NBA Play in Games - Circular Barplots of Simple Stats
15/4/2022
12-minute read
Today is April 15th, 2022 and the last round of the ’21-’22 NBA play-in games. Let’s look at some simple statistics of the four teams still fighting for a spot in the playoffs: Atlanta Hawks, Cleveland Cavaliers, New Orleans Pelicans, and the Los Angeles Clippers.
Load Packages
To make these plots I will use three packages:
rvest : to gather the data from Basketball Reference
tidyverse : to clean the data
NBAcharts : to create the charts.
Hopefully the first two are familiar to you. The last one I created to “save time and space”. Be warned that at this time this is a very simple package that currently has zero tests. I hope to update this soon.
# install.packages("devtools")
devtools::install_github("rbolt13/NBAcharts")
# load packages
library(rvest)
library(tidyverse)
library(NBAcharts)Scrape Data
To begin I’m creating a function, nba_data() that does four things:
Uses the rvest package to read two tables (Totals and Per Game Stat) from Basketball Reference.
Rename the second column in both tables to “Name”.
Merge the two tables together.
Replace NA values with 0 (for quantile function).
Then I am creating four data sets for the four teams still competing for a spot in the finals.
# Function : Scrape and merge total and per game data-------
nba_data <- function(slug){
# define team page URL
url <- paste0("https://www.basketball-reference.com/teams/",
slug,"/2022.html")
# Read total stats
ttl_stat <- url %>%
read_html %>%
html_node("#totals") %>%
html_table()
# Read Per Game stats
per_game_stat <- url %>%
read_html %>%
html_node("#per_game") %>%
html_table()
# Rename Column 2 to Name
names(ttl_stat)[2] <- "Name"
names(per_game_stat)[2] <- "Name"
# Replace NA values with 0 (for quantile function)
ttl_stat[is.na(ttl_stat)] <- 0
per_game_stat[is.na(per_game_stat)] <- 0
# Merge stats tables
merge(ttl_stat, per_game_stat,
by = c("Age", "Name", "G", "GS"))
}
# Data Sets : Total and Per Game Team Stats-------------------
hawks_data <- nba_data("ATL")
cavs_data <- nba_data("CLE")
pelicans_data <- nba_data("NOP")
clippers_data <- nba_data("LAC")Tidy Data Sets
There is a lot of simple stats to look at, but to narrow it down lets look at some big ones: steals, blocks, defensive rebounds, assists, turnovers, and points per game.
# Steals----------------------------------------------
team_stl <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(STL.x >= unname(quantile(t_data$STL.x))[4])
#create data frame
stl <- data.frame(
player = filtered_data$Name,
data = filtered_data$STL.x)
# arrange data
stl %>% arrange(-data)
}
# Blocks-----------------------------------------------
team_blk <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(BLK.x >= unname(quantile(t_data$BLK.x))[4])
#create data frame
blk <- data.frame(
player = filtered_data$Name,
data = filtered_data$BLK.x)
# arrange data
blk %>% arrange(-data)
}
# Defensive Rebounds------------------------------------
team_drb <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(DRB.x >= unname(quantile(t_data$DRB.x))[4])
#create data frame
drb <- data.frame(
player = filtered_data$Name,
data = filtered_data$DRB.x)
# arrange data
drb %>% arrange(-data)
}
# Assists -------------------------------------------------
team_ast <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(`AST.x` >= unname(quantile(t_data$`AST.x`))[4])
#create data frame
ast <- data.frame(
player = filtered_data$Name,
data = filtered_data$`AST.x`)
# arrange data
ast %>% arrange(-data)
}
# Turnovers ----------------------------------------------
team_tov <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(`TOV.x` >= unname(quantile(t_data$`TOV.x`))[4])
#create data frame
tov <- data.frame(
player = filtered_data$Name,
data = filtered_data$`TOV.x`)
# arrange data
tov %>% arrange(-data)
}
# Points per game --------------------------------------
team_ppg <- function(team_data){
# reassign data
t_data <- team_data
# filter data
filtered_data <- team_data %>%
filter(`PTS/G` >= unname(quantile(t_data$`PTS/G`))[4])
#create data frame
ppg <- data.frame(
player = filtered_data$Name,
data = filtered_data$`PTS/G`)
# arrange data
ppg %>% arrange(-data)
}Note: Within each of these function I am getting a value for the top 75% of
players in a certain area. Below is a smaller chunk of code that breaks down this line of code. You can easily change this by setting a specific numeric value (i.e. DRB.x >= 200) or by looking at a different quantile by changing the 4 to another value between 1 and 5.
