1.3 What is the timeline of hosting the world cup?

What was missing from the previous representation of the data is the time component. In this section we’ll explore a visualization method that would allow us to add this crucial aspect.

Let’s have a look on the data.

gt::gt(df_host1)

host_year	country_name	continent	country_code
1930	Uruguay	South America	UY
1934	Italy	Europe	IT
1938	France	Europe	FR
1950	Brazil	South America	BR
1954	Switzerland	Europe	CH
1958	Sweden	Europe	SE
1962	Chile	South America	CL
1966	England	Europe	NA
1970	Mexico	North America	MX
1974	Germany	Europe	DE
1978	Argentina	South America	AR
1982	Spain	Europe	ES
1986	Mexico	North America	MX
1990	Italy	Europe	IT
1994	United States	North America	US
1998	France	Europe	FR
2002	Japan, South Korea	Asia	JP, KR
2006	Germany	Europe	DE
2010	South Africa	Africa	ZA
2014	Brazil	South America	BR
2018	Russia	Europe	RU
2022	Qatar	Asia	QA
2026	Canada, Mexico, United States	North America	CA, MX, US

We’ll start by preparing the hosting data for visualization by filling in the gap years where the world cup stopped due to WWII

#add missing years in which world cup was cancelled
df_tm <- df_host1 %>% 
  #add rows with missing years in host_year column
  complete(host_year = full_seq(host_year, 4)) %>% 
  #fill the missing data in the other columns
  mutate(across(where(is.character),
                ~ ifelse(is.na(.x), "Cancelled", .x)
                )
         )

We’ll use a chronologically ordered tiles plot (AKA waffle plot) to visualize the timeline of hosting the world cup. Each tile would represent a world cup host. Thus, we need to create a 25-tiles plot (the number of 4-years intervals that represent hosted and cancelled world cup) and order hosts chronologically from left to right, and top to bottom.

#Define the position of each tile chronologically
df_tm <- df_tm %>%
  arrange(host_year) %>% 
  mutate(x = rep(1:5, 5), #number of columns (left to right order)
         y = rep(5:1, each = 5) #number of rows (top to bottom order)
         )
gt::gt(df_tm)

host_year	country_name	continent	country_code	x	y
1930	Uruguay	South America	UY	1	5
1934	Italy	Europe	IT	2	5
1938	France	Europe	FR	3	5
1942	Cancelled	Cancelled	Cancelled	4	5
1946	Cancelled	Cancelled	Cancelled	5	5
1950	Brazil	South America	BR	1	4
1954	Switzerland	Europe	CH	2	4
1958	Sweden	Europe	SE	3	4
1962	Chile	South America	CL	4	4
1966	England	Europe	NA	5	4
1970	Mexico	North America	MX	1	3
1974	Germany	Europe	DE	2	3
1978	Argentina	South America	AR	3	3
1982	Spain	Europe	ES	4	3
1986	Mexico	North America	MX	5	3
1990	Italy	Europe	IT	1	2
1994	United States	North America	US	2	2
1998	France	Europe	FR	3	2
2002	Japan, South Korea	Asia	JP, KR	4	2
2006	Germany	Europe	DE	5	2
2010	South Africa	Africa	ZA	1	1
2014	Brazil	South America	BR	2	1
2018	Russia	Europe	RU	3	1
2022	Qatar	Asia	QA	4	1
2026	Canada, Mexico, United States	North America	CA, MX, US	5	1

Let’s have a first look at the time line

df_tm %>% 
  ggplot(aes(x, y, fill = continent ))+
  geom_tile(color = "black", size = 1)+
  #add country names within each tile
  geom_text(aes(label = country_name),
                           size = 4,
                           show.legend = FALSE)+
  #fill by continent
  scale_fill_manual(values = conti_cols)

## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.

Nice! As you can see hosts are ordered so that the first world cup is in the left-top corner and the 2026 cup is in the right-bottom corner.

One can easily notice that for years where the world cup was cohosted by more than one country, the text doesn’t fit within the tile. We’ll fix this issue in two steps. First, we can split cohosts in different lines so that they are stacked on top of each other in each tile.

#split cohosts into separate lines
df_tm <- df_tm %>%
  mutate(#add a new line between cohosts
         country_name = str_replace_all(country_name, ", ", "\n"),
         #order countries by appearance in the timeline
         continent = factor(continent, levels = unique(continent)))

Second, we’ll use geom_fit_text instead of geom_text to confine the text to the area of each tile. Let’s have another look at the plot after modifying the text.

(plot <- df_tm %>% 
  ggplot()+
  geom_tile(aes(x, y, fill = continent ),
            color = "black", size = 1)+
  #fit country names within each tile
  ggfittext::geom_fit_text(aes(x, y, fill = continent,label = country_name), size = 13,show.legend = FALSE)+
  #fill by continent
  scale_fill_manual(values = conti_cols))

No doubt that countries’ names are more readable now! We can safely remove the plots axes since they don’t actually add anything and move the legend to the top.

(plot <- plot +
    theme_nothing()+
    theme(legend.position = "top"))

Despite the fact that tiles are ordered by year, what’s missing is the year of each cup. How can we add this to the plot? One way would be to indicate the range of years each to the left of each row.

df_range <- df_tm %>% 
  group_by(y) %>% 
  #range of years (first tile - last tile)
  mutate(range = glue::glue("({min(host_year)}-{max(host_year)})")) %>%
  ungroup() %>%
  distinct(y, range)

Let’s add the ranges to the plot and have yet another look

(plot <- plot+
  #add the time interval of each row
  geom_text(data = df_range,
            aes(y = y, label = range),
            x = -0.1,
            size = 4)+
  #expand the plotting area
  scale_x_discrete(expand = expansion(add = c(1.3, 0.2)))
)

This worked well! We can now easily know the year of hosting of each world cup. Finally, let’s beautify the plot by adding borders and some text for context.

plot+
  labs(title = "History of hosting FIFA world cup",
       subtitle = "Host countries of world cups chronologically orderd and colored by continent",
       caption = caption_cdc)+
  coord_fixed(0.7)+
  theme(title = element_text(size = 10),
        panel.border = element_rect(linewidth = 2,
                                    linetype = "solid",
                                    color = "black",
                                    fill = NA))

YES!! This is a condensed and clear representation of our data, which are two desirable features -when in balance- in data visualization!