College Football Visualizations with cfbfastR and sdvplotR

On this page

Ten charts and tables from the 2025 college football season: all 136 FBS teams on one scatter, conference small multiples, conference logos on an axis, a conference standings table, the national champion’s season with a logo in the title, the College Football Playoff bracket, a rivalry across two decades, a quarterback leaderboard with headshots, a ranked bar chart with logos on the y axis and a tier list. The last section makes the first chart interactive.

College football is the league where sdvplotR’s name matching earns its keep: the data names schools ("Ohio State", "San José State"), and sdvplotR resolves every FBS and FCS program, plus each conference, from those names, so the code below never builds a lookup table.

Setup

The play-by-play comes from the SportsDataverse releases through cfbfastR and needs no key. The schedule, ratings and series history come from the CollegeFootballData.com API, which needs a free key. Get one at https://collegefootballdata.com/key and register it before running the chunks that call cfbd_*() or load_cfb_schedules():

Sys.setenv(CFBD_API_KEY = "YOUR-API-KEY-HERE")
# to keep it across sessions, add CFBD_API_KEY=YOUR-API-KEY-HERE to
# ~/.Renviron (usethis::edit_r_environ()); see ?cfbfastR::register_cfbd

library(sdvplotR)
library(ggplot2)
library(dplyr, warn.conflicts = FALSE)
library(gt)

season <- 2025
caption <- "Data: cfbfastR | 2025 season"

pbp <- cfbfastR::load_cfb_pbp(season)

theme_set(
  theme_minimal(base_size = 12) +
    theme(
      plot.title = element_text(face = "bold"),
      plot.title.position = "plot",
      plot.caption = element_text(color = "grey45", size = 8)
    )
)

The schedule comes from one CFBD call:

schedule <- cfbfastR::load_cfb_schedules(season)

The output below uses a snapshot of that call taken on November 09, 2026 from cfbfastR::load_cfb_schedules(2025) (CFBD refused: the sportsdataverse-data cfb_schedules release), because CollegeFootballData.com caps a key’s calls each month and the site build cannot rely on one; data-raw/article_fixtures.R takes it with the same code. The ratings and the series history are live CFBD calls; without a key, or once its monthly calls are spent, the sections that need them show their code but do not run it.

# a key with calls left this month (a margin for the articles built after this one): CFBD answers HTTP 429 with no rows once they are spent
has_key <- isTRUE(tryCatch(suppressMessages(cfbfastR::cfbd_api_key_info())$remaining_calls > 10, error = function(e) FALSE))
if (!has_key) cat("CFBD_API_KEY is not set or has no calls left, so the SP+, SRS and series chunks are shown but not run.\n")

The schedule carries each team’s division and its conference that season. sdv_resolve() turns the school names into team ids:

fbs <- bind_rows(
  select(schedule, team = home_team, conference = home_conference, division = home_division),
  select(schedule, team = away_team, conference = away_conference, division = away_division)
) |>
  filter(division == "fbs") |>
  distinct(team, conference) |>
  mutate(
    # the index keys three conferences by their short codes
    conference = recode(conference, "American Athletic" = "AAC", "Conference USA" = "CUSA", "FBS Independents" = "FBS Indep."),
    # "Charlotte" alone is the 49ers or the Saints; the 49ers are the FBS team
    team_id = sdv_resolve(sub("^Charlotte$", "Charlotte 49ers", team), league = "cfb")
  )

nrow(fbs)
#> [1] 136
count(fbs, conference, sort = TRUE)
#>       conference  n
#> 1        Big Ten 18
#> 2            ACC 17
#> 3         Big 12 16
#> 4            SEC 16
#> 5            AAC 14
#> 6       Sun Belt 14
#> 7   Mid-American 13
#> 8           CUSA 12
#> 9  Mountain West 12
#> 10    FBS Indep.  2
#> 11        Pac-12  2