# Quantile (0%, 25%, 50%, 75%, 100%)
quantile(hawks_data$STL.x)## 0% 25% 50% 75% 100%
## 0.00 2.00 14.50 35.25 93.00# Unname
unname(quantile(hawks_data$STL.x))## [1] 0.00 2.00 14.50 35.25 93.00# Extract 4th data value (75%)
unname(quantile(hawks_data$STL.x))[4]## [1] 35.25Next to create tidy data sets for each team for each of the 6 stats.
# Steals Data -------------------------------
hawks_stls <- team_stl(hawks_data)
cavs_stls <- team_stl(cavs_data)
pelicans_stls <- team_stl(pelicans_data)
clippers_stls <- team_stl(clippers_data)
# Blocks Data -------------------------------
hawks_blks <- team_blk(hawks_data)
cavs_blks <- team_blk(cavs_data)
pelicans_blks <- team_blk(pelicans_data)
clippers_blks <- team_blk(clippers_data)
# Defensive Rebounds Data -------------------
hawks_drb <- team_drb(hawks_data)
cavs_drb <- team_drb(cavs_data)
pelicans_drb <- team_drb(pelicans_data)
clippers_drb <- team_drb(clippers_data)
# Points per game ---------------------------
hawks_ppg <- team_ppg(hawks_data)
cavs_ppg <- team_ppg(cavs_data)
pelicans_ppg <- team_ppg(pelicans_data)
clippers_ppg <- team_ppg(clippers_data)
# Assists Data -------------------------------------------
hawks_ast <- team_ast(hawks_data)
cavs_ast <- team_ast(cavs_data)
pelicans_ast <- team_ast(pelicans_data)
clippers_ast <- team_ast(clippers_data)
# Turnovers Data -----------------------------------------
hawks_tov <- team_tov(hawks_data)
cavs_tov <- team_tov(cavs_data)
pelicans_tov <- team_tov(pelicans_data)
clippers_tov <- team_tov(clippers_data)Color Palletes
To create color pallets for each team, I first assigned team_cols variables for each team with HTML links from the website Team Color Codes. Note that since the background of the graphs will be white, and the text black, I omitted these choices (black and white) upon assignment for aesthetics. Then I am using colorRampPalette() to create specific color palettes with size equal to length(team_stat$player).
# Team Colors ------------------------------------------------------------------
hawks_cols <- c("#C1D32F", "#E03A3E", "#26282A")
cavs_cols <- c("#860038", "#041E42", "#FDBB30")
pelicans_cols <- c("#0C2340", "#C8102E", "#85714D")
clippers_cols <- c("#C8102E", "#1D428A", "#BEC0C2")
# Steals color palette ---------------------------------------------------------
hawks_stls_pal <- colorRampPalette(hawks_cols)(length(hawks_stls$player))
cavs_stls_pal <- colorRampPalette(cavs_cols)(length(cavs_stls$player))
pelicans_stls_pal <- colorRampPalette(pelicans_cols)(length(pelicans_stls$player))
clippers_stls_pal <- colorRampPalette(clippers_cols)(length(clippers_stls$player))
# Blocks color palette ---------------------------------------------------------
hawks_blks_pal <- colorRampPalette(hawks_cols)(length(hawks_blks$player))
cavs_blks_pal <- colorRampPalette(cavs_cols)(length(cavs_blks$player))
pelicans_blks_pal <- colorRampPalette(pelicans_cols)(length(pelicans_blks$player))
clippers_blks_pal <- colorRampPalette(clippers_cols)(length(clippers_blks$player))
# Defensive Rebounds color palette ---------------------------------------------
hawks_drb_pal <- colorRampPalette(hawks_cols)(length(hawks_drb$player))
cavs_drb_pal <- colorRampPalette(cavs_cols)(length(cavs_drb$player))
pelicans_drb_pal <- colorRampPalette(pelicans_cols)(length(pelicans_drb$player))
clippers_drb_pal <- colorRampPalette(clippers_cols)(length(clippers_drb$player))
# Points per game (ppg) color palette ------------------------------------------
hawks_ppg_pal <- colorRampPalette(hawks_cols)(length(hawks_ppg$player))
cavs_ppg_pal <- colorRampPalette(cavs_cols)(length(cavs_ppg$player))
pelicans_ppg_pal <- colorRampPalette(pelicans_cols)(length(pelicans_ppg$player))
clippers_ppg_pal <- colorRampPalette(clippers_cols)(length(clippers_ppg$player))
