This is the github repository for the corresponding submission to the NFL Big Data Bowl 2024 competition on Kaggle. The full notebook can be viewed here.
Sadly, we narrowly missed a top-ten ranking, but we received an honorable mention as the only team from Europe.
In this contribution, we developed the metric PEP for quantifying the value of tackles. It allows practitioners to assess players, particularly in terms of their tackling abilities. Our approach allows for within-play conditional density estimation of the end-of-play yard line which serves as a basis for the evaluation of tackle performance measured by prevented expected points by artificially removing the tackler from the data. Importantly, our method incorporates distributional information, i.e., heteroscedasticity and multimodality, which would be lost when solely relying on point predictions. Therefore, the uncertainty can propagate to the level of expected points, leading to an accurate quantification of expected points prevented by the tackle.
In the prg directory all necessary code to reproduce our work is
found. Due to size restriction, we are not able to upload our random
forest model for the conditional density estimation. The necessary
scripts for the random forest are the preprocessing_new.R and the
train_eop_model_new.R files. To calculate our PEP metric, the ep model
uploaded in the models folder can be used (the code from the
ep_models.R script can also be used to create your own model).
Finally, the calc_tackle_value.R file can be used to obtain PEP values
for each tackle.