Conferences are rows of the team index too (sdv_teams("cfb", include_conferences = TRUE)), which holds each school’s conference today. Checking the season’s conferences against it shows who has moved. A tutorial on a past season should take conferences from that season’s data:

conf_rows <- sdv_teams("cfb", include_conferences = TRUE) |>
  filter(program == "conference")

fbs |>
  inner_join(select(sdv_teams("cfb"), team_id, conf_today = conference_id), by = "team_id") |>
  mutate(
    conf_today = conf_rows$team_id[match(conf_today, conf_rows$conference_id)],
    conf_2025 = coalesce(
      conf_rows$team_id[match(tolower(conference), tolower(conf_rows$team_id))],
      conf_rows$team_id[match(tolower(conference), tolower(conf_rows$short_name))]
    )
  ) |>
  filter(!is.na(conf_2025), conf_2025 != conf_today) |>
  arrange(conf_today, team) |>
  select(team, conf_2025, conf_today)
#>                team     conf_2025    conf_today
#> 1 Northern Illinois           MAC Mountain West
#> 2              UTEP          CUSA Mountain West
#> 3       Boise State Mountain West        Pac-12
#> 4    Colorado State Mountain West        Pac-12
#> 5      Fresno State Mountain West        Pac-12
#> 6   San Diego State Mountain West        Pac-12
#> 7       Texas State      Sun Belt        Pac-12
#> 8        Utah State Mountain West        Pac-12
#> 9    Louisiana Tech          CUSA      Sun Belt

1. All 136 FBS teams on one chart

Offensive EPA per play against defensive EPA per play allowed, for every FBS team, over the full season. With this many teams the logos have to be small: width = 0.032 makes each one 3.2% of the panel’s width. geom_mean_lines() marks the FBS averages, and the defense axis is reversed so the best teams sit top right.

scrimmage <- pbp |>
  filter(pass == 1 | rush == 1, !is.na(EPA))

offense <- scrimmage |>
  group_by(team = pos_team) |>
  summarise(off_epa = mean(EPA), off_sr = mean(epa_success), .groups = "drop")
defense <- scrimmage |>
  group_by(team = def_pos_team) |>
  summarise(def_epa = mean(EPA), .groups = "drop")

ratings <- fbs |>
  inner_join(offense, by = "team") |>
  inner_join(defense, by = "team") |>
  mutate(net_epa = off_epa - def_epa)

ggplot(ratings, aes(x = off_epa, y = def_epa)) +
  geom_mean_lines(aes(x0 = off_epa, y0 = def_epa), color = "grey50", linewidth = 0.4) +
  geom_sdv_logos(aes(team = team_id), league = "cfb", width = 0.032) +
  scale_y_reverse() +
  labs(
    title = "Every FBS offense and defense, 2025",
    x = "Offense EPA per play",
    y = "Defense EPA per play allowed (better is up)",
    caption = caption
  )

Scatter plot of all 136 FBS teams in the 2025 season, each drawn as a small logo, with offensive EPA per play on the horizontal axis and defensive EPA per play allowed on the vertical axis, reversed so better defenses sit higher, and dashed lines at the averages.

2. Conference small multiples

The same data, one panel per conference: every FBS team as a grey dot behind, the conference’s own teams as logos. The grey layer gets a copy of the data without the conference column, so facet_wrap() repeats it in every panel.

ggplot(ratings, aes(x = off_epa, y = def_epa)) +
  geom_point(data = select(ratings, -conference), color = "grey85", size = 1) +
  geom_sdv_logos(aes(team = team_id), league = "cfb", width = 0.1) +
  scale_y_reverse() +
  facet_wrap(~conference, ncol = 4) +
  labs(
    title = "FBS offense and defense by conference, 2025",
    x = "Offense EPA per play",
    y = "Defense EPA per play allowed (better is up)",
    caption = caption
  ) +
  theme(panel.grid.minor = element_blank())

Eleven small scatter plots, one per FBS conference in 2025, of offensive against defensive EPA per play, with the conference's teams drawn as logos over grey dots for every other FBS team.

