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.
To install the CRAN version, use:
install.packages(teamcolors)
To install the development version from GitHub, use:
devtools::install_github("beanumber/teamcolors")
library(teamcolors)
head(teamcolors)
## # A tibble: 6 × 11
## name league primary secon…¹ terti…² quate…³ divis…⁴ locat…⁵ mascot sport…⁶
## <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
## 1 A&M-Cor… ncaa #0067C5 #007F3E <NA> <NA> Southl… A&M-Co… Chris… A&M-Co…
## 2 AFC Bou… epl #e62333 #000000 <NA> <NA> <NA> AFC Bourn… AFC Bo…
## 3 Abilene… ncaa #461D7C #FFFFFF <NA> <NA> WAC Abilene Chris… Abilen…
## 4 Air For… ncaa #003087 #8A8D8F #B1B3B3 #FFC72C MWC Air Fo… <NA> Air Fo…
## 5 Akron ncaa #041E42 #A89968 <NA> <NA> MAC Akron <NA> Akron
## 6 Alabama ncaa #9E1B32 #828A8F #FFFFFF <NA> SEC Alabama <NA> Alabama
## # … with 1 more variable: logo <chr>, and abbreviated variable names
## # ¹secondary, ²tertiary, ³quaternary, ⁴division, ⁵location, ⁶sportslogos_name
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")
## Atlanta Hawks Boston Celtics Brooklyn Nets
## "#e13a3e" "#008348" "#061922"
## Charlotte Hornets Chicago Bulls Cleveland Cavaliers
## "#1d1160" "#ce1141" "#860038"
## Dallas Mavericks Denver Nuggets Detroit Pistons
## "#007dc5" "#4d90cd" "#ed174c"
## Golden State Warriors Houston Rockets Indiana Pacers
## "#fdb927" "#ce1141" "#ffc633"
## Los Angeles Clippers Los Angeles Lakers Memphis Grizzlies
## "#ed174c" "#fdb927" "#0f586c"
## Miami Heat Milwaukee Bucks Minnesota Timberwolves
## "#98002e" "#00471b" "#005083"
## New Orleans Pelicans New York Knicks Oklahoma City Thunder
## "#002b5c" "#006bb6" "#007dc3"
## Orlando Magic Philadelphia 76ers Phoenix Suns
## "#007dc5" "#ed174c" "#e56020"
## Portland Trail Blazers Sacramento Kings San Antonio Spurs
## "#e03a3e" "#724c9f" "#bac3c9"
## Toronto Raptors Utah Jazz Washington Wizards
## "#ce1141" "#002b5c" "#002b5c"
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)
)
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()
## Warning: The `guide` argument in `scale_*()` cannot be `FALSE`. This was deprecated in
## ggplot2 3.3.4.
## ℹ Please use "none" instead.
## Warning: ggrepel: 4 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
pythag <- pythag %>%
left_join(teamcolors, by = "name")
with(pythag, plot(wpct, exp_wpct, bg = primary, col = secondary, pch = 21, cex = 3))
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()
## Warning: The `<scale>` argument of `guides()` cannot be `FALSE`. Use "none" instead as
## of ggplot2 3.3.4.
## ℹ The deprecated feature was likely used in the teamcolors package.
## Please report the issue at <�]8;;https://github.com/beanumber/teamcolors/issues�https://github.com/beanumber/teamcolors/issues�]8;;�>.
To view color palettes for college teams, use the show_ncaa_col()
function.
show_ncaa_col()
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))
)
## # A tibble: 10 × 3
## league num_teams num_logos
## <chr> <int> <int>
## 1 cfl 9 8
## 2 epl 20 0
## 3 mlb 30 30
## 4 mls 22 21
## 5 nba 30 30
## 6 ncaa 363 235
## 7 nfl 32 31
## 8 nhl 31 31
## 9 nwsl 9 7
## 10 wnba 12 12
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")