In this speaker talk, we'll learn how Machine Learning and statistics are used to model how well a basketball player is likely to perform based off of individual and team performance.

DARKO is a statistical metric that combines bayesian inference with machine learning to develop game-by-game estimates of how well a player is likely to perform in the future.

It's considered the most useful player metric that's publicly available, and is used by NBA front offices across the league.

In this talk we'll hear from its creator about how it's derived, the details of how it learns from data, as well as design principles that need to be considered when incorporating Machine Learning and analytics for sports.

This is a fantastic opportunity for anyone who has an interest in sports analytics, or who wants to see first hand how analytics has to be crafted to fit very particular business use cases.