Is expected goals (xG) in neutral gamestates a better predictor of future performance than xG across all gamestates?
Quantifying how unlikely a player’s season-long shooting performance is, factoring in their prior shot history
Identifying where passes (even incomplete ones) have the most positive impact on the pitch.
All-play scores, optimal lineup choices, and more
Finnicking around (with an xG model) and finding out
Calculating expected goal difference (xGD) with respect to game state, using FBref data.
Using calibration to improve classifier model performance
To what extent does a state differentiate between players? How much does the stat’s value change over time?
Using empirical Bayes and the Gamma-Poisson conjugate pair
Evaluating Opta’s xG model performance with Brier skill score (BSS) and calibration curves
Evaluating how we can use match outcome probabilites for season-long insights
Calculating and comparing expected points from different expected goals sources
Quantifying soccer pass networks with weighted maximum cuts
An Alternative to the Classic Approach to Dimension Reduction + Clustering
With Atomic VAEP
Brute Force Programming Go Brrr
With data.table, reticulate, and spatstat
With S3 Classes and vctrs
A Newbie’s Guide to Making A Pull Request (for an R package)