# Assists (ast) ----------------------------------------------------------------
hawks_ast_pal <- colorRampPalette(hawks_cols)(length(hawks_ast$player))
cavs_ast_pal <- colorRampPalette(cavs_cols)(length(cavs_ast$player))
pelicans_ast_pal <- colorRampPalette(pelicans_cols)(length(pelicans_ast$player))
clippers_ast_pal <- colorRampPalette(clippers_cols)(length(clippers_ast$player))
# Turnovers (tov) --------------------------------------------------------------
hawks_tov_pal <- colorRampPalette(hawks_cols)(length(hawks_tov$player))
cavs_tov_pal <- colorRampPalette(cavs_cols)(length(cavs_tov$player))
pelicans_tov_pal <- colorRampPalette(pelicans_cols)(length(pelicans_tov$player))
clippers_tov_pal <- colorRampPalette(clippers_cols)(length(clippers_tov$player))Graphs
Since 6 different stats for four different teams equals 24 graphs, I created the package NBAcharts to minimize the amount of code needed to recreate each graph. For more information on the function I use be sure to check out it’s documentation by typing ?circ_col_chart() into your RStudio console, or checking out the Github repository here.
# Variables (team names) -------------------------------------------------------
hawks <- "Atlanta Hakws"
cavs <- "Clevland Cavaliers"
pels <- "New Orleans Pelicans"
clips <- "Los Angeles Clippers"
# Variables (captions) ---------------------------------------------------------
all_caption <- c("\n\n Data Visualisation by Randi Bolt\n https://www.rbolt.me/\nSource: Basketball Reference \nLink: https://www.basketball-reference.com/\nDate: April 15th, 2022")
steal_caption <- c("This visualization shows the total number of steals\n by the top 25% of stealers of the `21/`22 ")
blocks_caption <- c("This visualization shows the total number of blocks\n by the top 25% of blockers of the `21/`22 ")
drb_caption <- c("This visualization shows the total number of defensive\n rebounds by the top 25% of defensive rebounders of the \n`21/`22 ")
ppg_caption <- c("This visualization shows the average amount of points\n per game by the top 25% of scorers of the `21/`22 ")
ast_caption <- c("This visualization shows the total number of assists\n by the top 25% of assisters of the `21/`22 ")
tov_caption <- c("This visualization shows the total number of turn overs\n by the top 25% of turn overers of the `21/`22 ")
# Steals -----------------------------------------------------------------------
# Hawks
hawks_stls_png <- circ_col_chart(player_data = hawks_stls,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = hawks_stls_pal,
chart_title = c("Hawks Total Steals"),
chart_subtitle = paste(steal_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_stls.png"))
# Cavs
cavs_stls_png <- circ_col_chart(player_data = cavs_stls,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = cavs_stls_pal,
chart_title = c("Cavaliers Total Steals"),
chart_subtitle = paste(steal_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_stls.png"))
# Pelicans
pelicans_stls_png <- circ_col_chart(player_data = pelicans_stls,
ring_scale = 50,
ring_inner = 55,
ring_middle = 105,
ring_outer = 155,
ringlab_inner = "50",
ringlab_middle = "100",
ringlab_outer = "150",
col_pal = pelicans_stls_pal,
chart_title = c("Pelicans Total Steals"),
chart_subtitle = paste(steal_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_stls.png"))
# Clippers
clippers_stls_png <- circ_col_chart(player_data = clippers_stls,
ring_scale = 50,
ring_inner = 55,
ring_middle = 105,
ring_outer = 155,
ringlab_inner = "50",
ringlab_middle = "100",
ringlab_outer = "150",
col_pal = clippers_stls_pal,
chart_title = c("Clippers Total Steals"),
chart_subtitle = paste(steal_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_stls.png"))
# Blocks -----------------------------------------------------------------------
# Hawks
hawks_blks_png <- circ_col_chart(player_data = hawks_blks,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = hawks_blks_pal,
