topChef is a collection of data sets detailing events across all seasons of Top Chef US and Top Chef Masters US and one season of Top Chef Canada. It includes Chef information, challenge descriptions, challenge winners, episode information, guest judge names, and reward/prize information.
install.packages("topChef")I am continually improving the datasets, so github version is likely to be more up-to-date. If it’s not appearing to be updated, restart your R sessions, install it again, and call it into your library.
devtools::install_github("celevitz/topChef")Data were collected manually while watching each season of Top Chef. Additional data were collected from https://en.wikipedia.org/wiki/Top_Chef. My Top Chef data journey was inspired by https://topchefstats.com/.
Huge thanks to https://github.com/doehm for all his support!
Across datasets, key joining variables include:
chefseasonseasonNumberseriesepisodeA tibble containing information about Chefs for each season they are in, including placement and gender. For some but not all seasons, there is also information on hometown, current city of residence, age, a flag for whether they are a person of color, and their occupation.
chefdetails
#> # A tibble: 416 × 13
#> name chef hometown city state age season seasonNumber series placement
#> <chr> <chr> <chr> <chr> <chr> <dbl> <chr> <dbl> <chr> <dbl>
#> 1 Richar… Rich… <NA> <NA> <NA> 38 All S… 8 US 1
#> 2 Mike I… Mike… <NA> <NA> <NA> 35 All S… 8 US 2
#> 3 Antoni… Anto… <NA> <NA> <NA> 34 All S… 8 US 3
#> 4 Tiffan… Tiff… <NA> <NA> <NA> 27 All S… 8 US 4
#> 5 Carla … Carl… <NA> <NA> <NA> 46 All S… 8 US 5
#> 6 Dale T… Dale… <NA> <NA> <NA> 32 All S… 8 US 6
#> 7 Angelo… Ange… <NA> <NA> <NA> 35 All S… 8 US 7
#> 8 Fabio … Fabi… <NA> <NA> <NA> 32 All S… 8 US 8
#> 9 Tre Wi… Tre … <NA> <NA> <NA> 34 All S… 8 US 9
#> 10 Marcel… Marc… <NA> <NA> <NA> 30 All S… 8 US 10
#> # ℹ 406 more rows
#> # ℹ 3 more variables: personOfColor <chr>, occupation <chr>, gender <chr>A tibble containing information about each challenge that the Chefs compete in.
challengedescriptions
#> # A tibble: 641 × 17
#> season seasonNumber series episode challengeType outcomeType
#> <chr> <dbl> <chr> <dbl> <chr> <chr>
#> 1 Kentucky 16 US 1 Quickfire Team
#> 2 Kentucky 16 US 1 Elimination Team
#> 3 Kentucky 16 US 2 Quickfire Individual
#> 4 Kentucky 16 US 2 Elimination Team
#> 5 Kentucky 16 US 3 Quickfire Individual
#> 6 Kentucky 16 US 3 Elimination Individual
#> 7 Kentucky 16 US 4 Quickfire Individual
#> 8 Kentucky 16 US 4 Elimination Team
#> 9 Kentucky 16 US 5 Elimination Team
#> 10 Kentucky 16 US 6 Elimination Individual
#> # ℹ 631 more rows
#> # ℹ 11 more variables: challengeDescription <chr>, shopTime <chr>,
#> # shopBudget <chr>, prepTime <dbl>, cookTime <dbl>, productPlacement <chr>,
#> # advantage <chr>, lastChanceKitchenWinnerEnters <chr>,
#> # restaurantWarWinner <chr>, restaurantWarEliminated <chr>,
#> # didJudgesVisitWinningTeamFirst <chr>A tibble containing win and loss data for each chef in each episode.
challengewins
#> # A tibble: 9,667 × 9
#> season seasonNumber series episode inCompetition chef challengeType outcome
#> <chr> <dbl> <chr> <dbl> <lgl> <chr> <chr> <chr>
#> 1 All St… 8 US 1 TRUE Ange… Quickfire LOW
#> 2 All St… 8 US 1 TRUE Anto… Quickfire WIN
#> 3 All St… 8 US 1 TRUE Carl… Quickfire LOW
#> 4 All St… 8 US 1 TRUE Case… Quickfire HIGH
#> 5 All St… 8 US 1 TRUE Dale… Quickfire HIGH
#> 6 All St… 8 US 1 TRUE Dale… Quickfire WIN
#> 7 All St… 8 US 1 TRUE Elia… Quickfire LOW
#> 8 All St… 8 US 1 TRUE Fabi… Quickfire LOW
#> 9 All St… 8 US 1 TRUE Jami… Quickfire LOW
#> 10 All St… 8 US 1 TRUE Jenn… Quickfire HIGH
#> # ℹ 9,657 more rows
#> # ℹ 1 more variable: rating <dbl>A tibble containing information about who were the guest judges for each challenge.
