Get Started with cfb4th

First load the packages and some data. load_4th_pbp() loads cfbfastR data and computes 4th down probabilities (depending on your computer, this may take up to a minute or two per season).

library(cfb4th)
library(dplyr)
library(knitr)
library(tibble)
library(gt)

Easy mode: using cfbfastR data

Here’s what the data obtained using load_4th_pbp() looks like:

suppressWarnings(
  data <- cfb4th::load_4th_pbp(2020) %>% 
    dplyr::filter(.data$week %in% 1:14)
)
data %>%
  dplyr::filter(!is.na(go_boost)) %>%
  utils::head(10) %>%
  dplyr::select(
    pos_team, distance, yards_to_goal, go_boost, first_down_prob, 
    wp_fail, wp_succeed, go_wp, fg_make_prob, miss_fg_wp, make_fg_wp, 
    fg_wp, punt_wp
  ) %>%
  knitr::kable(digits = 2)

Or we can add some filters to look up a certain game:

data %>%
  dplyr::filter(week == 12, pos_team == "Utah", down == 4) %>%
  dplyr::select(
    pos_team, distance, yards_to_goal, go_boost, first_down_prob, 
    wp_fail, wp_succeed, go_wp, fg_make_prob, miss_fg_wp, make_fg_wp, 
    fg_wp, punt_wp
  ) %>%
  knitr::kable(digits = 2)

Calculations from user input

The below shows the bare minimum amount of information that has to be fed to cfb4th in order to compute 4th down decision recommendations. The main function on user-input data is add_4th_probs().

Teams are included to help the model easily track the simulations.

one_play <-
  tibble::tibble(
      # Game Info
      home = "Utah",
      away = "BYU",
      pos_team = "Utah",
      def_pos_team = "BYU",
      spread = -7,
      over_under = 55,

      # Situation Info
      half = 2,
      period = 3, # Quarter
      TimeSecsRem = 900, # Half Seconds Remaining
      adj_TimeSecsRem = 900, # Game Seconds Remaining
      down = 4,
      distance = 4,
      yards_to_goal = 40,
      pos_score_diff_start = 7,

      pos_team_receives_2H_kickoff = 1,
      pos_team_timeouts_rem_before = 3,
      def_pos_team_timeouts_rem_before = 3

    )
one_play %>%
  cfb4th::add_4th_probs() %>%
  dplyr::select(
    pos_team, distance, yards_to_goal, go_boost, first_down_prob, 
    wp_fail, wp_succeed, go_wp, fg_make_prob, miss_fg_wp, make_fg_wp, 
    fg_wp, punt_wp
  ) %>%
  knitr::kable(digits = 2)
#> Computing probabilities for 1 plays. . .
#> [04:25:36] WARNING: src/learner.cc:553: 
#>   If you are loading a serialized model (like pickle in Python, RDS in R) generated by
#>   older XGBoost, please export the model by calling `Booster.save_model` from that version
#>   first, then load it back in current version. See:
#> 
#>     https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html
#> 
#>   for more details about differences between saving model and serializing.
#> 
#> [04:25:36] WARNING: src/learner.cc:1203: 
#>   If you are loading a serialized model (like pickle in Python, RDS in R) generated by
#>   older XGBoost, please export the model by calling `Booster.save_model` from that version
#>   first, then load it back in current version. See:
#> 
#>     https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html
#> 
#>   for more details about differences between saving model and serializing.
#> 
#> [04:25:36] WARNING: src/learner.cc:888: Found JSON model saved before XGBoost 1.6, please save the model using current version again. The support for old JSON model will be discontinued in XGBoost 2.3.
#> [04:25:36] WARNING: src/learner.cc:553: 
#>   If you are loading a serialized model (like pickle in Python, RDS in R) generated by
#>   older XGBoost, please export the model by calling `Booster.save_model` from that version
#>   first, then load it back in current version. See:
#> 
#>     https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html
#> 
#>   for more details about differences between saving model and serializing.

pos_team	distance	yards_to_goal	go_boost	first_down_prob	wp_fail	wp_succeed	go_wp	fg_make_prob	miss_fg_wp	make_fg_wp	fg_wp	punt_wp
Utah	4	40	1.4	0.49	0.81	0.92	0.87	0.3	0.81	0.91	0.84	0.85

Make a summary table

Let’s put the play above into a table using the provided function make_table_data(), which makes it easier to interpret the recommendations for a play. This function only works with one play at a time since it makes a table using the results from the play.

one_play %>%
  cfb4th::add_4th_probs() %>%
  cfb4th::make_table_data() %>%
  knitr::kable(digits = 1)
#> Computing probabilities for 1 plays. . .

choice	choice_prob	success_prob	fail_wp	success_wp
Go for it	86.5	49.4	81.2	92.0
Punt	85.1	NA	NA	NA
Field goal attempt	84.0	30.4	80.7	91.3

Looking at the table, the offense would be expected to have 86.5% win probability if they had gone for it and 85% if they punted.

Getting 4th down plays from a live game

cfbfastR isn’t available for live games and typing all the plays in by hand is annoying. So how does the 4th down bot work? With thanks to the ESPN API, which can be accessed using get_4th_plays().

game <- cfbfastR::cfbd_game_info(year = 2019, team = "Utah", week = 4)
plays <- cfb4th::get_4th_plays(game) %>% 
  tail(1)
plays %>% 
  dplyr::select("desc", "TimeSecsRem")
#> # A tibble: 1 × 2
#>   desc                        TimeSecsRem
#>   <chr>                             <dbl>
#> 1 Jadon Redding 38 yd FG GOOD         241
plays %>% 
  cfb4th::add_4th_probs() %>%
  cfb4th::make_table_data() %>%
  knitr::kable(digits = 1)
#> Computing probabilities for 1 plays. . .

choice	choice_prob	success_prob	fail_wp	success_wp
Field goal attempt	7.5	72.3	1.8	9.7
Go for it	7.5	29.0	2.3	20.2
Punt	NA	NA	NA	NA

Ben Baldwin & Jared Lee

Easy mode: using cfbfastR data

Calculations from user input

Make a summary table

Getting 4th down plays from a live game