FIDE Chess Player Ratings

library(tidyverse) # ggplot, lubridate, dplyr, stringr, readr...
library(praise)
library(DT)
library(fontawesome)
fontawesome::fa_html_dependency()

fide_ratings_august_OG <- read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-09-23/fide_ratings_august.csv")
fide_ratings_september_OG <- read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-09-23/fide_ratings_september.csv")

fide_ratings_august <- fide_ratings_august_OG |> 
  #filter(games > 0) |>
  mutate(level = case_when(title %in% c("GM", "WGM") | wtitle == "WGM" ~ "Grand Master",
                           title %in% c("IM", "WIM") | wtitle == "WIM" ~ "International Master",
                           title %in% c("FM", "WFM") | wtitle == "WFM" ~ "Fide Master",
                           title %in% c("CM", "WCM") | wtitle == "WCM" ~ "Candidate Master",
                           rating <= 1400 ~ "Beginner",
                           rating <= 1800 ~ "Intermediate",
                           TRUE ~ "Advanced")) |> 
  mutate(level = factor(level, levels = c("Grand Master", "International Master", 
                           "Fide Master", "Candidate Master", 
                           "Advanced", "Intermediate",
                           "Beginner")))

fide_ratings_august |> 
  filter(fed %in% c("ESP", "GER", "USA")) |>  
  ggplot(aes(x = fed, fill = level)) + 
  geom_bar(position = "dodge")

Joining August & September

fide_ratings <- fide_ratings_august_OG |> 
  inner_join(fide_ratings_september_OG, by = "id") |> 
  mutate(level_aug = case_when(title.x %in% c("GM", "WGM") | wtitle.x == "WGM" ~ "Grand Master",
                           title.x %in% c("IM", "WIM") | wtitle.x == "WIM" ~ "International Master",
                           title.x %in% c("FM", "WFM") | wtitle.x == "WFM" ~ "Fide Master",
                           title.x %in% c("CM", "WCM") | wtitle.x == "WCM" ~ "Candidate Master",
                           rating.x <= 1400 ~ "Beginner",
                           rating.x <= 1800 ~ "Intermediate",
                           TRUE ~ "Advanced")) |> 
  mutate(level_aug = factor(level_aug, levels = c("Grand Master", "International Master", 
                           "Fide Master", "Candidate Master", 
                           "Advanced", "Intermediate",
                           "Beginner"))) |>
  mutate(level_sep = case_when(title.y %in% c("GM", "WGM") | wtitle.y == "WGM" ~ "Grand Master",
                           title.y %in% c("IM", "WIM") | wtitle.y == "WIM" ~ "International Master",
                           title.y %in% c("FM", "WFM") | wtitle.y == "WFM" ~ "Fide Master",
                           title.y %in% c("CM", "WCM") | wtitle.y == "WCM" ~ "Candidate Master",
                           rating.y <= 1400 ~ "Beginner",
                           rating.y <= 1800 ~ "Intermediate",
                           TRUE ~ "Advanced")) |> 
  mutate(level_sep = factor(level_sep, levels = c("Grand Master", "International Master", 
                           "Fide Master", "Candidate Master", 
                           "Advanced", "Intermediate",
                           "Beginner"))) |> 
  mutate(rating_diff = rating.y - rating.x)

Below are all the people who did not play any games in September or August, yet their rating across the two months changed. No idea why!

fide_ratings |> 
  filter(games.x == 0, games.y == 0, rating_diff != 0) |> 
  ggplot(aes(x = level_sep, y = rating_diff)) +
  geom_point(aes(color = fed.x)) + 
  ggrepel::geom_text_repel(aes(label = name.x), size = 2) + 
  labs(y = "", 
       x = "",
       color = "country",
       title = "September - August rating", 
       subtitle = "players with zero games and rating change") + 
  scale_color_brewer(palette = "Dark2") + 
  theme_minimal()

Maybe the 35 people who changed (but didn’t play any games), actually did play some games, which weren’t recorded. Seems likely to be a recording error.

fide_ratings |> 
  filter(games.x == 0, games.y == 0) |> 
  mutate(rate0 = ifelse(rating_diff == 0 , "no change", "change")) |>  
  group_by(rate0) |> 
  summarize(n())

# A tibble: 2 × 2
  rate0      `n()`
  <chr>      <int>
1 change        35
2 no change 129924

fide_ratings |> 
  slice_max(order_by = abs(rating_diff), n = 25) |> 
  select(name.x, rating.x, games.x, rating.y, games.y, title.x, title.y, rating_diff)

# A tibble: 26 × 8
   name.x          rating.x games.x rating.y games.y title.x title.y rating_diff
   <chr>              <dbl>   <dbl>    <dbl>   <dbl> <chr>   <chr>         <dbl>
 1 Kulshrestha Ta…     1573       0     2242       9 <NA>    <NA>            669
 2 Montes Molina,…     2202      17     1573       0 <NA>    <NA>           -629
 3 Gil Jimenez, D…     1970       4     1409       0 <NA>    <NA>           -561
 4 Guru Darshan P      1409       5     1958       9 <NA>    <NA>            549
 5 Ayushman Biyala     1492       2     1901       0 <NA>    <NA>            409
 6 Dea Ruiz, Jofre     1901       9     1492       0 <NA>    <NA>           -409
 7 Mussauly, Nura…     2115      15     1716       0 <NA>    <NA>           -399
 8 Plotnikov, Ale…     1564       0     1960       8 <NA>    <NA>            396
 9 Schell, Finn        1956       7     1564       0 <NA>    <NA>           -392
10 Plzak, David        1488       0     1850      18 <NA>    <NA>            362
# ℹ 16 more rows

praise()

[1] "You are unreal!"