Evergreen games

First, load some R packages I will need later:

library(dplyr) 
library(purrr)
library(rvest)
library(httr)
library(stringr)
library(glue)

The following is a function which goes to the helpful page https://boardgamegeek.com/plays/bygame/subtype/boardgame/start/2018-01-01/end/2018-12-31?sortby=distinctusers for a given year, retrieves all the IDs of the games and the number of unqiue players, and then makes a call to the XML API to get the publication years.

get_players <- function(year) {
  #first create the link with the appropriate year substituted in and read that page in
  players_link <- glue('https://boardgamegeek.com/plays/bygame/subtype/boardgame/start/{year}-01-01/end/{year}-12-31?sortby=distinctusers', year = year)
  
  players_page <- read_html(players_link) 
  
  
  # strip out all the game IDs (used https://selectorgadget.com/ to get the matching path)
  ids <- players_page %>%
    html_nodes('#main_content td:nth-child(3) a') %>%
    html_attr('href') %>%
    str_replace('/playstats/thing/','')
  
  # same for the number of unique players
  players <- players_page %>%
    html_nodes('#main_content td:nth-child(3) a') %>%
    html_text() %>%
    str_replace('\n\t\t','') %>%
    as.integer()
  
  # now get the XML API page for the IDs pulled out - it can do them all in one call
  ids_xml <- GET('https://boardgamegeek.com',
                       accept_xml(),
                       path = '/xmlapi2/thing',
                       query = list('id' = paste(ids, collapse = ',')))
  
  # from the XML page returned pull out the publication year
  pubyear <- content(ids_xml) %>%
    xml_find_all('//item/yearpublished') %>%
    xml_attr('value') %>%
    as.integer()
  
  # and the title
  title <- content(ids_xml) %>%
    xml_find_all('//item/name[@type="primary"]') %>%
    xml_attr('value')
  
  # and now stick everything in a data frame
  data_frame(title, players, pubyear)
}

With this function created, I can now just repeat it for all the years I’m interested in, combining them into a single data frame:

players_df <- map_df(2005:2018, get_players, .id = 'year') %>%
  mutate(year = as.integer(year) + 2004)

And finally I do a bit of manipulation to get only the ‘evergreen games’ and display them in the desired format.

players_df %>%
  filter(year - pubyear >= 10) %>%
  arrange(desc(year), desc(players)) %>%
  group_by(year) %>%
  top_n(10, players) %>%
  ungroup()