In beanumber/teamcolors: Color Palettes for Pro Sports Teams

teamcolors

An R package providing color palettes for pro and amateur sports teams. The palettes are provided by Jim Neilsen's Team Colors website and offered with only minimal alterations. NCAA colors come from teamcolorcodes.com, via the ncaahoopR package. Other sports include the Women's National Basketball Association, National Women's Soccer League, and the Canadian Football League.

Install

To install the CRAN version, use:

install.packages(teamcolors)

To install the development version from GitHub, use:

devtools::install_github("beanumber/teamcolors")

Load

library(teamcolors)
head(teamcolors)

Show palettes

Sometimes you need to work with a named vector of colors. Other times you can use the built-in scale_color_teams() and scale_color_fill() functions.

league_pal("nba")

Plot

In baseball, Pythagorean expectation relates expected winning percentage to runs allowed and runs scored. How well does it work?

library(Lahman)
library(tidyverse)
pythag <- Teams %>%
  filter(yearID == 2016) %>%
  select(name, teamID, yearID, W, L, R, RA) %>%
  mutate(
    wpct = W / (W + L), exp_wpct = 1 / (1 + (RA / R)^2),
    # note name discrepancy!
    name = ifelse(name == "Los Angeles Angels of Anaheim", "Los Angeles Angels", name)
  )

ggplot2

ggplot(pythag, aes(x = wpct, y = exp_wpct, color = name, fill = name)) +
  geom_abline(slope = 1, intercept = 0, linetype = 3) +
  geom_point(shape = 21, size = 3) +
  scale_fill_teams(guide = FALSE) +
  scale_color_teams(2, guide = FALSE) +
  ggrepel::geom_text_repel(aes(label = teamID)) +
  scale_x_continuous("Winning Percentage", limits = c(0.3, 0.7)) +
  scale_y_continuous("Expected Winning Percentage", limits = c(0.3, 0.7)) +
  theme_light() +
  labs(
    title = "Real and Pythagorean winning % by team",
    subtitle = paste(first(pull(pythag, yearID)), "MLB Season", sep = " "),
    caption = "Source: Lahman baseball database. Using teamcolors R pkg"
  ) +
  coord_equal()

Base R

pythag <- pythag %>%
  left_join(teamcolors, by = "name")
with(pythag, plot(wpct, exp_wpct, bg = primary, col = secondary, pch = 21, cex = 3))

Key

You can see the color palettes using existing functionality from the scales package, but it won't show the names of the teams.

scales::show_col(league_pal("mlb"), borders = league_pal("mlb", 2))

So, instead, use show_team_col(). Note that this only shows color palettes for non-NCAA teams.

show_team_col()

To view color palettes for college teams, use the show_ncaa_col() function.

show_ncaa_col()

Logos

Links to team logos are provided by (http://www.sportslogos.net/).

teamcolors %>%
  filter(grepl("New Y", name)) %>%
  pull(logo) %>%
  knitr::include_graphics()

Note that we don't have any coverage for the EPL.

teamcolors %>%
  group_by(league) %>%
  summarize(
    num_teams = n(),
    num_logos = sum(!is.na(logo))
  )

References

For more examples see:

• Lopez, M.J., Matthews, G.J., Baumer, B.S., "How often does the best team win? A unified approach to understanding randomness in North American sport," The Annals of Applied Statistics, vol. 12, no. 4, 2018, pp. 2483--2516. URL (https://doi.org/10.1214/18-AOAS1165)

To cite this package in your work, see:

citation("teamcolors")

Notes

beanumber/teamcolors documentation built on Feb. 13, 2023, 10:16 a.m.