chart_title = c("Hawks Total Blocks"),
chart_subtitle = paste(blocks_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_blks.png"))
# Cavs
cavs_blks_png <- circ_col_chart(player_data = cavs_blks,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = cavs_blks_pal,
chart_title = c("Cavaliers Total Blocks"),
chart_subtitle = paste(blocks_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_blks.png"))
# Pelicans
pelicans_blks_png <- circ_col_chart(player_data = pelicans_blks,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = pelicans_blks_pal,
chart_title = c("Pelicans Total Blocks"),
chart_subtitle = paste(blocks_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_blks.png"))
# Clippers
clippers_blks_png <- circ_col_chart(player_data = clippers_blks,
ring_scale = 30,
ring_inner = 35,
ring_middle = 65,
ring_outer = 95,
ringlab_inner = "30",
ringlab_middle = "60",
ringlab_outer = "90",
col_pal = clippers_blks_pal,
chart_title = c("Clippers Total Blocks"),
chart_subtitle = paste(blocks_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_blks.png"))
# Defensive Rebounds -----------------------------------------------------------
# Hawks
hawks_drb_png <- circ_col_chart(player_data = hawks_drb,
ring_scale = 200,
ring_inner = 220,
ring_middle = 420,
ring_outer = 620,
ringlab_inner = "200",
ringlab_middle = "400",
ringlab_outer = "600",
col_pal = hawks_drb_pal,
chart_title = c("Hawks Total Defensive Rebounds"),
chart_subtitle = paste(drb_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_drb.png"))
# Cavs
cavs_drb_png <- circ_col_chart(player_data = cavs_drb,
ring_scale = 200,
ring_inner = 220,
ring_middle = 420,
ring_outer = 620,
ringlab_inner = "200",
ringlab_middle = "400",
ringlab_outer = "600",
col_pal = cavs_drb_pal,
chart_title = c("Cavaliers Total Defensive Rebounds"),
chart_subtitle = paste(drb_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_drb.png"))
# Pelicans
pelicans_drb_png <- circ_col_chart(player_data = pelicans_drb,
ring_scale = 200,
ring_inner = 220,
ring_middle = 420,
ring_outer = 620,
ringlab_inner = "200",
ringlab_middle = "400",
ringlab_outer = "600",
col_pal = pelicans_drb_pal,
chart_title = c("Pelicans Total Defensive Rebounds"),
chart_subtitle = paste(drb_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_drb.png"))
# Clippers
clippers_drb_png <- circ_col_chart(player_data = clippers_drb,
ring_scale = 200,
ring_inner = 220,
ring_middle = 420,
ring_outer = 620,
ringlab_inner = "200",
ringlab_middle = "400",
ringlab_outer = "600",
col_pal = clippers_drb_pal,
chart_title = c("Clippers Total Defensive Rebounds"),
chart_subtitle = paste(drb_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_drb.png"))
# Points per game (ppg) --------------------------------------------------------
# Hawks
hawks_ppg_png <- circ_col_chart(player_data = hawks_ppg,
ring_scale = 10,
ring_inner = 12,
ring_middle = 22,
ring_outer = 32,
ringlab_inner = "10",
ringlab_middle = "20",
ringlab_outer = "30",
col_pal = hawks_ppg_pal,
chart_title = c("Hawks Points per Game"),
chart_subtitle = paste(ppg_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_ppg.png"))
# Cavs
cavs_ppg_png <- circ_col_chart(player_data = cavs_ppg,
ring_scale = 10,
ring_inner = 12,
ring_middle = 22,
ring_outer = 32,
ringlab_inner = "10",
ringlab_middle = "20",
ringlab_outer = "30",
col_pal = cavs_ppg_pal,
chart_title = c("Cavaliers Points per Game"),
chart_subtitle = paste(ppg_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_ppg.png"))
# Pelicans
pelicans_ppg_png <- circ_col_chart(player_data = pelicans_ppg,
ring_scale = 10,
ring_inner = 12,
ring_middle = 22,
ring_outer = 32,
ringlab_inner = "10",
ringlab_middle = "20",
ringlab_outer = "30",
col_pal = pelicans_ppg_pal,