3. Conference logos on an axis

How strong is each conference top to bottom? Each team’s SP+ rating (cfbfastR::cfbd_ratings_sp()) is a logo on its conference’s row, with a line at the conference median. The rows are labelled with the conference names, and scale_y_sdv() with theme_y_sdv() draws those names as conference logos: sdvplotR keeps the conferences’ marks alongside the teams’.

sp <- cfbfastR::cfbd_ratings_sp(year = season) |>
  select(team, sp = rating, sp_rank = ranking)

by_conf <- ratings |>
  inner_join(sp, by = "team") |>
  group_by(conference) |>
  mutate(median_sp = median(sp)) |>
  # three staggered rows per conference, so neighboring logos overlap less
  arrange(sp, team, .by_group = TRUE) |>
  mutate(offset = (row_number() %% 3 - 1) * 0.22) |>
  ungroup() |>
  mutate(conference = reorder(conference, median_sp))

medians <- distinct(by_conf, conference, median_sp)

ggplot(by_conf, aes(x = sp, y = conference)) +
  # the discrete layer that sets the rows; the layers below place logos and
  # medians between them in numbers (row 1, row 2, ...)
  geom_blank() +
  geom_vline(xintercept = 0, color = "grey70", linewidth = 0.3) +
  geom_segment(
    aes(x = median_sp, xend = median_sp, y = as.numeric(conference) - 0.4, yend = as.numeric(conference) + 0.4),
    data = medians, color = "grey30", linewidth = 0.6
  ) +
  geom_sdv_logos(aes(y = as.numeric(conference) + offset, team = team_id), league = "cfb", height = 0.03) +
  scale_y_sdv(league = "cfb", size = 28) +
  labs(
    title = "Conference strength by SP+, 2025",
    subtitle = "Each FBS team's SP+ rating; the line is the conference median",
    x = "SP+ rating (points better than an average team)",
    y = NULL,
    caption = "Data: CollegeFootballData.com via cfbfastR | 2025 final SP+"
  ) +
  theme_y_sdv() +
  theme(panel.grid.major.y = element_blank(), panel.grid.minor = element_blank())

Strip chart of the 2025 SP+ rating of every FBS team, one row per conference with the conference's logo as the row label, rows sorted by median rating with the SEC and Big Ten at the top, each team drawn as a small logo and a vertical line at each conference median.

4. A conference standings table

SEC records built from the schedule: conference games and all regular-season games, before the title game and bowls. gt_sdv_logos() turns the abbreviation column into logos, gt_fmt_tally() writes each pair of win and loss columns as one record, and gt_theme_athletic() styles the table. The last column is each team’s real postseason: the bowl or playoff round from the schedule’s notes.

regular <- schedule |>
  filter(completed, season_type == "regular", week <= 14)

sides <- bind_rows(
  regular |> transmute(team = home_team, conf = home_conference, conference_game, pf = home_points, pa = away_points),
  regular |> transmute(team = away_team, conf = away_conference, conference_game, pf = away_points, pa = home_points)
) |>
  mutate(win = pf > pa)

# Title games and bowls, one row per team per game, in date order
late <- schedule |>
  filter(completed, season_type == "postseason" | week > 14) |>
  arrange(start_date)
postseason <- bind_rows(
  transmute(late, start_date, team = home_team, won = home_points > away_points, notes),
  transmute(late, start_date, team = away_team, won = away_points > home_points, notes)
) |>
  arrange(start_date) |>
  mutate(
    # "College Football Playoff Quarterfinal at the Allstate Sugar Bowl" -> "CFP Quarterfinal"
    game = sub("^College Football Playoff (.*?)( Game| at the .*| Presented by .*)?$", "CFP \\1", notes, perl = TRUE),
    game = paste(if_else(won, "W", "L"), coalesce(game, "conference title game"))
  ) |>
  group_by(team) |>
  summarise(postseason = paste(game, collapse = ", "), .groups = "drop")

