mlb_cpr

Data: MLB.com via {baseballr} | Last Updated: October 5, 2023 at 08:07:46

Contents

Team Rankings

Runs Scored v Runs Allowed

Composite Performance Rating (CPR) Rankings

Scorigami (2023 Only)

Historic MLB Scorigami (Since 1901)

Top Team Analysis

Team Margins Plot

Scatterplot of Margins of Victory and Defeat

Margins of Victory and Defeat

One-Run Games

Best Records This Month

Runs Scored v Runs Allowed This Month

Eras Records

Home and Away Performance

Monthly v Season Win Percentages

Win Percentage v Run Differential as Percent of Runs Scored

Which teams play the closest games?

Pythagorean Wins

## [1] "Run-adjusted margin is more correlated than pythagorean wins (0.942 vs. 0.941)"

Close Games

Best Team of Past Month?

# getting vector of all team names
all_teams = mlb_teams(season = 2023) |>
  filter(sport_name == "Major League Baseball") |>
  pull(team_full_name) |>
  sort()

# getting end game data and adding info
end_games = mlb_schedule(season = 2023, level_ids = "1") |>
  filter(teams_away_team_name %in% all_teams &
           teams_home_team_name %in% all_teams &
           series_description == "Regular Season" &
           status_detailed_state == "Final") |>
  select(date, game_pk,
         away_team = teams_away_team_name, away_score = teams_away_score,
         home_score = teams_home_score, home_team = teams_home_team_name) |>
  mutate(date = as_date(date),
         win_team = ifelse(home_score > away_score, home_team, away_team),
         win_score = ifelse(home_score > away_score, home_score, away_score),
         lose_score = ifelse(home_score > away_score, away_score, home_score),
         lose_team = ifelse(home_score > away_score, away_team, home_team),
         final_score = paste0(win_score, "-", lose_score),
         description = paste0(win_team, " def. ", lose_team, " ", final_score))

# best records of the past month
# end_games |>
#   filter(date >= Sys.Date() - 31) |>
#   count(win_team) |>
#   rename(team = win_team, wins = n) |>
#   inner_join(end_games |>
#   filter(date >= Sys.Date() - 31) |>
#   count(lose_team) |>
#   rename(team = lose_team, losses = n), by = "team") |>
#   mutate(win_pct = round(wins / (wins + losses), 3)) |>
#   arrange(desc(win_pct))

# function to get runs scored in past month
get_past_month_runs_scored = function(team) {
  home_runs = end_games |> filter(date >= Sys.Date() - 31 & home_team == team) |> pull(home_score) |> sum()
  away_runs = end_games |> filter(date >= Sys.Date() - 31 & away_team == team) |> pull(away_score) |> sum()
  return(home_runs + away_runs)
}

# function to get runs allowed in past month
get_past_month_runs_allowed = function(team) {
  home_runs = end_games |> filter(date >= Sys.Date() - 31 & home_team == team) |> pull(away_score) |> sum()
  away_runs = end_games |> filter(date >= Sys.Date() - 31 & away_team == team) |> pull(home_score) |> sum()
  return(home_runs + away_runs)
}

# team color codes
team_color_codes = c("#A71930", "#CE1141", "#DF4601", "#BD3039", "#0E3386",
                     "#27251F", "#C6011F", "#00385D", "#333366", "#0C2340",
                     "#EB6E1F", "#004687", "#BA0021", "#005A9C", "#00A3E0",
                     "#FFC52F", "#D31145", "#FF5910", "#003087", "#003831",
                     "#E81828", "#FDB827", "#FFC425", "#FD5A1E", "#005C5C",
                     "#C41E3A", "#8FBCE6", "#C0111F", "#134A8E", "#FFB7C5")

# getting run-adjusted margins for each team in past month
past_month_ram = data.frame(team = all_teams) |>
  mutate(scored = sapply(team, get_past_month_runs_scored),
         allowed = sapply(team, get_past_month_runs_allowed),
         diff = scored - allowed,
         ram = round(diff / scored, 3),
         pos_lab = ifelse(ram >= 0, ram, ""),
         neg_lab = ifelse(ram < 0, ram, ""))

past_month_ram |>
  ggplot(aes(reorder(team, ram), ram)) +
  geom_col(aes(fill = team), show.legend = F) +
  geom_text(aes(label = pos_lab), size = 3, hjust = -0.25) +
  geom_text(aes(label = neg_lab), size = 3, hjust = 1.25) +
  annotate("text", x = 1.5, y = 0.25, label = "Data: MLB.com via {baseballr}", size = 3, fontface = "italic") +
  scale_fill_manual(values = team_color_codes) +
  scale_y_continuous(breaks = seq(-3, 3, by = 0.1)) +
  coord_flip(ylim = c(min(past_month_ram$ram) * 1.1, max(past_month_ram$ram) * 1.1)) +
  labs(x = NULL, y = "RAM",
       title = "Past Month Run-Adjusted Margin (RAM)",
       subtitle = "RAM = (Runs Scored - Runs Allowed) / Runs Scored")


margin_results = end_games |>
  mutate(win_margin = win_score - lose_score,
         margin_description = paste0(win_team, " win by ", win_margin),
         loser_desc = paste0(lose_team, " lose by ", win_margin)) |>
  count(margin_description) |>
  mutate(desc = paste0(margin_description, " (", n, ")")) |>
  arrange(desc(n), margin_description) |>
  pull(desc)

