|
Abstract:
|
Continuous-time assessments of game outcomes in sports have become increasingly common in the last decade. In the National Football League (NFL), only discrete-time estimates of play value were possible, since public NFL datasets were recorded at the play-by-play level. While measures such as expected points and win probability are useful for evaluating plays and game situations, there has been no research into how these values change throughout the course of a play. In this work, we make two main contributions: First, we introduce a general, modular framework for continuous-time within-play valuation using NFL player-tracking data, with several modular sub-models to easily incorporate recent player-tracking research. Second, we use a LSTM recurrent neural network to estimate how many yards the ball-carrier is expected to gain from their current position, conditional on the locations and trajectories of the ball-carrier, their teammates and opponents. Additionally, we demonstrate an extension with conditional density estimation so that the expectation of any measure of play value can be calculated in continuous-time, which was never before possible at such a granular level.
|