chart_title = c("Pelicans Points per Game"),
chart_subtitle = paste(ppg_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_ppg.png"))
# Clippers
clippers_ppg_png <- circ_col_chart(player_data = clippers_ppg,
ring_scale = 10,
ring_inner = 12,
ring_middle = 22,
ring_outer = 32,
ringlab_inner = "10",
ringlab_middle = "20",
ringlab_outer = "30",
col_pal = clippers_ppg_pal,
chart_title = c("Clippers Points per Game"),
chart_subtitle = paste(ppg_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_ppg.png"))
# Assists (ast) ----------------------------------------------------------------
# Hawks
hawks_ast_png <- circ_col_chart(player_data = hawks_ast,
ring_scale = 250,
ring_inner = 260,
ring_middle = 510,
ring_outer = 760,
ringlab_inner = "250",
ringlab_middle = "500",
ringlab_outer = "750",
col_pal = hawks_ast_pal,
chart_title = c("Hawks Total Assists"),
chart_subtitle = paste(ast_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_ast.png"))
# Cavs
cavs_ast_png <- circ_col_chart(player_data = cavs_ast,
ring_scale = 250,
ring_inner = 260,
ring_middle = 510,
ring_outer = 760,
ringlab_inner = "250",
ringlab_middle = "500",
ringlab_outer = "750",
col_pal = cavs_ast_pal,
chart_title = c("Cavaliers Total Assists"),
chart_subtitle = paste(ast_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_ast.png"))
# Pelicans
pelicans_ast_png <- circ_col_chart(player_data = pelicans_ast,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = pelicans_ast_pal,
chart_title = c("Pelicans Total Assists"),
chart_subtitle = paste(ast_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_ast.png"))
# Clippers
clippers_ast_png <- circ_col_chart(player_data = clippers_ast,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = clippers_ast_pal,
chart_title = c("Clippers Total Assists"),
chart_subtitle = paste(ast_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_ast.png"))
# Turnovers (tov) --------------------------------------------------------------
# Hawks
hawks_tov_png <- circ_col_chart(player_data = hawks_tov,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = hawks_tov_pal,
chart_title = c("Hawks Total Turnovers"),
chart_subtitle = paste(tov_caption, hawks),
chart_source = all_caption,
cc_chart_filename = c("hawks_tov.png"))
# Cavs
cavs_tov_png <- circ_col_chart(player_data = cavs_tov,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = cavs_tov_pal,
chart_title = c("Cavaliers Total Turnovers"),
chart_subtitle = paste(tov_caption, cavs),
chart_source = all_caption,
cc_chart_filename = c("cavs_tov.png"))
# Pelicans
pelicans_tov_png <- circ_col_chart(player_data = pelicans_tov,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = pelicans_tov_pal,
chart_title = c("Pelicans Total Turnovers"),
chart_subtitle = paste(tov_caption, pels),
chart_source = all_caption,
cc_chart_filename = c("pelicans_tov.png"))
# Clippers
clippers_tov_png <- circ_col_chart(player_data = clippers_tov,
ring_scale = 100,
ring_inner = 110,
ring_middle = 210,
ring_outer = 310,
ringlab_inner = "100",
ringlab_middle = "200",
ringlab_outer = "300",
col_pal = clippers_tov_pal,
chart_title = c("Clippers Total Turnovers"),
chart_subtitle = paste(tov_caption, clips),
chart_source = all_caption,
cc_chart_filename = c("clippers_tov.png"))To display the images I am going to put the teams who played each other side by side using settings within the r chunk.
Below are the charts for the Clippers and the Pelicans.
Resources
insert side by side png images using knitr
Additional Resources
Hawks at Cavaliers | Full Game Highlights | April 15, 2022
Pelicans at Clippers | Full Game Highlights | April 15, 2022
LA Clippers’ Paul George tests positive for COVID-19, out for play-in game vs. New Orleans Pelicans
CJ McCollum gets 90-second standing ovation in Portland return | it was special to be part of it