margin_results_L = end_games |>
  mutate(win_margin = win_score - lose_score,
         margin_description = paste0(win_team, " win by ", win_margin),
         loser_desc = paste0(lose_team, " lose by ", win_margin)) |>
  count(loser_desc) |>
  mutate(desc = paste0(loser_desc, " (", n, ")")) |>
  arrange(desc(n), loser_desc) |>
  pull(desc)

upper_limit = 10

for (i in 1:upper_limit) {
  print(margin_results[i])
}

## [1] "Cincinnati Reds win by 1 (34)"
## [1] "Miami Marlins win by 1 (33)"
## [1] "Baltimore Orioles win by 1 (30)"
## [1] "Milwaukee Brewers win by 1 (29)"
## [1] "Philadelphia Phillies win by 1 (29)"
## [1] "Washington Nationals win by 1 (29)"
## [1] "Cleveland Guardians win by 1 (27)"
## [1] "New York Mets win by 1 (25)"
## [1] "San Francisco Giants win by 1 (25)"
## [1] "Seattle Mariners win by 1 (25)"

print("========================================")

## [1] "========================================"

for (i in 1:upper_limit) {
  print(margin_results_L[i])
}

## [1] "Cleveland Guardians lose by 1 (31)"
## [1] "Chicago White Sox lose by 1 (30)"
## [1] "Cincinnati Reds lose by 1 (29)"
## [1] "New York Mets lose by 1 (29)"
## [1] "Minnesota Twins lose by 1 (27)"
## [1] "Oakland Athletics lose by 1 (27)"
## [1] "Seattle Mariners lose by 1 (26)"
## [1] "St. Louis Cardinals lose by 1 (26)"
## [1] "Boston Red Sox lose by 1 (25)"
## [1] "Tampa Bay Rays lose by 1 (25)"

end_games |>
  mutate(diff = win_score - lose_score) |>
  filter(diff >= 3) |>
  count(win_team) |>
  arrange(desc(n)) |>
  rename(`Team` = win_team,
         `Wins by 3+` = n)

## # A tibble: 30 × 2
##    Team                `Wins by 3+`
##    <chr>                      <int>
##  1 Los Angeles Dodgers           74
##  2 Tampa Bay Rays                64
##  3 Texas Rangers                 62
##  4 Atlanta Braves                61
##  5 San Diego Padres              57
##  6 Toronto Blue Jays             55
##  7 Baltimore Orioles             54
##  8 Chicago Cubs                  53
##  9 Houston Astros                52
## 10 New York Yankees              48
## # ℹ 20 more rows

reg23 = end_games |>
  mutate(total_score = home_score + away_score)

post22 = mlb_schedule(season = 2022, level_ids = "1") |>
  filter(teams_away_team_name %in% all_teams &
           teams_home_team_name %in% all_teams &
           series_description %in% c("Division Series", "League Championship Series", "Wild Card Game", "World Series") &
           status_detailed_state == "Final") |>
  select(date, game_pk,
         away_team = teams_away_team_name, away_score = teams_away_score,
         home_score = teams_home_score, home_team = teams_home_team_name) |>
  mutate(date = as_date(date),
         win_team = ifelse(home_score > away_score, home_team, away_team),
         win_score = ifelse(home_score > away_score, home_score, away_score),
         lose_score = ifelse(home_score > away_score, away_score, home_score),
         lose_team = ifelse(home_score > away_score, away_team, home_team),
         final_score = paste0(win_score, "-", lose_score),
         description = paste0(win_team, " def. ", lose_team, " ", final_score),
         total_score = home_score + away_score)

# most postseason games had six or less runs scored
# post22 |>
#   count(total_score <= 6)

reg23 |>
  filter(total_score <= 6) |>
  count(win_team) |>
  rename(team = win_team, wins = n) |>
  inner_join(reg23 |>
  filter(total_score <= 6) |>
  count(lose_team) |>
  rename(team = lose_team, losses = n), by = "team") |>
  mutate(win_pct = round(wins / (wins + losses), 3)) |>
  arrange(desc(win_pct))

## # A tibble: 30 × 4
##    team                  wins losses win_pct
##    <chr>                <int>  <int>   <dbl>
##  1 Atlanta Braves          24     10   0.706
##  2 Los Angeles Dodgers     26     12   0.684
##  3 Baltimore Orioles       30     17   0.638
##  4 Toronto Blue Jays       31     20   0.608
##  5 Seattle Mariners        31     21   0.596
##  6 Arizona Diamondbacks    25     17   0.595
##  7 Pittsburgh Pirates      29     22   0.569
##  8 Los Angeles Angels      26     20   0.565
##  9 Tampa Bay Rays          24     19   0.558
## 10 Chicago Cubs            25     20   0.556
## # ℹ 20 more rows

end_games |>
  mutate(runs_scored = home_score + away_score) |>
  count(runs_scored) |>
  ggplot(aes(runs_scored, n)) +
  geom_col(fill = "springgreen4") +
  scale_x_continuous(breaks = seq(0, 30, by = 1)) +
  scale_y_continuous(breaks = seq(0, 250, by = 25)) +
  labs(x = "Runs Scored", y = "Number of Games")