judges
#> # A tibble: 755 × 9
#> season seasonNumber series episode challengeType outcomeType guestJudge
#> <chr> <dbl> <chr> <dbl> <chr> <chr> <chr>
#> 1 All Stars: … 8 US 1 Quickfire Team Tom Colic…
#> 2 All Stars: … 8 US 1 Elimination Individual Anthony B…
#> 3 All Stars: … 8 US 2 Quickfire Individual Joe Jonas
#> 4 All Stars: … 8 US 2 Elimination Team Katie Lee
#> 5 All Stars: … 8 US 3 Quickfire Team David Cha…
#> 6 All Stars: … 8 US 3 Elimination Team Anthony B…
#> 7 All Stars: … 8 US 3 Elimination Team Kate Krad…
#> 8 All Stars: … 8 US 4 Quickfire Individual Tony Mant…
#> 9 All Stars: … 8 US 4 Elimination Team Tony Mant…
#> 10 All Stars: … 8 US 5 Quickfire Individual Tom Colic…
#> # ℹ 745 more rows
#> # ℹ 2 more variables: competedOnTC <chr>, otherShows <chr>A tibble containing information about rewards and prizes won by challenge.
rewards
#> # A tibble: 328 × 9
#> season seasonNumber series episode challengeType outcomeType rewardType
#> <chr> <dbl> <chr> <dbl> <chr> <chr> <chr>
#> 1 All Stars: … 8 US 1 Elimination Individual Money
#> 2 All Stars: … 8 US 3 Quickfire Team Money
#> 3 All Stars: … 8 US 3 Quickfire Team Money
#> 4 All Stars: … 8 US 3 Quickfire Team Money
#> 5 All Stars: … 8 US 3 Quickfire Team Money
#> 6 All Stars: … 8 US 3 Elimination Team Prize
#> 7 All Stars: … 8 US 4 Quickfire Individual Money
#> 8 All Stars: … 8 US 4 Elimination Team Prize
#> 9 All Stars: … 8 US 5 Quickfire Individual Prize
#> 10 All Stars: … 8 US 6 Elimination Team Prize
#> # ℹ 318 more rows
#> # ℹ 2 more variables: reward <chr>, chef <chr>A tibble containing information about each episode.
episodeinfo
#> # A tibble: 358 × 8
#> season seasonNumber series overallEpisodeNumber episode episodeName
#> <chr> <dbl> <chr> <dbl> <dbl> <chr>
#> 1 Canada 6 6 Canada 60 1 The Next Wave
#> 2 Canada 6 6 Canada 61 2 The Chicken o…
#> 3 Canada 6 6 Canada 62 3 Que Rico
#> 4 Canada 6 6 Canada 63 4 Restaurant Wa…
#> 5 Canada 6 6 Canada 64 5 Double Overti…
#> 6 Canada 6 6 Canada 65 6 That's a Lot …
#> 7 Canada 6 6 Canada 66 7 Wine of the T…
#> 8 Canada 6 6 Canada 67 8 Finale Four W…
#> 9 San Francisco 1 US 1 1 Who Deserves …
#> 10 San Francisco 1 US 2 2 Food of Love
#> # ℹ 348 more rows
#> # ℹ 2 more variables: airDate <date>, nCompetitors <dbl>
library(ggplot2)
chefdetails %>%
# Keep just winners and relevant variables (just the winning chef's name)
filter(placement == 1) %>%
select(chef, gender,series,season) %>%
# Bring on the challenge result data
left_join(challengewins) %>%
# Keep just elimination & sudden death quickfire challenges
filter(challengeType %in% c("Elimination","Quickfire Elimination"
,"Sudden Death Quickfire")) %>%
# Get the number of wins for each person
group_by(series,season,seasonNumber,chef,gender) %>%
filter(outcome %in% c("WIN","WINNER")) %>%
summarise(wins=n()) %>%
# Plot
ggplot(aes(x=wins,y=chef,color=gender,fill=gender,shape=series)) +
geom_point(size=4) +
theme_minimal() +
# beautification
scale_color_manual(values=c("#1170AA", "#A3ACB9")) +
scale_shape_manual(values=c(0,17,1)) +
scale_x_continuous(name="Number of elimination wins",
breaks=seq(1,9,1),
limits=c(.8,9.5),
labels=seq(1,9,1)) +
theme(panel.grid = element_blank(),
title=element_text(size=12),
axis.line.x=element_line(color="black"),
axis.ticks.x=element_line(color="black"),
axis.text.x=element_text(size=12),
axis.title.x=element_text(size=12)) +
labs(title="Number of elimination wins of Top Chefs"
,caption="I should have ordered by # of wins...")
