150: Jaccard

Load data

library(rlang)

all_games <- readr::read_csv("all_games_jaccard.csv")

# h2h function using collegefootballdata
# get_all_team_opponents_with_h2h <- function(team_name) {
#  games_2025 <- get_games(year = 2025) |>
#    dplyr::filter(
#      seasonType == "regular",
#      homeClassification == "fbs", 
#      awayClassification == "fbs"
#    )
  
#  opponents <- dplyr::bind_rows(
#    games_2025 |> dplyr::filter(homeTeam == team_name) |> dplyr::select(opponent = awayTeam),
#    games_2025 |> dplyr::filter(awayTeam == team_name) |> dplyr::select(opponent = homeTeam)
#  ) |>
#    dplyr::pull(opponent) |>
#    unique()
  
#  return(opponents)  # Don't add the team itself
# }

# load Texas, Alabama, Penn State, Ohio State opponents via CSV 

psu_opp <- readr::read_csv("psu_opp.csv")
osu_opp <- readr::read_csv("osu_opp.csv")

texas_opp <- readr::read_csv("texas_opp.csv")
bama_opp <- readr::read_csv("bama_opp.csv")


# save table for data wrapper sharing 
# datawrapper_tbl <- all_games |> 
#  dplyr::filter(year == 2025) |> 
#  dplyr::mutate(
#    most_sim_final = dplyr::case_when(
      # All three have the same Jaccard score
#      most_sim_jaccard == second_sim_jaccard & second_sim_jaccard == third_sim_jaccard ~ 
#        paste0(most_similar, ", ", second_similar, ", ", third_similar),
      
      # First two have the same Jaccard score (but third is different)
#      most_sim_jaccard == second_sim_jaccard ~ 
#        paste0(most_similar, ", ", second_similar),
      
      # Only the first one is unique (default case)
#      TRUE ~ most_similar
#    )
#  ) |>
#  dplyr::select(team, most_sim_final, shared_fbs = most_sim_shared, jaccard = most_sim_jaccard)

Beeswarm plot

beeswarm_data <- all_games |>
  dplyr::filter(year %in% c(1995, 2005, 2015, 2025)) |>
  dplyr::select(year, most_sim_jaccard) |>
  dplyr::filter(!is.na(most_sim_jaccard))

# Convert to percentage version
summary_stats <- beeswarm_data |>
  dplyr::group_by(year) |>
  dplyr::summarise(
    median_jaccard = median(most_sim_jaccard, na.rm = TRUE),
    mean_jaccard = mean(most_sim_jaccard, na.rm = TRUE),
    n_teams = dplyr::n(),
    .groups = "drop"
  ) |>
  dplyr::mutate(
    # Create annotation labels with percentages
    annotation_label = paste0(
      n_teams,
      " teams \n",
      round(median_jaccard * 100, 1),
      "%"
    )
  )

beeswarm_plot <- ggplot2::ggplot(
  beeswarm_data,
  ggplot2::aes(
    x = factor(year),
    y = most_sim_jaccard * 100,
    color = factor(year)
  )
) +
  ggbeeswarm::geom_quasirandom(alpha = 0.7, 
                               size = 1.6, 
                               width = 0.3,
                               method = "tukey"
  ) +
  ggplot2::annotate(
    "label",
    x = c(1, 2, 3, 4),
    y = 10.5,
    label = summary_stats$annotation_label,
    size = 3.5,
    color = "#333333",
    fill = "floralwhite",
    family = "Roboto Condensed"
  ) +
  ggplot2::annotate(
    "label",
    x = 1.95,
    y = 93.5,
    label = "Jaccard similarity example  \nUTEP and BYU shared 10 opponents in 1995  \n83.3% of unique opponents between  \n the two teams were shared",
    size = 3.5,
    color = "#333333",
    fill = "#42049EFF",
    alpha = 0.3,
    family = "Roboto Condensed"
  ) +
  ggplot2::annotate(
    geom = "curve",
    color = "#474747",
    x = 1.5,
    y = 90,
    xend = 1.15,
    yend = 83.3,
    curvature = .3,
    arrow = ggplot2::arrow(length = grid::unit(2, "mm"))
  ) + 
  ggplot2::scale_color_viridis_d(
    name = "Year",
    option = "plasma",
    begin = 0.1,
    end = 0.9
  ) +
  ggplot2::labs(
    title = "Similarity in college football schedules over time",
    subtitle = "Distribution of highest schedule similarities between FBS teams (regular season only).",
    x = "Year",
    y = "Schedule Overlap (%)",
    caption = "Shows number of FBS teams and median similarity for a sample of four seasons.  \ndata via collegefootballdata.com | Viz by Chris at Bless your Chart"
  ) +
  hrbrthemes::theme_ipsum_rc() +
  ggplot2::theme(
    panel.grid.minor = ggplot2::element_blank(),
    legend.position = "none",
    plot.subtitle = ggtext::element_markdown(),
    plot.caption = ggtext::element_markdown(color = "gray60",
                              family = "Roboto Condensed",
                              hjust = 1),
    plot.margin = ggplot2::margin(t = 5, r = 20, b = 5, l = 5, unit = "pt")
  ) +
  ggplot2::ylim(0, 100)


