A set of functions to easily access Football Outsiders data, merge-ready NFL names and a wrapper for nflreadr::load_rosters()

Installation

if (!require("remotes")) install.packages("remotes")
remotes::install_github("SCasanova/personalfix")

Functions

Football Outsiders Data

foutsiders_data() This function accepts page name, season, username and password to scrape from Football Outsiders. Current supported pages are:

• "team-offense"
• "team-defense"
• "special-teams"
• "qb"
• "wr"
• "rb"
• "te"
• "basic-offensive-line"
• "basic-defensive-line"

Example:

foutsiders_data("team-offense", 2020, "username", "password")

# A tibble: 32 × 10
   Team  `Total DVOA` `Weighted DVOA` `Pass DVOA` `Rush DVOA`
   <chr>        <dbl>           <dbl>       <dbl>       <dbl>
 1 GB           0.291           0.316       0.52        0.029
 2 KC           0.239           0.234       0.49       -0.057
 3 TB           0.198           0.246       0.371      -0.02 
 4 TEN          0.184           0.191       0.395       0.061
 5 BUF          0.156           0.178       0.433      -0.151
 6 SEA          0.137           0.095       0.304      -0.014
 7 NO           0.107           0.094       0.166       0.095
 8 MIN          0.063           0.06        0.188       0.007
 9 CLE          0.054           0.081       0.209      -0.008
10 LAR          0.044          -0.048       0.124       0.035
# … with 22 more rows, and 5 more variables:
#   Unadj. Total VOA <dbl>, Unadj. Pass VOA <dbl>,
#   Unadj. Rush VOA <dbl>, Variance <dbl>, Schedule <dbl>

Adjust Numeric Formats

fix_num() This function will take a vector, detect its format (character, numeric or percentage) and will adjust to numeric when possible

Examples:

fix_num("20%")
[1] 0.2

fix_num("20")
[1] 20

fix_num("twenty")
[1] "twenty"

fix_num("2,543")
[1] 2543

df %>% dplyr::mutate(dplyr::across(dplyr::everything(), fix_num))

Scale/Standardize variables

standardize() This function will take a vector and standardize it with a mean of 0 and a standard deviation of 1. Wrapper for scales() with a nicer output.

Examples:

standardize(c(12,342,54,22,2,453))
[1] -0.6858083  0.9844260 -0.4732330 -0.6351952 -0.7364215
[6]  1.5462320

mean(standardize(c(12,342,54,22,2,453)))
[1] 9.233789e-18

sd(standardize(c(12,342,54,22,2,453)))
[1] 1

df %>% dplyr::mutate(dplyr::across(col1:col24, standardize))

Merge-ready NFL Names

name_key() This functions takes arguments name, arg_1, arg_2 and draft_year (easily obtainable from clean_rosters()) where arg_* are modular and can be anything the user wants. The output is a merge-ready name for situations where IDs are not available. It also uses ffscrapr's mismatched names database to fix name variations. Additional arguments are recommended to be draft round, draft team, or position to avoid inconsistencies across sources or years.

Examples:

name_key(c('Zeke Elliott', 'Josh Jacobs'), c(1,1), c('RB','RB'), c(2017, 2019))
[1] "ezeelliott171RB" "josjacobs191RB" 

name_key(c('Chatarius Atwell', 'Sean Bunting'), c('LAR','TB'))
[1] "tutatwellLAR"    "seamurphy-buntingTB"

df %>% dplyr::mutate(merge_name = name_key(name, draft_team, position, draft_year))

Modified Rosters

clean_rosters() This function is a wrapper for nflreadr::load_rosters() and outputs all headshot urls (including 2021 rookies) with https: prefix and integrates FB, HB and RB, into RB. Also includes draft data from ffscrapr::dp_playerids()