library(ggplot2)
library(topChef)
## Won the episode
# Data set up
# how many episodes each chef won in each season
wonepi <- challengewins %>% select(!rating) %>%
mutate(challengeType=case_when(challengeType %in% c("Quickfire Elimination","Sudden Death Quickfire") ~ "Quickfire"
,TRUE ~ challengeType)) %>%
filter(outcome %in% c("WIN","WINNER") & challengeType %in% c("Elimination","Quickfire")) %>%
distinct() %>%
pivot_wider(names_from=challengeType,values_from=outcome) %>%
filter(Quickfire == "WIN" & (Elimination %in% c("WIN","WINNER"))) %>%
# were there multiple winners
group_by(series,season,seasonNumber, episode) %>%
mutate(nchefs=n()
,chef=ifelse(nchefs > 1,paste0(chef,"*"),chef)) %>%
# how many times did each chef win an episode?
ungroup() %>% group_by(series,season,seasonNumber,chef) %>%
summarise(n=n()) %>%
select(series,season,seasonNumber,chef,n) %>%
distinct() %>%
mutate(seasonNumberchar=case_when(seasonNumber <= 9 ~paste0("0",as.character(seasonNumber))
,TRUE ~as.character(seasonNumber)) ) %>%
filter(series == "US")
# visualize
wonepivizdata <- wonepi %>%
# add on placement of chefs, just for those who have won an episode
left_join(topChef::chefdetails %>% select(season,seasonNumber,chef,placement)) %>%
mutate(placement=as.character(case_when(placement > 5 ~ 5
,TRUE ~ placement))
,placement = case_when(placement == "5" ~ "5th or lower"
,placement == "1" ~ "1st"
,placement == "2" ~ "2nd"
,placement == "3" ~ "3rd"
,placement == "4" ~ "4th"
,TRUE ~ placement))
wonepivizdata <- wonepivizdata[order(wonepivizdata$seasonNumberchar,wonepivizdata$n,wonepivizdata$chef),]
## Visualization
wonepivizdata %>%
ggplot(aes(x=seasonNumberchar,y=n,label=chef,fill=factor(placement))) +
scale_fill_manual(values=c("#141B41","#1170AA","#5fa2ce","#a3cce9","#ababab")) +
geom_col(position="stack",color="white") +
geom_text(size = 1, position = position_stack(vjust=.5),angle=15
,color=case_when(wonepivizdata$placement == "1st"~ "white"
,wonepivizdata$placement == "2nd" ~ "white"
,TRUE ~ "black"
,is.na(wonepivizdata$placement) ~ "black")) +
labs(title="A Chef has won an episode in all but Seasons 1 and 16"
,subtitle="Winning an episode means that one chef won both the Quickfire and Elimination Challenges.\nThis has happened 35 times. Eight Top Chefs won an episode in their season.\n
Brooke is the only one to do it in two different seasons. In all but three seasons, only
one or two chefs won an episode. In seasons four and seven, three chefs won an
episode. In season 11, six chefs won an episode. Stefan and Gregory G. are the only
chefs to have ever won three episodes in one season. Four chefs have won two
episodes in one season: Angelo S., Ed C., Tiffany D., and Brooke W.
* Travis and Carlos won the team quickfire and the paired elimination challenge."
