nflpredictr 
The nflpredictr R package provides a lightweight R client for
retrieving predictions for upcoming (and past) NFL games via the NFL
Win Predictor API. Details on the
model used to make the predictions can be found
here
nflpredictr provides four main functionalities:
- Retrieves new data to use as the inputs into a ML model trained to predict the outcomes of NFL games
- Provide the new data to the API and return predictions
- Visualize results.
- Retrieve Las Vegas Odds and lines to compare with the predictions
generated from
nflpredictr
Installation
You can install the development version of nflpredictr from
GitHub with:
# install.packages("devtools")
devtools::install_github("anguswg-ucsb/nflpredictr")Example
nflpredictr provides utility functions for accessing game
predictions from a Logistic Regression ML model that was trained on ~20
years of historic NFL data and correctly predicts the outcomes of games
~69% of the
time.
If no inputs are given to predict_games(), the default behavior is
to make a prediction for week 2 of the 2021 NFL season.
# Load package
library(nflpredictr)
# Make an API request using predict_games(), default week is the upcoming week of the season
nflpredictr::predict_games(
year = 2021,
week = 2
)
#>
#>
#> Sending request to nflwinpredictor API...
#> Request URL:
#> http://68.183.25.9:8000/predict-new-data?year=2021&pred_week=2
#> # A tibble: 16 × 8
#> season week game_id home_team away_team win home_win_prob away_w…¹
#> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl>
#> 1 2021 2 2021_02_SF_PHI PHI SF 1 0.699 0.301
#> 2 2021 2 2021_02_TEN_SEA SEA TEN 1 0.694 0.306
#> 3 2021 2 2021_02_MIN_ARI ARI MIN 1 0.692 0.308
#> 4 2021 2 2021_02_ATL_TB TB ATL 1 0.644 0.356
#> 5 2021 2 2021_02_LV_PIT PIT LV 1 0.618 0.382
#> 6 2021 2 2021_02_DAL_LAC LAC DAL 1 0.572 0.428
#> 7 2021 2 2021_02_BUF_MIA MIA BUF 1 0.544 0.456
#> 8 2021 2 2021_02_KC_BAL BAL KC 1 0.540 0.460
#> 9 2021 2 2021_02_HOU_CLE CLE HOU 1 0.509 0.491
#> 10 2021 2 2021_02_NYG_WAS WAS NYG 0 0.498 0.502
#> 11 2021 2 2021_02_NO_CAR CAR NO 0 0.471 0.529
#> 12 2021 2 2021_02_NE_NYJ NYJ NE 0 0.449 0.551
#> 13 2021 2 2021_02_CIN_CHI CHI CIN 0 0.443 0.557
#> 14 2021 2 2021_02_DEN_JAX JAX DEN 0 0.417 0.583
#> 15 2021 2 2021_02_LA_IND IND LA 0 0.413 0.587
#> 16 2021 2 2021_02_DET_GB GB DET 0 0.312 0.688
#> # … with abbreviated variable name ¹away_win_probMake predictions on past games
Predictions can also be requested for weeks of past NFL seasons, going as far back as the 2016 season.
# Make an API request using predict_games() for a specific year and week
nflpredictr::predict_games(
year = 2018,
week = 8
)
#>
#>
#> Sending request to nflwinpredictor API...
#> Request URL:
#> http://68.183.25.9:8000/predict-new-data?year=2018&pred_week=8
#> # A tibble: 14 × 8
#> season week game_id home_team away_team win home_win_prob away_w…¹
#> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl>
#> 1 2018 8 2018_08_GB_LA LA GB 1 0.758 0.242
#> 2 2018 8 2018_08_CLE_PIT PIT CLE 1 0.729 0.271
#> 3 2018 8 2018_08_NYJ_CHI CHI NYJ 1 0.683 0.317
#> 4 2018 8 2018_08_DEN_KC KC DEN 1 0.650 0.350
#> 5 2018 8 2018_08_MIA_HOU HOU MIA 1 0.639 0.361
#> 6 2018 8 2018_08_SF_ARI ARI SF 1 0.563 0.437
#> 7 2018 8 2018_08_TB_CIN CIN TB 1 0.561 0.439
#> 8 2018 8 2018_08_BAL_CAR CAR BAL 1 0.555 0.445
#> 9 2018 8 2018_08_SEA_DET DET SEA 1 0.552 0.448
#> 10 2018 8 2018_08_NO_MIN MIN NO 1 0.524 0.476
#> 11 2018 8 2018_08_PHI_JAX JAX PHI 0 0.440 0.560
#> 12 2018 8 2018_08_WAS_NYG NYG WAS 0 0.406 0.594
#> 13 2018 8 2018_08_IND_LV LV IND 0 0.397 0.603
#> 14 2018 8 2018_08_NE_BUF BUF NE 0 0.317 0.683
#> # … with abbreviated variable name ¹away_win_probMake a tile plot of win probabilities
nflpredictr contains a few functions for quick plotting the outputs
from predict_games(). The plot_tile() function will plot
each match up side by side on a tile plot and color code teams by their
predicted win probability from the NFL Win Prediction API.
# Plot the outputs from predict_games()
tile_plot <- nflpredictr::plot_tile(
predictions = nflpredictr::predict_games(
year = 2021,
week = 5
)
)
Plot the teams favored to win that week
The plot_favored() function will plot the teams favored by the NFL Win
Prediction API model in order of win probability.
# Plot the outputs from predict_games()
fav_plot <- nflpredictr::plot_favored(
predictions = nflpredictr::predict_games(
year = 2021,
week = 5
),
prob_alpha = FALSE
)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent