Goal

Collect data to understand the circumstances of what happened in Chelsea’s Graham Potter era.

Actual analysis will be done on other medium later.

Prep

library(rvest)
library(tidyverse)
library(janitor)
library(lubridate)
library(knitr)

Transfermarkt

Get HTML with rvest.

tmarkt <- read_html("https://www.transfermarkt.com/graham-potter/leistungsdatenDetail/trainer/23954/plus/1?saison_id=&verein_id=631&liga=&wettbewerb_id=&trainer_id=")
tmarkt_df <- tmarkt %>% html_elements(".items") %>% html_table() %>% as.data.frame()
tmarkt_df %>% kable()
Date Competition Season MD Home.Team Home.Team.1 wappen Result wappen.1 Away.Team Away.Team.1 Tactic
Apr 1, 2023 Premier League 22/23 29 Chelsea NA 0:2 NA Aston Villa 3-4-3 NA NA
Mar 18, 2023 Premier League 22/23 28 Chelsea NA 2:2 NA Everton 3-4-3 NA NA
Mar 11, 2023 Premier League 22/23 27 Leicester NA 1:3 NA Chelsea 3-4-3 NA NA
Mar 7, 2023 UEFA Champions League 22/23 last 16 Chelsea NA 2:0 NA Bor. Dortmund 3-4-3 NA NA
Mar 4, 2023 Premier League 22/23 26 Chelsea NA 1:0 NA Leeds 3-4-3 NA NA
Feb 26, 2023 Premier League 22/23 25 Tottenham NA 2:0 NA Chelsea 4-2-3-1 NA NA
Feb 18, 2023 Premier League 22/23 24 Chelsea NA 0:1 NA Southampton 4-2-3-1 NA NA
Feb 15, 2023 UEFA Champions League 22/23 last 16 Bor. Dortmund NA 1:0 NA Chelsea 4-2-3-1 NA NA
Feb 11, 2023 Premier League 22/23 23 West Ham NA 1:1 NA Chelsea 4-2-3-1 NA NA
Feb 3, 2023 Premier League 22/23 22 Chelsea NA 0:0 NA Fulham 4-3-3 Attacking NA NA
Jan 21, 2023 Premier League 22/23 21 Liverpool NA 0:0 NA Chelsea 4-1-4-1 NA NA
Jan 15, 2023 Premier League 22/23 20 Chelsea NA 1:0 NA Crystal Palace 4-2-3-1 NA NA
Jan 12, 2023 Premier League 22/23 7 Fulham NA 2:1 NA Chelsea 3-5-2 flat NA NA
Jan 8, 2023 FA Cup 22/23 Third Round Man City NA 4:0 NA Chelsea 4-3-3 Attacking NA NA
Jan 5, 2023 Premier League 22/23 19 Chelsea NA 0:1 NA Man City 4-2-3-1 NA NA
Jan 1, 2023 Premier League 22/23 18 Nottm Forest NA 1:1 NA Chelsea 4-3-3 Attacking NA NA
Dec 27, 2022 Premier League 22/23 17 Chelsea NA 2:0 NA Bournemouth 4-3-3 Attacking NA NA
Nov 12, 2022 Premier League 22/23 16 Newcastle NA 1:0 NA Chelsea 3-4-2-1 NA NA
Nov 9, 2022 EFL Cup 22/23 Third Round Man City NA 2:0 NA Chelsea 3-4-3 NA NA
Nov 6, 2022 Premier League 22/23 15 Chelsea NA 0:1 NA Arsenal 4-2-3-1 NA NA
Nov 2, 2022 UEFA Champions League 22/23 Group Stage Chelsea NA 2:1 NA Dinamo Zagreb 4-3-1-2 NA NA
Oct 29, 2022 Premier League 22/23 14 Brighton NA 4:1 NA Chelsea 3-4-2-1 NA NA
Oct 25, 2022 UEFA Champions League 22/23 Group Stage RB Salzburg NA 1:2 NA Chelsea 3-4-1-2 NA NA
Oct 22, 2022 Premier League 22/23 13 Chelsea NA 1:1 NA Man Utd 3-4-2-1 NA NA
Oct 19, 2022 Premier League 22/23 12 Brentford NA 0:0 NA Chelsea 3-4-2-1 NA NA
Oct 16, 2022 Premier League 22/23 11 Aston Villa NA 0:2 NA Chelsea 3-4-2-1 NA NA
Oct 11, 2022 UEFA Champions League 22/23 Group Stage AC Milan NA 0:2 NA Chelsea 3-5-2 NA NA
Oct 8, 2022 Premier League 22/23 10 Chelsea NA 3:0 NA Wolves 4-2-3-1 NA NA
Oct 5, 2022 UEFA Champions League 22/23 Group Stage Chelsea NA 3:0 NA AC Milan 3-4-2-1 NA NA
Oct 1, 2022 Premier League 22/23 9 Crystal Palace NA 1:2 NA Chelsea 4-3-1-2 NA NA
Sep 14, 2022 UEFA Champions League 22/23 Group Stage Chelsea NA 1:1 NA RB Salzburg 4-2-3-1 NA NA