,caption= "Data: github.com/celevitz/topChef /// Twitter@carlylevitz") +
ylab("\n# of times a chef won an episode this season\n") +
xlab("Season") +
theme_minimal() +
theme(panel.grid = element_blank()
,axis.line.x=element_line(color="black")
,axis.ticks.x=element_line(color="black")
,axis.line.y=element_line(color="black")
,axis.ticks.y=element_line(color="black")
,axis.text.x=element_text(color="black")
,axis.text.y=element_text(color="black")
,axis.title.x = element_text(color="black")
,axis.title.y = element_text(color="black") ) +
guides(fill = guide_legend(title = "Placement"))
I created a weighted index to compare chefs within and across seasons. I am still working on it, but this is how things currently stand: The weighted index is calculated by assigning a certain number of points to different outcomes. For the sake of simplicity, Sudden Death Quickfires and Quickfire Elimination Challenges are counted as Elimination Challenges. The scoring is as follows:
The weightedindex function takes the following
parameters:
series: US, US Masters, or Canadaseasonnumber: Values between 1 and 20 for Top Chef US;
1 through 5 for US Masters; and 6 for Canadanumberofelimchalls: Number of elimination challenges
through which you want to calculate the index. Values between 1 and
20.numberofquickfires: Number of quickfire challenges
through which you want to calculate the index. Values between 1 and
20.Currently, the function will include more Quickfire challenges than you have specified in the function. It will keep the episodes in each season through the episode with the Nth Elimination Challenge or the Nth Quickfire, whichever is the higher episode. This was a way to hold constant the number of challenges that have occurred so that chefs could be more comparable across seasons.
The weightedindex function exports a tibble with the
variables of: series, season name, season number, chef, chef’s
placement, number of elimination challenges wins/highs/lows/outs, number
of quickfire challenge wins/highs/lows, and the weighted index
score.
weightedindex function: full
seasons
library(topChef)
library(ggplot2)
library(tidyr)
library(dplyr)
## Get the index for all seasons
allseasons <- weightedindex("US",1,20,20)
for (season in seq(2,20,1)) {
allseasons <- rbind(allseasons,weightedindex("US",season,20,20))
}
## Graph it
# for sorting reasons, have the season be a character
allseasons$seasonnumchar[allseasons$seasonNumber <= 9] <-
paste0("0",as.character(allseasons$seasonNumber[allseasons$seasonNumber <= 9]))
allseasons$seasonnumchar[allseasons$seasonNumber > 9] <-
as.character(allseasons$seasonNumber[allseasons$seasonNumber > 9])
# for sorting reasons, have the placement as a character
allseasons$placementchar[allseasons$placement <= 9] <-
paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))
allseasons$placementchar[allseasons$placement > 9] <-
as.character(allseasons$placement[allseasons$placement > 9])
# Distribution of scores at the end of the competition/seasons
graphalltitles <- function(dataset) {
dataset %>%
ggplot(aes(x=placement,y=indexWeight,label=chef)) +
facet_wrap(~paste0("Season ",seasonnumchar)) +
geom_hline(yintercept=0,color="#ffbc69")+
geom_text(hjust=0.5,size=2) +
theme_minimal() +
labs(title=paste0("Top Chef Weighted Index Scores at the End of Seasons")
,subtitle="Comparing All Chefs Across All Seasons\n")+
ylab("Index Score") + xlab("Placement") +
scale_y_continuous(lim=c(-25,50)) +
scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +
theme(panel.grid = element_blank()
,axis.text.x=element_text(size=6,color="black")
,axis.text.y=element_text(size=6,color="black")
,axis.ticks=element_line(color="gray15")
,axis.line=element_line(color="gray15")
,strip.background=element_rect(fill="darkcyan")
,strip.text=element_text(color="black"))
}
graphnotitlenocaption <- function(dataset) {
dataset %>%
ggplot(aes(x=placement,y=indexWeight,label=chef)) +
facet_wrap(~paste0("Season ",seasonnumchar)) +
geom_hline(yintercept=0,color="#ffbc69")+
geom_text(hjust=0.5,size=2) +
theme_minimal() +
ylab("Index score") + xlab("Placement") +
scale_y_continuous(lim=c(-25,50)) +
scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +
theme(panel.grid = element_blank()