sec <- sides |>
  filter(conf == "SEC") |>
  group_by(team) |>
  summarise(
    conf_w = sum(win & conference_game),
    conf_l = sum(!win & conference_game),
    w = sum(win),
    l = sum(!win),
    pf = sum(pf),
    pa = sum(pa),
    .groups = "drop"
  ) |>
  left_join(sp, by = "team") |>
  left_join(postseason, by = "team") |>
  # conference record, then overall record: not the SEC's tiebreakers
  arrange(desc(conf_w), desc(w), desc(sp), team) |>
  mutate(logo = sdv_resolve(team, league = "cfb")) |>
  select(logo, team, conf_w, conf_l, w, l, pf, pa, sp, postseason)

sec |>
  gt() |>
  gt_sdv_logos(columns = "logo", league = "cfb", height = 24) |>
  gt_fmt_tally(columns = c(conf_w, conf_l), label = "SEC") |>
  gt_fmt_tally(columns = c(w, l), label = "Overall") |>
  fmt_number(columns = "sp", decimals = 1) |>
  sub_missing(columns = "postseason", missing_text = "") |>
  cols_align(align = "left", columns = "postseason") |>
  data_color(columns = "sp", palette = c("#f7f7f7", "#1b7837")) |>
  cols_label(logo = "", team = "", pf = "PF", pa = "PA", sp = "SP+", postseason = "Postseason") |>
  tab_header(title = "SEC standings, 2025", subtitle = "Regular season, before the SEC title game") |>
  tab_source_note("Data: CollegeFootballData.com via cfbfastR") |>
  gt_theme_athletic(density = "compact")

		SEC	Overall	PF	PA	SP+	Postseason
SEC standings, 2025
Regular season, before the SEC title game
	Georgia	7-1	11-1	387	200	24.1	W SEC Championship, L CFP Quarterfinal
	Ole Miss	7-1	11-1	447	241	24.0	W CFP First Round, W CFP Quarterfinal, L CFP Semifinal
	Texas A&M	7-1	11-1	436	263	20.7	L CFP First Round
	Alabama	7-1	10-2	399	198	14.8	L SEC Championship, W CFP First Round, L CFP Quarterfinal
	Vanderbilt	6-2	10-2	473	263	20.3	L ReliaQuest Bowl
	Oklahoma	6-2	10-2	317	167	18.3	L CFP First Round
	Texas	6-2	9-3	355	237	16.2	W Cheez-It Citrus Bowl
	Tennessee	4-4	8-4	489	345	15.0	L Liberty Mutual Music City Bowl
	Missouri	4-4	8-4	386	233	14.4	L TaxSlayer Gator Bowl
	LSU	3-5	7-5	262	220	10.3	L Kinder's Texas Bowl
	Kentucky	2-6	5-7	276	317	1.8
	Florida	2-6	4-8	259	288	3.5
	Auburn	1-7	5-7	321	248	11.6
	Mississippi State	1-7	5-7	366	350	4.1	L Duke's Mayo Bowl
	South Carolina	1-7	4-8	272	265	5.9
	Arkansas	0-8	2-10	395	406	5.1
Data: CollegeFootballData.com via cfbfastR

5. The national champion’s season, with a logo in the title

Indiana won every game. Each bar is one game’s margin in Indiana’s primary color, the opponent’s logo above it, and ggtitle_image() puts the Indiana logo beside the title. The shaded games are the Big Ten title game and the playoff.

champion <- schedule |>
  filter(grepl("National Championship", notes)) |>
  mutate(winner = if_else(home_points > away_points, home_team, away_team)) |>
  pull(winner)

games <- schedule |>
  filter(completed, home_team == champion | away_team == champion) |>
  arrange(start_date) |>
  mutate(
    game = row_number(),
    home = home_team == champion,
    opponent = if_else(home, away_team, home_team),
    margin = if_else(home, home_points - away_points, away_points - home_points),
    title_run = season_type == "postseason" | week > 14
  )

record <- sprintf("%d-%d", sum(games$margin > 0), sum(games$margin < 0))
run_start <- min(games$game[games$title_run])

