,
Ben Williams [aut] ,
Will Palmquist [aut] | Title: | Simulating Plays and Drives in the NFL |
|---|---|
| Description: | The intent here is to enable the simulation of plays/drives and evaluate game-play strategies in the National Football League (NFL). Built-in strategies include going for it on fourth down and varying the proportion of passing/rushing plays during a drive. The user should be familiar with nflscrapR data before trying to write his/her own strategies. This work is inspired by a blog post by Mike Lopez, currently the Director of Data and Analytics at the NFL, Lopez (2019) <https://statsbylopez.netlify.app/post/resampling-nfl-drives/>. |
| Authors: | Ryan Elmore [cre, aut] ,
Ben Williams [aut] ,
Will Palmquist [aut] |
| Maintainer: | Ryan Elmore <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.4.0 |
| Built: | 2025-02-23 03:31:44 UTC |
| Source: | https://github.com/rtelmore/nflsimulator |
The down and distance updater will run a play and update various game-based statistics accordingly.
down_distance_updater( what_down, yards_to_go, yards_from_own_goal, play_by_play_data, ... )down_distance_updater( what_down, yards_to_go, yards_from_own_goal, play_by_play_data, ... )
what_down |
The current down (1st, 2nd, 3rd, or 4th down) |
yards_to_go |
Number of yards to go until a first down or TD |
yards_from_own_goal |
The number of yards from the possession team's own goal |
play_by_play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
... |
Additional arguments for different strategies |
A data.frame object
## Not run: down_distance_updater(what_down = 1, yards_to_go = 10, yards_from_own_goal = 25, play_by_play_data = pbp_data, strategy = "normal") ## End(Not run)## Not run: down_distance_updater(what_down = 1, yards_to_go = 10, yards_from_own_goal = 25, play_by_play_data = pbp_data, strategy = "normal") ## End(Not run)
This function will return the expected points for several 4th down decision. The options are "go for it", "field goal", or "punt". This should be primarily used within the 'NFLSimulatoR::sample_play()' function.
expected_pts_fourth(yards_from_goal, yards_to_go, play_data)expected_pts_fourth(yards_from_goal, yards_to_go, play_data)
yards_from_goal |
The number of yards until a team scores a touchdown |
yards_to_go |
Number of yards to go until a first down or TD |
play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
A data.frame of the expected points of three fourth down options
## Not run: expected_pts_fourth(what_down = 1, yards_to_go = 10, yards_from_own_goal = 25, play_by_play_data = reg_pbp_2018) ## End(Not run)## Not run: expected_pts_fourth(what_down = 1, yards_to_go = 10, yards_from_own_goal = 25, play_by_play_data = reg_pbp_2018) ## End(Not run)
Add necessary columns to nflscrapR data
prep_pbp_data(data)prep_pbp_data(data)
data |
An nflscrapR or nflfastR data set. Note that stringsAsFactors = FALSE is assumed. |
a data.table object
## Not run: dt <- prep_pbp_data(nflscrapr_pbp_data) ## End(Not run)## Not run: dt <- prep_pbp_data(nflscrapr_pbp_data) ## End(Not run)
Sample a Series of Drives, a strategy to test verses the normal strategy
sample_drives( n_sims, from_yard_line = 25, play_by_play_data, strategy = "normal", single_drive = FALSE, progress = TRUE, ... )sample_drives( n_sims, from_yard_line = 25, play_by_play_data, strategy = "normal", single_drive = FALSE, progress = TRUE, ... )
n_sims |
The number of simulations |
from_yard_line |
The starting field position (defaults to 25) |
play_by_play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
strategy |
"normal", "passes_rushes", or "fourth_downs" |
single_drive |
TRUE indicates only a single drive, otherwise, drives sampled until a score occurs |
progress |
logical for inclusion of a progress bar |
... |
Additional arguments for different strategies |
A data.frame of drives
## Not run: sample_drives(2, 25, dt) ## End(Not run)## Not run: sample_drives(2, 25, dt) ## End(Not run)
This function will return a sample play from the nflscrapR play-by-play data for a given down, distance, yards from the team's goal, using a given strategy on fourth down. The strategies are: empirical, always going for it on fourth down, never going for it on fourth down, go for it if one is less than a certain distance from a first down/touchdown, and go for it if it maximizes one's expected points. This should be primarily used within the 'NFLSimulatoR::sample_play()' function.
