Projects

nflverse#

Contributor to the open-source NFL data package family nflverse. Found obscure API endpoints and built scripts to hit these endpoints, download and clean the sourced data, and then make the data available via the {nflreadr} package. You can view the nflverse family of packages here .

Predicting Scrabble Point Values#

Finished 1st overall in a Kaggle community competition to predict the point value of the 20th turn of a Scrabble game given only the 19 previous turns. Relied on a neural network to model Scrabble turns, then adjusted point values using a mixed model to account for individual player skill. You can view my writeup of the project here .

Kaggle 2021 March Mania Contest#

Finished 27th overall in Kaggle’s 2021 March Mania contest relying on a variety of team-strength metrics, domain knowledge, and other features to predict performance in the 2021 Men’s Basketball National Championship. You can view my code for this project here .

Big Data Bowl 2020#

Worked with a team to evaluate the pass-defense of individual players using NFL tracking data, creating an R package to estimate and evaluate defensive skills of specific NFL players. Our writeup for the project can be viewed here , our code can be viewed here , and a Shiny app to visualize our results can be viewed here .

NERD Score Twitter Bot#

Put together a Twitter bot to produce daily NERD score values for pitchers and games (NERD estimates the entertainment factor for a given game or pitcher). The bot is previewed on FanGraphs here and the code can be found here , though the bot is no longer maintained or updated.