NFLTables.jl
Functions to access NFL data. Please create a GitHub issue to request a new dataset or report a bug.
NFLTables provides enumerations and functions to make it convenient to access existing NFL data.
Where relevant, variable, module, and function names follow the conventions for the julia language.
julia> using NFLTables
julia> varinfo(NFLTables)
name size summary
–––––––––– ––––––––––– ––––––––––
NFLFastR 23.407 KiB Module
NFLScrapR 26.434 KiB Module
NFLTables 273.709 KiB Module
POST 4 bytes SeasonPart
PRE 4 bytes SeasonPart
REG 4 bytes SeasonPart
Schedules 35.631 KiB Module
SeasonPart 128 bytes DataType
SuperBowl 128 bytes DataType
julia> SuperBowl(2021)
SB_LVI::SuperBowl = 2021
julia> instances(SeasonPart)
(PRE, REG, POST)
Data retrieved in the submodules is cached locally with Scratch. By convention, data can be downloaded again with redownload=true in the appropriate function.
NFL Schedules
Schedules are obtained by parsing data from www.nfl.com.
julia> scheduledf = Schedules.schedule(2019);
julia> size(scheduledf)
(332, 9)
julia> last(scheduledf, 5)
5×9 DataFrame
Row │ season seasonpart week date states home away homescore awayscore
│ Int64 String7… Int64 Date… String15… String3… String3… Int64 Int64
─────┼────────────────────────────────────────────────────────────────────────────────────────────
1 │ 2019 POST 2 2020-01-12 FINAL GB SEA 28 23
2 │ 2019 POST 2 2020-01-12 FINAL KC HOU 51 31
3 │ 2019 POST 3 2020-01-19 FINAL KC TEN 35 24
4 │ 2019 POST 3 2020-01-19 FINAL SF GB 37 20
5 │ 2019 POST 4 2020-02-02 FINAL KC SF 31 20
nflscrapR Data
The data is available at https://github.com/ryurko/nflscrapR-data for the 2009-2019 seasons.
Use NFLScrapR.getgamedata to retrieve game information as a dataframe.
julia> gamesdf = NFLScrapR.getgamedata(2018, REG);
julia> size(gamesdf)
(256, 10)
julia> names(gamesdf)
10-element Vector{String}:
"type"
"game_id"
"home_team"
"away_team"
"week"
"season"
"state_of_game"
"game_url"
"home_score"
"away_score"
julia> first(gamesdf, 5)
5×10 DataFrame
Row │ type game_id home_team away_team week season state_of_game game_url home_score away_score
│ String3… Int64 String3… String3… Int64 Int64 String7… String Int64 Int64
─────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
1 │ reg 2018090600 PHI ATL 1 2018 POST http://www.nfl.com/liveupdate/ga… 18 12
2 │ reg 2018090900 BAL BUF 1 2018 POST http://www.nfl.com/liveupdate/ga… 47 3
3 │ reg 2018090907 NYG JAX 1 2018 POST http://www.nfl.com/liveupdate/ga… 15 20
4 │ reg 2018090906 NO TB 1 2018 POST http://www.nfl.com/liveupdate/ga… 40 48
5 │ reg 2018090905 NE HOU 1 2018 POST http://www.nfl.com/liveupdate/ga… 27 20
Play-by-play information is also available; use the getplaydata function.
julia> pbpdf = NFLScrapR.getplaydata(2018, REG);
julia> size(pbpdf)
(45156, 256)
julia> names(pbpdf)[1:10]
10-element Vector{String}:
"play_id"
"game_id"
"home_team"
"away_team"
"posteam"
"posteam_type"
"defteam"
"side_of_field"
"yardline_100"
"game_date"
nflfastR Data
For data more recent than nflscrapR, use NFLFastR.
The data is hosted at https://github.com/nflverse/nflfastR, and the README describes the improvements over nflscrapR.
This is the recommended data for new analysis.
julia> fastdf = NFLFastR.getdata(2018);
julia> size(fastdf)
(47874, 372)
julia> fastdf[1:10, 1:6]
10×6 DataFrame
Row │ play_id game_id old_game_id home_team away_team season_type
│ Int64 String15… Int64 String3… String3… String7…
─────┼──────────────────────────────────────────────────────────────────────────
1 │ 1 2018_01_ATL_PHI 2018090600 PHI ATL REG
2 │ 37 2018_01_ATL_PHI 2018090600 PHI ATL REG
3 │ 52 2018_01_ATL_PHI 2018090600 PHI ATL REG
4 │ 75 2018_01_ATL_PHI 2018090600 PHI ATL REG
5 │ 104 2018_01_ATL_PHI 2018090600 PHI ATL REG
6 │ 125 2018_01_ATL_PHI 2018090600 PHI ATL REG
7 │ 146 2018_01_ATL_PHI 2018090600 PHI ATL REG
8 │ 168 2018_01_ATL_PHI 2018090600 PHI ATL REG
9 │ 190 2018_01_ATL_PHI 2018090600 PHI ATL REG
10 │ 214 2018_01_ATL_PHI 2018090600 PHI ATL REG