ggplot2::ggsave(
  "beeswarm_plot.png",
  beeswarm_plot,
  w =10,
  h = 8,
  dpi = 600,
  type = 'cairo',
  bg = "white"
)

beeswarm_plot

Venn diagram

opponent_lists <- list(
  "Penn State" = psu_opp[[2]],
  "Ohio State" = osu_opp[[2]]  
)

venn_plot <- ggvenn::ggvenn(
    opponent_lists,
    fill_color = c("#041E42", "#BB0000"),  # Penn State navy and Ohio State scarlet
    fill_alpha = 0.3,
    stroke_size = 1.5,
    set_name_size = 0,
    show_elements = FALSE,   
    text_color = "#333333",
    text_size = 4
) +
    hrbrthemes::theme_ipsum_rc(base_family = "Roboto Condensed") +
    ggplot2::labs(
        title = "Penn State and Ohio State: 2025 Shared Opponents",
        subtitle = "While both teams play head-to-head, the two Big Ten schools share only two common opponents.",
        x = "",
        y = "",
        caption = "Out of 20 unique opponents, only 10% (two total) are shared  \ndata via collegefootballdata.com | Viz by Chris at Bless your Chart"
    ) +
    ggplot2::theme(
        text = ggplot2::element_text(family = "Roboto Condensed"),
        plot.subtitle = ggtext::element_markdown(),
        plot.caption = ggtext::element_markdown(color = "gray60",
                                                family = "Roboto Condensed",
                                                hjust = 1),
        plot.margin = ggplot2::margin(t = 5, r = 20, b = 5, l = 5, unit = "pt"),
        
        # Remove ALL axis elements including numbers
        axis.title = ggplot2::element_blank(),
        axis.text = ggplot2::element_blank(),
        axis.text.x = ggplot2::element_blank(), 
        axis.text.y = ggplot2::element_blank(),  
        axis.ticks = ggplot2::element_blank(),
        axis.line = ggplot2::element_blank(),
        
        # Remove panel elements
        panel.grid.major = ggplot2::element_blank(),
        panel.grid.minor = ggplot2::element_blank(),
        panel.border = ggplot2::element_blank(),
        panel.background = ggplot2::element_blank(),
        
        # Remove plot background
        plot.background = ggplot2::element_blank(),
        
        # Remove legend elements (if you want)
        legend.background = ggplot2::element_blank(),
        legend.key = ggplot2::element_blank(),
        
        # Remove strip elements (for facets)
        strip.background = ggplot2::element_blank(),
        strip.text = ggplot2::element_blank()
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0,
        y =-0.30,
        team = "UCLA",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0,
        y =-0.50,
        team = "Rutgers",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = .01,
        y = 0.25,
        team = "Penn State",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.01,
        y = 0.45,
        team = "Ohio State",
        height = .085,
    ) + # show penn st opponents 
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.80,
        y = 0.55,
        team = "Oregon",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -1.2,
        y = 0.55,
        team = "Indiana",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.48,
        y = 0.55,
        team = "Michigan State",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.80,
        y = 0.35,
        team = "Nebraska",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -1.2,
        y = 0.35,
        team = "Northwestern",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.48,
        y = 0.35,
        team = "Iowa",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.8,
        y = -0.30,
        team = "Florida International",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -0.5,
        y = -0.30,
        team = "Nevada",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = -1.2,
        y = -0.30,
        team = "Villanova",
        height = .085,
    ) + # ohio state opponents 
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 1.15,
        y = -0.30,
        team = "Texas",
        height = .090,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.80,
        y = -0.30,
        team = "Ohio",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.45,
        y = -0.30,
        team = "Grambling",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.45,
        y = 0.55,
        team = "Washington",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.80,
        y = 0.55,
        team = "Michigan",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 1.15,
        y = 0.55,
        team = "Illinois",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.45,
        y = 0.35,
        team = "Minnesota",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 0.80,
        y = 0.35,
        team = "Wisconsin",
        height = .085,
    ) +
    ggplot2::annotate(
        cfbplotR::GeomCFBlogo,
        x = 1.15,
        y = 0.35,
        team = "Purdue",
        height = .085,
    ) +
    # Add team name annotations with Roboto Condensed font
    ggplot2::annotate(
        "text",
        x = -0.8,
        y = 0.8,
        label = "Penn State",
        family = "Roboto Condensed",
        size = 5,
        fontface = "bold",
        color = "#041E42"
    ) +
    ggplot2::annotate(
        "text", 
        x = 0.8,
        y = 0.8,
        label = "Ohio State",
        family = "Roboto Condensed", 
        size = 5,
        fontface = "bold",
        color = "#BB0000"
    )

ggplot2::ggsave(
  "vd_plot.png",
  venn_plot,
  w =10,
  h = 8,
  dpi = 600,
  type = 'cairo',
  bg = "white"
)

venn_plot