Drop empty columns.

tmarkt_df <- tmarkt_df %>% 
  select(-6, -8, -11, -12)

Rename column headers.

tmarkt_df <- tmarkt_df %>% 
  clean_names() %>% 
  rename(
    score = wappen,
    away_team = wappen_1,
    tactic = away_team
  )

Tidy the table. Add venue identifier where Chelsea played their game.

tmarkt_df <- tmarkt_df %>% 
  mutate(
    venue = case_when(
      home_team == "Chelsea" ~ "home",
      away_team == "Chelsea" ~ "away"
    )
  )

Split the score into two columns.

tmarkt_df <- tmarkt_df %>% 
  separate(score, into = c("home_score", "away_score"), sep = ":")

Add goal_for and goal_against for chelsea.

tmarkt_df <- tmarkt_df %>% 
  mutate(
    goal_for = case_when(
      home_team == "Chelsea" ~ home_score,
      away_team == "Chelsea" ~ away_score
    ),
    goal_against = case_when(
      home_team != "Chelsea" ~ home_score,
      away_team != "Chelsea" ~ away_score,
    )
  )

Convert ‘date’ column to date class.

tmarkt_df$date <- mdy(tmarkt_df$date)
tmarkt_df["date"]
##          date
## 1  2023-04-01
## 2  2023-03-18
## 3  2023-03-11
## 4  2023-03-07
## 5  2023-03-04
## 6  2023-02-26
## 7  2023-02-18
## 8  2023-02-15
## 9  2023-02-11
## 10 2023-02-03
## 11 2023-01-21
## 12 2023-01-15
## 13 2023-01-12
## 14 2023-01-08
## 15 2023-01-05
## 16 2023-01-01
## 17 2022-12-27
## 18 2022-11-12
## 19 2022-11-09
## 20 2022-11-06
## 21 2022-11-02
## 22 2022-10-29
## 23 2022-10-25
## 24 2022-10-22
## 25 2022-10-19
## 26 2022-10-16
## 27 2022-10-11
## 28 2022-10-08
## 29 2022-10-05
## 30 2022-10-01
## 31 2022-09-14
tmarkt_df %>% head() %>% kable()
date competition season md home_team home_score away_score away_team tactic venue goal_for goal_against
2023-04-01 Premier League 22/23 29 Chelsea 0 2 Aston Villa 3-4-3 home 0 2
2023-03-18 Premier League 22/23 28 Chelsea 2 2 Everton 3-4-3 home 2 2
2023-03-11 Premier League 22/23 27 Leicester 1 3 Chelsea 3-4-3 away 3 1
2023-03-07 UEFA Champions League 22/23 last 16 Chelsea 2 0 Bor. Dortmund 3-4-3 home 2 0
2023-03-04 Premier League 22/23 26 Chelsea 1 0 Leeds 3-4-3 home 1 0
2023-02-26 Premier League 22/23 25 Tottenham 2 0 Chelsea 4-2-3-1 away 0 2

Export to csv

tmarkt_df %>% 
  write.csv(
    file = "tmarkt_df"
  )