ggplot(games, aes(x = game, y = margin)) +
  annotate(
    "rect", xmin = run_start - 0.5, xmax = max(games$game) + 0.5, ymin = -Inf, ymax = Inf,
    fill = "grey93"
  ) +
  annotate(
    "text", x = (run_start + max(games$game)) / 2, y = max(games$margin) + 16,
    label = "Big Ten title game and playoff", size = 3.2, color = "grey40"
  ) +
  geom_col(fill = sdv_team_colors("cfb", champion), width = 0.7) +
  geom_sdv_logos(aes(y = margin + 6, team = opponent), league = "cfb", height = 0.085) +
  scale_x_continuous(breaks = games$game) +
  scale_y_continuous(limits = c(0, max(games$margin) + 18), expand = c(0, 0)) +
  ggtitle_image(
    title_image = champion,
    title = sprintf("%s's %s national title season", champion, record),
    subtitle = "Margin of victory in every game, opponents' logos above the bars",
    image_height = 30,
    league = "cfb"
  ) +
  labs(x = "Game", y = "Margin of victory", caption = "Data: CollegeFootballData.com via cfbfastR") +
  theme_title_image(size = 15, face = "bold") +
  theme(panel.grid.major.x = element_blank(), panel.grid.minor = element_blank())

Bar chart of Indiana's margin of victory in each of its 16 games of the 2025 season in Indiana crimson, with each opponent's logo above its bar, the Big Ten title game and three playoff games shaded, and the Indiana logo beside the title.

6. The College Football Playoff bracket

The 11 playoff games, from the schedule’s notes, drawn as a bracket. Each column is a round; the team that advanced is drawn solid and the team that lost faded, with the score where the two meet. Teams with a first-round bye enter in the quarterfinal column. The layout walks back from the champion: a team’s slot in one round sits between the two teams of the game it won the round before.

rounds <- c("First round", "Quarterfinals", "Semifinals", "Final", "Champion")

cfp <- schedule |>
  filter(grepl("College Football Playoff", notes)) |>
  mutate(
    round = case_when(
      grepl("First Round", notes) ~ 1,
      grepl("Quarterfinal", notes) ~ 2,
      grepl("Semifinal", notes) ~ 3,
      grepl("Championship", notes) ~ 4
    ),
    winner = if_else(home_points > away_points, home_team, away_team),
    loser = if_else(home_points > away_points, away_team, home_team),
    score = paste0(pmax(home_points, away_points), "-", pmin(home_points, away_points))
  )

# A team's slot in round r: a leaf if it played no game in round r - 1 (a
# bye, or the first round), otherwise centered on that game's two teams
slots <- data.frame()
edges <- data.frame()
leaf <- 0
place <- function(team, r) {
  prev <- cfp[cfp$round == r - 1 & cfp$winner == team, ]
  if (nrow(prev) == 0) {
    leaf <<- leaf + 1
    y <- leaf
  } else {
    ys <- c(place(prev$winner, r - 1), place(prev$loser, r - 1))
    y <- mean(ys)
    edges <<- rbind(edges, data.frame(x = r - 1, y1 = ys[1], y2 = ys[2], y = y, score = prev$score))
  }
  lost <- any(cfp$round == r & cfp$loser == team)
  slots <<- rbind(slots, data.frame(team = team, x = r, y = y, lost = lost))
  y
}
invisible(place(cfp$winner[cfp$round == 4], 5))
slots$team_id <- sdv_resolve(slots$team, league = "cfb")

