# function from hoopR to fetch data
acc <- c("North Carolina", "Duke", "Wake Forest", "N.C. State", "Pittsburgh",
"Virginia Tech", "Georgia Tech", "Notre Dame", "Virginia", "Boston College",
"Louisville", "Syracuse", "Clemson", "Florida St.", "Miami FL")
acc_team <- function(id) {
kp_team_schedule(team = id, year = 2023)
}
# function to get ACC games
teams_acc <- lapply(acc, acc_team)
# build table of only conference games
acc_results_2023 <- as.data.frame(do.call(rbind, teams_acc)) %>%
mutate(conference_game = if_else(postseason == "Atlantic Coast Conference Tournament", TRUE, conference_game)) %>%
filter(conference_game == TRUE) %>%
mutate(full_result = str_split(result, ",", simplify = T),
points_for = str_split(full_result[,2], "-", simplify = T),
fr = full_result[,1],
pf = points_for[,1],
pa = points_for[,2]) %>%
select(game_date, game_id, team, opponent, opponent_rk, location, fr, pf, pa) %>%
mutate(diff = as.numeric(pf) - as.numeric(pa),
year = 2023) %>%
rename(result = fr, points_scored = pf, points_allowed = pa)
# find wins and losses
wins_losses <- acc_results_2023 %>%
group_by(team, result) %>%
summarise(total = n()) %>%
pivot_wider(names_from = result, values_from = total)
# find opp count
opps <- acc_results_2023 %>%
group_by(team, opponent, opponent_rk) %>%
summarise(total = n())
# fix it up for uva and duke
opps$total[opps$team == 'Duke' & opps$opponent == 'Virginia'] <- 3
opps$total[opps$team == 'Virginia' & opps$opponent == 'Duke'] <- 3
# find the opponent ranking
opp_rk <- opps %>%
mutate(agg_rk = (opponent_rk * total)) %>%
group_by(team) %>%
summarise(opp_rk = mean(agg_rk)) %>%
mutate(rk = dense_rank(desc(-opp_rk))) %>%
mutate(team = if_else(team == "N.C. State", "NC State", team))
# combine table
full_acc_sos <- merge(opp_rk, wins_losses, by = "team") %>%
select(team, W, L, opp_rk)
# net ratings function
net_ranks <- function(url) {
net_page <- read_html(url)
net_rk <- net_page %>%
html_nodes("table") %>%
.[1] %>%
html_table(fill = TRUE)
net_table <- as.data.frame(net_rk)
net_table <- net_table %>%
mutate(date = gsub("[^0-9.-]", "", colnames(net_table[3]))) %>%
mutate(date = str_sub(date, end = -2)) %>%
mutate(record = str_split(WL, "-", simplify = T),
wins = record[,1],
losses = record[,2],
conf_record = str_split(Conf..Record, "-", simplify = T),
wins_conf = conf_record[,1],
losses_conf = conf_record[,2],
non_conf_record = str_split(Non.Conference.Record, "-", simplify = T),
wins_non_conf = non_conf_record[,1],
losses_non_conf = non_conf_record[,2],
first_q = str_split(Quadrant.1, "-", simplify = T),
q1_win = first_q[,1],
q1_loss = first_q[,2],
second_q = str_split(Quadrant.2, "-", simplify = T),
q2_win = second_q[,1],
q2_loss = second_q[,2],
third_q = str_split(Quadrant.3, "-", simplify = T),
q3_win = third_q[,1],
q3_loss = third_q[,2],
fourth_q = str_split(Quadrant.4, "-", simplify = T),
q4_win = fourth_q[,1],
q4_loss = fourth_q[,2],
) %>%
rename(team = Team, conf = Conference, net = NET, prev_net = Prev.NET,
sos = NET.SOS, non_conf_sos = NET.NonConf.SOS) %>%
select(team, conf, net, prev_net, sos, non_conf_sos, wins,
losses, wins_non_conf, losses_non_conf, q1_win, q1_loss,
q2_win, q2_loss, q3_win, q3_loss, q4_win, q4_loss, date) %>%
mutate_at(vars(-team, -conf, -date), as.numeric)
}
net <- net_ranks(url = "https://stats.ncaa.org/selection_rankings/nitty_gritties/30928")
# find net ratings
acc_non_sos <- net %>% filter(conf == "ACC") %>%
arrange(non_conf_sos) %>%
select(team, non_conf_sos) %>%
mutate(team = if_else(team == "Miami (FL)", "Miami FL", team))
# combine two
sos_two <- merge(opp_rk, acc_non_sos, by = "team") %>%
select(team, rk, opp_rk, non_conf_sos)
