The sixteenth issue of Unexpected Points Added!
Please do me a favor and share it with at least one person and one pet this week.
All over the place this week. We look at route statistics for the best NFL TE ever (?), review the components of expected goals in soccer, learn how shot tracking fixes NBA shots, refute NHL momentum, and look at a regression model for R.
This Week's Lineup
Travis Kelce 2020 Route Statistics
In 2020, Kelce became the first TE to lead all receivers in DYAR (adds up total player performance on a per play basis for the entire year), and this link looks at which route patterns were most successful.
Past Goals to Create Future Goals in Soccer
Breaks down all the components of the improved xG (expected goals) model. Video included too.
Shot-Tracking Is Changing The Way Basketball Players Fix Their Game
By using Noahlytics Data Service, and their high-quality motion tracking cameras during practice, teams can analyze players' shots to identify what needs to be fixed.
NHL Momentum? Think Again.
This research paper applies entropy as an unbiased measure to further refute the idea of momentum in sports. Specifically, it looks at the 2020 St. Louis Blues and their improbable run to the Stanley Cup. Ultimately, game outcomes are not dependent on previous games’ outcomes and conclude that the theory of momentum across the season is a fallacy that should not affect behavior.
Getting predictions from an isotonic regression model
Pass the output of the isoreg function to as.stepfun to make an isotonic regression model into a black box object that takes in uncalibrated predictions and outputs calibrated ones.
Was this email forwarded to you? Sign up here >>
Unexpected Points Added is curated and maintained by Patrick Hayes. If you have questions or suggestions for the newsletter, just reply to this email. I answer every single one.