ggplot() +
  # each game: from both teams to the round's midline, then on to the winner
  geom_segment(aes(x = x + 0.22, xend = x + 0.5, y = y1, yend = y1), data = edges, color = "grey60") +
  geom_segment(aes(x = x + 0.22, xend = x + 0.5, y = y2, yend = y2), data = edges, color = "grey60") +
  geom_segment(aes(x = x + 0.5, xend = x + 0.5, y = y1, yend = y2), data = edges, color = "grey60") +
  geom_segment(aes(x = x + 0.5, xend = x + 0.78, y = y, yend = y), data = edges, color = "grey60") +
  geom_label(
    aes(x = x + 0.5, y = y, label = score),
    data = edges, size = 3, linewidth = 0, fill = "white", color = "grey25"
  ) +
  geom_sdv_logos(
    aes(x = x, y = y, team = team_id, alpha = if_else(lost, 0.3, 1)),
    data = slots, league = "cfb", height = 0.075
  ) +
  scale_alpha_identity() +
  scale_x_continuous(breaks = 1:5, labels = rounds, position = "top", limits = c(0.6, 5.4)) +
  scale_y_reverse() +
  labs(
    title = "The 2025-26 College Football Playoff",
    x = NULL,
    y = NULL,
    caption = "Data: CollegeFootballData.com via cfbfastR"
  ) +
  theme(panel.grid = element_blank(), axis.text.y = element_blank(), axis.text.x = element_text(face = "bold"))

The 2025-26 College Football Playoff bracket drawn with team logos, rounds from left to right: first round, quarterfinals, semifinals, final and champion. Winners are solid, losers faded, each game's score is printed where the two teams meet, and Indiana is the champion on the right.

7. A rivalry, season by season

Ohio State against Michigan since 2004, from one CFBD call (cfbfastR::cfbd_team_matchup()). Each bar is the margin from Ohio State’s side, colored for the winner, with the winner’s logo at the end of the bar. There was no game in 2020.

series <- cfbfastR::cfbd_team_matchup("Ohio State", "Michigan", min_year = 2004, max_year = season) |>
  filter(!is.na(home_score)) |>
  mutate(
    osu_margin = if_else(home_team == "Ohio State", home_score - away_score, away_score - home_score),
    winner_id = sdv_resolve(winner, league = "cfb"),
    tip = osu_margin + if_else(osu_margin > 0, 6, -6)
  )

ggplot(series, aes(x = season, y = osu_margin)) +
  geom_col(aes(fill = winner_id), width = 0.75) +
  geom_hline(yintercept = 0, color = "grey30") +
  geom_sdv_logos(aes(y = tip, team = winner_id), league = "cfb", height = 0.075) +
  annotate("text", x = 2020, y = 2, label = "no game", angle = 90, hjust = 0, size = 3, color = "grey45") +
  scale_fill_sdv(league = "cfb") +
  scale_x_continuous(breaks = seq(2004, season, 3)) +
  scale_y_continuous(limits = c(-35, 50)) +
  labs(
    title = sprintf("The Game: Ohio State vs Michigan, 2004-%d", season),
    subtitle = sprintf(
      "Ohio State %d wins, Michigan %d",
      sum(series$osu_margin > 0), sum(series$osu_margin < 0)
    ),
    x = NULL,
    y = "Ohio State margin",
    caption = "Data: CollegeFootballData.com via cfbfastR"
  ) +
  theme(legend.position = "none", panel.grid.minor = element_blank())

Bar chart of Ohio State's margin against Michigan in each season from 2004 to 2025, bars above zero in Ohio State scarlet and below zero in Michigan blue, each with the winner's logo at the end of the bar and a note that there was no game in 2020.

8. A quarterback leaderboard with headshots

cfbfastR’s play-by-play names the passer on each dropback by role (completion_player_id, incompletion_player_id, …); the first that is set is the quarterback’s ESPN athlete id, which is what geom_sdv_headshots() takes for college players. FBS quarterbacks only, 300 or more dropbacks, all games.

