Distribution of kimarite (winning techniques) by weight and height of rikishi (sumo wrestlers)

library(dplyr)
library(plotly)
options(stringsAsFactors = FALSE)

 
 
 
Three years, 18 tournaments:

tournament <- apply(
    expand.grid(
        2014:2016, # years
        sprintf("%02d", seq(1, 11, 2)) # six tournaments a year
    ),
    1,
    paste, collapse = "."
)
  
m <- matrix(tournament[order(tournament)], ncol = 6, byrow = TRUE)
  
prmatrix(
    m,
    rowlab = rep("", ncol(m)),
    collab = rep("", ncol(m)),
    quote = FALSE
)

                                                
 2014.01 2014.03 2014.05 2014.07 2014.09 2014.11
 2015.01 2015.03 2015.05 2015.07 2015.09 2015.11
 2016.01 2016.03 2016.05 2016.07 2016.09 2016.11

 
 
 
Fetch makuuchi (top division) results from Cervus1983/sumodb repository:

bout <- do.call(
    rbind,
    lapply(
        tournament,
        function(x) read.csv(
            paste0("https://raw.githubusercontent.com/Cervus1983/sumodb/master/CSV/", x, ".results.csv")
        )
    )
)
  
bout

 
 
 
Most frequent kimarite is yorikiri — 31.3% of all bouts (including forfeits — fusen):

bout %>% count(kimarite) %>% arrange(-n)

 
 
 
Banzuke for each tournament:

banzuke <- do.call(
    rbind,
    lapply(
        tournament,
        function(x) read.csv(
            paste0("https://raw.githubusercontent.com/Cervus1983/sumodb/master/CSV/", x, ".banzuke.csv")
        )
    )
)
  
banzuke

 
 
 
Height & weight distribution:

banzuke %>%
    group_by(rikishi) %>%
    summarise(height = mean(height)) %>%
    plot_ly(width = 910) %>%
    add_histogram(
        x = ~height,
        xbins = list(
            start = floor(min(banzuke$height)),
            end = ceiling(max(banzuke$height)),
            size = 1
        )
    ) %>%
    layout(
        xaxis = list(title = "Height (cm)"),
        yaxis = list(
            showgrid = FALSE,
            showticklabels = FALSE
        )
    )

banzuke %>%
    group_by(rikishi) %>%
    summarise(weight = mean(weight)) %>%
    plot_ly(width = 910) %>%
    add_histogram(
        x = ~weight,
        xbins = list(
            start = floor(min(banzuke$weight)),
            end = ceiling(max(banzuke$weight)),
            size = 5
        )
    ) %>%
    layout(
        xaxis = list(title = "Weight (kg)"),
        yaxis = list(
            showgrid = FALSE,
            showticklabels = FALSE
        )
    )

 
 
 
Add height & weight data to bout results, group less frequent kimarite as other, and plot kimarite distribution:

bout2 <- left_join(
    left_join(
        bout %>% mutate(rikishi = ifelse(win1 == 1, shikona1, shikona2)),
        banzuke
    ) %>%
        select(-rank) %>%
        rename(winner.height = height, winner.weight = weight) %>%
        mutate(rikishi = ifelse(win1 == 0, shikona1, shikona2)),
    banzuke
) %>%
    select(-rank) %>%
    rename(loser.height = height, loser.weight = weight)
  
bout3 <- inner_join(
    bout2,
    bout2 %>%
        count(kimarite) %>%
        arrange(-n) %>%
        mutate(
            freq_rank = row_number(),
            kimarite2 = ifelse(freq_rank > 7, "other", kimarite) # top 7 kimarite, everything else as "other"
        ) %>%
        select(kimarite, kimarite2)
) %>% filter(
        complete.cases(.), # omit records with missing weight/height
        kimarite != "fusen" # omit forfeits
    )
  
bout3 %>%
    plot_ly(width = 910) %>%
    add_markers(
        x = ~winner.height,
        y = ~loser.height,
        color = ~kimarite2,
        hoverinfo = "text",
        text = ~paste(basho, "~", shikona1, "v", shikona2, "~", kimarite)
    ) %>%
    layout(
        xaxis = list(
            title = "Winner's height"
        ),
        yaxis = list(
            title = "Loser's height"
        )
    )

bout3 %>%
    plot_ly(width = 910) %>%
    add_markers(
        x = ~winner.weight,
        y = ~loser.weight,
        color = ~kimarite2,
        hoverinfo = "text",
        text = ~paste(basho, "~", shikona1, "v", shikona2, "~", kimarite)
    ) %>%
    layout(
        xaxis = list(
            title = "Winner's weight"
        ),
        yaxis = list(
            title = "Loser's weight"
        )
    )