sample_fourth_down_strategy( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, fourth_down_strategy = "empirical", yards_less_than = 5 )sample_fourth_down_strategy( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, fourth_down_strategy = "empirical", yards_less_than = 5 )
what_down |
The current down (1st, 2nd, 3rd, or 4th down) |
yards_to_go |
Number of yards to go until a first down or TD |
yards_from_own_goal |
The number of yards from the possession team's own goal |
window_yards_from_own_goal |
Precision parameter for "yards_from_own_goal" (a value of 1 means the sampling will occur within plus or minus 1 of the "yards_from_own_goal" value) |
play_by_play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
fourth_down_strategy |
The specific fourth down strategy 'empirical', 'always_go_for_it', 'never_go_for_it','yds_less_than','exp_pts' |
yards_less_than |
Parameter for 'yds_less_than' strategy. If using 'yds_less_than' strategy and one is less than 'yards_less_than' yards from first down/touchdown, then go for it on fourth down |
A tibble containing lots of info
## Not run: sample_fourth_down_strategy(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, window_yards_from_own_goal = 2, play_by_play_data = reg_pbp_2018, fourth_down_strategy = "empirical") ## End(Not run)## Not run: sample_fourth_down_strategy(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, window_yards_from_own_goal = 2, play_by_play_data = reg_pbp_2018, fourth_down_strategy = "empirical") ## End(Not run)
This function will return a sample play from the nflscrapR play-by-play data for a given down, distance, yards from the team's goal, using a given pass/rush play strategy. The user may choose a value for the proportion of passing plays to be sampled. Thus one can test strategies in which the team always passes, always runs, or some distribution of the two. This strategy is only intended for downs 1 - 3, and uses an empirical strategy for fourth downs. This should be primarily used within the 'NFLSimulatoR::sample_play()' function.
sample_passes_rushes_strategy( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, prop_passes = 0.5 )sample_passes_rushes_strategy( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, prop_passes = 0.5 )
what_down |
The current down (1st, 2nd, 3rd, or 4th down) |
yards_to_go |
Number of yards to go until a first down or TD |
yards_from_own_goal |
The number of yards from the possession team's own goal |
window_yards_from_own_goal |
Precision parameter for "yards_from_own_goal" (a value of 1 means the sampling will occur within plus or minus 1 of the "yards_from_own_goal" value) |
play_by_play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
prop_passes |
Proportion of plays that should be pass plays, between 0 and 1, inclusive |
A tibble containing lots of info
## Not run: sample_passes_rushes_strategy(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, window_yards_from_own_goal = 2, play_by_play_data = reg_pbp_2018, prop_passes = 0.5) ## End(Not run)## Not run: sample_passes_rushes_strategy(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, window_yards_from_own_goal = 2, play_by_play_data = reg_pbp_2018, prop_passes = 0.5) ## End(Not run)
This function will return a sample play from the nflscrapR play-by-play data for a given down, distance, yards from the team's goal, using the usual NFL-coaching strategy.
sample_play( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, strategy = "normal", ... )sample_play( what_down, yards_to_go, yards_from_own_goal, window_yards_from_own_goal = 1, play_by_play_data, strategy = "normal", ... )
what_down |
The current down (1st, 2nd, 3rd, or 4th down) |
yards_to_go |
Number of yards to go until a first down or TD |
yards_from_own_goal |
The number of yards from the possession team's own goal |
window_yards_from_own_goal |
Precision parameter for "yards_from_own_goal" (a value of 1 means the sampling will occur within plus or minus 1 of the "yards_from_own_goal" value) |
play_by_play_data |
A data file from nflscrapR prepped using the prep_pbp_data.R function |
strategy |
A string describing the strategy to be used, default is "normal", others include: "fourth_downs" and "passes_rushes" which implement some strategy regarding 4th downs and proportion of plays that are passing plays, respectively. |
... |
Additional arguments for different strategies |
A tibble containing lots of info
## Not run: sample_play_test(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, play_by_play_data = pbp_data, strategy = "normal") ## End(Not run)## Not run: sample_play_test(what_down = 3, yards_to_go = 2, yards_from_own_goal = 45, play_by_play_data = pbp_data, strategy = "normal") ## End(Not run)