qbs <- pbp |>
  filter(pass == 1, !is.na(EPA)) |>
  mutate(
    qb_id = coalesce(completion_player_id, incompletion_player_id, interception_thrown_player_id, sack_taken_player_id),
    qb = coalesce(completion_player, incompletion_player, interception_thrown_player, sack_taken_player)
  ) |>
  filter(!is.na(qb_id), pos_team %in% fbs$team) |>
  arrange(game_id, id_play) |>
  group_by(qb_id) |>
  summarise(
    qb = last(qb),
    team = last(pos_team),
    dropbacks = n(),
    epa = mean(EPA),
    .groups = "drop"
  ) |>
  filter(dropbacks >= 300) |>
  arrange(desc(epa), qb_id) |>
  head(15) |>
  mutate(team_id = sdv_resolve(team, league = "cfb"), qb_id = as.character(qb_id))

ggplot(qbs, aes(x = epa, y = reorder(qb, epa))) +
  geom_col(aes(fill = team_id), width = 0.75) +
  geom_sdv_logos(aes(x = -0.016, team = team_id), league = "cfb", height = 0.05) +
  geom_sdv_headshots(aes(x = epa + 0.022, player_id = qb_id), league = "cfb", height = 0.06) +
  scale_fill_sdv(league = "cfb") +
  scale_x_continuous(breaks = seq(0, 0.4, 0.1), expand = expansion(mult = c(0.02, 0.08))) +
  labs(
    title = "Top 15 FBS quarterbacks by EPA per dropback, 2025",
    subtitle = "300+ dropbacks, all games",
    x = "EPA per dropback",
    y = NULL,
    caption = caption
  ) +
  theme(legend.position = "none", panel.grid.major.y = element_blank())

Horizontal bar chart of the 15 FBS quarterbacks with the highest EPA per dropback in the 2025 season (300 or more dropbacks), each bar in the team's color with the team's logo before it and the quarterback's headshot at its end.

9. A ranked bar chart with logos on the y axis

The final SP+ top 25. The bars use the teams’ colors, and with the team abbreviations on the y axis, scale_y_sdv() plus theme_y_sdv() replaces each with its logo.

top25 <- sp |>
  inner_join(fbs, by = "team") |>
  arrange(sp_rank, team) |>
  head(25) |>
  mutate(team_id = factor(team_id, levels = rev(team_id)))

ggplot(top25, aes(x = sp, y = team_id)) +
  geom_col(aes(fill = team_id), width = 0.75) +
  geom_text(aes(label = sprintf("%.1f", sp)), hjust = -0.2, size = 3) +
  scale_fill_sdv(league = "cfb") +
  scale_y_sdv(league = "cfb", size = 18) +
  scale_x_continuous(expand = expansion(mult = c(0, 0.08))) +
  labs(
    title = "Final SP+ top 25, 2025",
    x = "SP+ rating",
    y = NULL,
    caption = "Data: CollegeFootballData.com via cfbfastR"
  ) +
  theme_y_sdv() +
  theme(legend.position = "none", panel.grid.major.y = element_blank())

Horizontal bar chart of the top 25 teams in the final 2025 SP+ ratings, bars in team colors with each team's logo in place of its name on the vertical axis and the rating printed at the end of each bar.

10. Tiers of a ranking you compute

A composite ranking: the average of each FBS team’s rank in three systems, SP+, ESPN’s FPI (cfbfastR::espn_ratings_fpi()) and the Simple Rating System (cfbfastR::cfbd_ratings_srs()). The top 32 go into five tiers with sdv_team_tiers(). FPI comes keyed by ESPN abbreviation, which sdv_resolve() turns into a team id.

fpi <- cfbfastR::espn_ratings_fpi(year = season) |>
  transmute(team_id = suppressWarnings(sdv_resolve(team_abbreviation, league = "cfb")), fpi = as.numeric(fpi))
srs <- cfbfastR::cfbd_ratings_srs(year = season) |>
  select(team, srs = rating)

composite <- fbs |>
  inner_join(sp, by = "team") |>
  inner_join(srs, by = "team") |>
  inner_join(fpi, by = "team_id") |>
  mutate(score = (rank(-sp) + rank(-fpi) + rank(-srs)) / 3) |>
  arrange(score, team_id) |>
  head(32)

sizes <- c(4, 6, 7, 7, 8) # teams per tier, top to bottom
composite$tier_no <- rep(seq_along(sizes), sizes)

sdv_team_tiers(
  select(composite, tier_no, team = team_id),
  league = "cfb",
  title = "College football tiers, 2025",
  subtitle = "Average rank in SP+, FPI and SRS among FBS teams",
  caption = "Data: CollegeFootballData.com and ESPN via cfbfastR",
  alpha = 1,
  tier_desc = c(
    "1" = "Elite",
    "2" = "Contenders",
    "3" = "Very good",
    "4" = "Good",
    "5" = "Solid"
  )
)

College football team logos in five labeled tiers, from Elite to Solid, for the 32 best FBS teams of 2025 by their average rank in SP+, FPI and SRS, best first within each tier.