,axis.text.x=element_text(size=6,color="black")
,axis.text.y=element_text(size=6,color="black")
,axis.ticks=element_line(color="gray15")
,axis.line=element_line(color="gray15")
,strip.background=element_rect(fill="darkcyan")
,strip.text=element_text(color="black"))
}
graphonlycaption <- function(dataset) {
dataset %>%
ggplot(aes(x=placement,y=indexWeight,label=chef)) +
facet_wrap(~paste0("Season ",seasonnumchar)) +
geom_hline(yintercept=0,color="#ffbc69")+
geom_text(hjust=0.5,size=2) +
theme_minimal() +
ylab("Index Score") + xlab("Placement") +
scale_y_continuous(lim=c(-25,50)) +
scale_x_continuous(lim=c(0,20),breaks=seq(1,18,2),labels = seq(1,18,2)) +
theme(panel.grid = element_blank()
,axis.text.x=element_text(size=6,color="black")
,axis.text.y=element_text(size=6,color="black")
,axis.ticks=element_line(color="gray15")
,axis.line=element_line(color="gray15")
,strip.background=element_rect(fill="darkcyan")
,strip.text=element_text(color="black")) +
labs(caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")
}
graphalltitles(allseasons %>%
filter(seasonNumber %in% c(1,2,3,4,5,6)) )
graphnotitlenocaption(allseasons %>%
filter(seasonNumber %in% c(7,8,9,10,11,12)) )
graphnotitlenocaption(allseasons %>%
filter(seasonNumber %in% c(13,14,15,16,17,18)) )
graphonlycaption(allseasons %>%
filter(seasonNumber %in% c(19,20)) )
weightedindex function 10 Elimination
Challenges or 7 Quickfires into the seasonThe circles represent outliers: Michael in Las Vegas, Richard in All Stars: New York, Paul in Texas, and Kristen in Seattle. The thick bar within the rectangles are the median scores of chefs scores in that season.
This example shows the index scores for the Top Four chefs in all seasons with the remaining six chefs from Top Chef World All Stars. Ali, Buddha, and Amar thus far have the highest index scores.
library(topChef)
library(ggplot2)
library(tidyr)
library(dplyr)
## Get the index for all seasons
allseasons <- weightedindex("US",1,10,7)
for (season in seq(2,20,1)) {
allseasons <- rbind(allseasons,weightedindex("US",season,10,7))
}
# drop unneeded variables
allseasons <- allseasons[,c("chef","season","seasonNumber","placement","indexWeight")]
## Graph it
# for sorting reasons, have the season be a character
allseasons$seasonnumchar[allseasons$seasonNumber <= 9] <-
paste0("0",as.character(allseasons$seasonNumber[allseasons$seasonNumber <= 9]))
allseasons$seasonnumchar[allseasons$seasonNumber > 9] <-
as.character(allseasons$seasonNumber[allseasons$seasonNumber > 9])
# for sorting reasons, have the placement as a character
allseasons$placementchar[allseasons$placement <= 9] <-
paste0("0",as.character(allseasons$placement[allseasons$placement <= 9]))
allseasons$placementchar[allseasons$placement > 9] <-
as.character(allseasons$placement[allseasons$placement > 9])
# Distribution of scores at this stage of the competition
allseasons %>%
ggplot(aes(x=seasonnumchar,y=indexWeight) ) +
geom_boxplot() +
# add horizontal line at 0
geom_hline(yintercept=0, color="#ffbc69") +
theme_minimal() +
labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")
,subtitle="Comparing All Chefs Across All Seasons\n"
,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+
scale_x_discrete(labels=unique(allseasons$season[order(allseasons$seasonNumber)])) +
theme_minimal() +
ylab("Index score") + xlab("") +
theme(panel.grid = element_blank()
,axis.text.x=element_text(angle=90)
)
# Top Four: Distribution of scores at this stage of the competition
allseasons[allseasons$placement <= 4,] %>%
ggplot(aes(x=placement,y=indexWeight,label=chef) ) +
# add horizontal line at 0
geom_hline(yintercept=0, color="#ffbc69") +
geom_text(size=2) +
theme_minimal() +
labs(title=paste0("Top Chef Weighted Index: 7 quickfires & 10 elimination challenges\ninto each season")
,subtitle="Comparing Top Four Chefs Across All Seasons\n"
,caption="Scoring: Elimination win = 7 points. Elimination high = 3. Elimination low = -3. Eliminated = -7.\nQuickfire win = 4. Quickfire high = 2. Quickfire low = -2.\nData github.com/celevitz/topChef ||| Twitter @carlylevitz")+
scale_x_continuous(lim=c(.5,4.5),breaks=c(1,1.5,2,3,4),labels=c("1","Current\nseason","2","3","4")) +
theme_minimal() +
ylab("Index score") + xlab("Placement") +
theme(panel.grid = element_blank()
)