Sports Analytics Weekly by kubeia.io - 47/2025
- Week 47
- Week 46
- Week 45
- Week 44
- Week 43
- Week 42
- Week 41
- Week 40
- Week 39
- Week 38
- Week 37
- Week 36
- Week 35
- Week 33
- Week 32
- Week 31
- Week 30
- Week 29
- Week 28
- Week 27
- Week 26
- Week 25
- Week 24
- Week 23
- Week 22
- Week 21
- Week 19
- Week 17
- Week 16
- Week 14
- Week 13
- Week 12
- Week 11
- Week 10
- Week 9
- Week 7
- Week 6
- Week 5
- Week 4
- Week 3
- Week 2
- Week 1
đź“ť Sports Analytics
The 12 Essential Reads from Year 1 of The xG Football Club
substack.com - Alex Marin Felices
As The xG Football Club turns one year old, this post brings together the 12 defining analytics pieces that shaped our understanding of the game this past season. Think of this edition as a technical map of where football analytics is heading — the models that advanced our thinking, the papers that challenged our assumptions, and the insights that bridged academic research with real recruitment and performance work.
Which xG Data Should You Trust?
pythonfootball.com - Martin
StatsBomb loves reminding us that “not all xG is created equal.”So I decided to put that to the test.Welcome to The Python Football Review
Inside the Mind of a Football Team
substack.com - Alex Marin Felices
The following summary critically reviews the research paper titled “Assessing Team Strategy using Spatiotemporal Data” by Patrick Lucey, Dean Oliver, Peter Carr, Joe Roth and Iain Matthews. This paper from 2013 motivated a shift from hand-labeled “what happened” event stats toward continuous spatiotemporal signals that can describe where and how actions unfold. It argues that while organizations now collect ball and player locations at each instant, these data are “continuous, high-dimensional and difficult to segment into semantic groups,” which makes strategy inference hard without labels to condition on. The authors framed this as the emerging space of “Sports Spatiotemporal Analytics” and proposed using ball-movement–based representations to characterize team behavior in football.
🤖 Machine Learning
The Forgotten Paper That Invented Gradient Boosting Before Gradient Boosting
medium.com - Valeriy Manokhin
When we talk about the evolution of boosting algorithms, the usual storyline jumps straight from AdaBoost to Friedman’s groundbreaking work on Gradient Boosting Machines (GBM). That narrative makes sense: Friedman’s perspective reshaped how we think about boosting and directly paved the way for modern tools like XGBoost, LightGBM, and CatBoost.But the real history is more nuanced — and far more interesting.
Time Series Forecasting Using Foundation Models
manning.com - Marco Peixeiro
You don’t need to spend weeks—or even months—coding and training your own models for time series forecasting. Time Series Forecasting Using Foundation Models shows you how to make accurate predictions using flexible pretrained models.In Time Series Forecasting Using Foundation Models you will discover:The inner workings of large time modelsZero-shot forecasting on custom datasetsFine-tuning foundation forecasting modelsEvaluating large time modelsTime Series Forecasting Using Foundation Models teaches you how to do efficient forecasting using powerful time series models that have already been pretrained on billions of data points. You’ll appreciate the hands-on examples that show you what you can accomplish with these amazing models. Along the way, you’ll learn how time series foundation models work, how to fine-tune them, and how to use them with your own data.