11. Make it interactive

The 136-team scatter from section 1 is crowded in the middle, which is where an interactive chart helps most. This one is native plotly: a transparent marker under each team carries the hover text, and add_sdv_logos() places the logos as layout images at the same coordinates. Hover a logo for the school, its 2025 conference and its EPA per play on each side; drag a box to zoom into the cluster and double-click to zoom back out. plotly loads each logo from its URL, so 136 of them add almost nothing to the page.

pad <- 0.03
fbs_hover <- ratings |>
  mutate(tip = sprintf(
    "<b>%s</b> (%s)<br>Offense: %+.3f EPA/play<br>Defense: %+.3f EPA/play allowed<br>Net: %+.3f",
    team, conference, off_epa, def_epa, net_epa
  ))
avg_line <- list(color = "grey", dash = "dash", width = 1)

plotly::plot_ly(
  fbs_hover, x = ~off_epa, y = ~def_epa, type = "scatter", mode = "markers",
  marker = list(size = 14, opacity = 0), text = ~tip, hoverinfo = "text"
) |>
  plotly::layout(
    title = list(text = "Every FBS offense and defense, 2025", x = 0.02),
    xaxis = list(title = "Offense EPA per play", range = range(fbs_hover$off_epa) + c(-pad, pad), zeroline = FALSE),
    yaxis = list(
      title = "Defense EPA per play allowed (better is up)", zeroline = FALSE,
      range = rev(range(fbs_hover$def_epa) + c(-pad, pad))
    ),
    shapes = list(
      list(type = "line", x0 = mean(fbs_hover$off_epa), x1 = mean(fbs_hover$off_epa), yref = "paper", y0 = 0, y1 = 1, line = avg_line),
      list(type = "line", y0 = mean(fbs_hover$def_epa), y1 = mean(fbs_hover$def_epa), xref = "paper", x0 = 0, x1 = 1, line = avg_line)
    )
  ) |>
  add_sdv_logos(fbs_hover$off_epa, fbs_hover$def_epa, fbs_hover$team_id, "cfb", size = 0.04)

A range given high to low reverses the axis, so good defenses sit at the top as in the static chart. Interactive charts covers plotly, ggiraph and leaflet, and the CFB stadium map recipe puts the same logos on a map of the stadiums with addSdvLogoMarkers().

Next steps

cfbfastR documents the play-by-play columns and every cfbd_*() endpoint used here.
sdv_surface("cfb", team) draws a college field in a team’s colors; the NFL tutorial puts plays on one.
cfbseedR applies conference tiebreakers and simulates the playoff field.

Our Authors

Citation

To cite the sdvplotR R package in publications, use:

BibTeX Citation

@misc{gilani_carl_lee_kay_weatherman_sdvplotR,
  author = {Gilani, Saiem and Carl, Sebastian and Lee, Jared and Kay, Camden and Weatherman, Andrew},
  title = {sdvplotR: The SportsDataverse's R Package for Multi-League Sports Plotting.},
  url = {https://sdvplotR.sportsdataverse.org},
  year = {2026}
}

sdvplot - the Python plotting package
sdvplot-js - the JavaScript / TypeScript plotting packages
cfbfastR - college football
hoopR - men’s basketball
wehoop - women’s basketball
baseballr - baseball
fastRhockey - hockey
oddsapiR - betting odds
sportyR - playing surfaces
sportsdataverse-py - the Python package
sportsdataverse-R - the R meta-package

Saiem Gilani