If anyone has leads on where to find live stats from the games listed as “missing”, feel free to log an issue and let me know!

Hex logos

Courtesy of @PythonUnicornCoder - thank you!!