Sports Analytics Weekly by kubeia.io - 40/2026
- Week 40
- Week 39
- Week 38
- Week 37
- Week 36
- Week 35
- Week 32
- Week 31
- Week 30
- Week 29
- Week 25
- Week 24
- Week 23
- Week 22
- Week 21
- Week 20
- Week 19
- Week 17
- Week 16
- Week 15
- Week 13
- Week 12
- Week 11
- Week 10
- Week 9
- Week 8
- Week 7
- Week 6
- Week 5
- Week 3
- Week 2
📝 Sports Analytics
It Is the Era of Pressing, We Need Better Metrics
substack.com - Alex Marin Felices
A Tracking-Based Framework Using Player Position, Speed, Direction and Time-to-Intercept to Quantify Defensive Pressure at Every Moment.
Player Performance Is Noisy, Bayesian Models Can Help
substack.com - Alex Marin Felices
A Bayesian study shows why availability, missing data and uncertainty can matter as much as the player rating itself
🤖 Machine Learning
TabPFN vs XGBoost: benchmark measured on an RTX 4070 Ti
efraingaray.com - Efrain Garay
The claim behind TabPFN and TabICL is that they predict on a table without ever training on it and still beat tuned boosting. I measured it on fourteen datasets from the Grinsztajn benchmark, with the same split and the same clock for everyone. The one that does not train wins, the advantage holds up to 32,000 rows instead of breaking, and the most-cited model can no longer be downloaded without an account.
NVIDIA Kumo Tabular Sets a New Accuracy-Efficiency Frontier for Tabular Prediction
huggingface.co - Jingang Qu, Valter Hudovernik...
NVIDIA Kumo Tabular, part of the NVIDIA Kumo Structured model collection, is an open foundation model for tabular data now available on Hugging Face. Given a table of labeled rows, it predicts the labels of new rows in a single forward pass, with no training, no tuning, and no feature engineering, for both classification and regression. It was pretrained only on artificial data, comes in three sizes (28M to 215M parameters), runs through our open-source library, and is released under the OpenMDW-1.1 license for commercial use. It ranks first on the four benchmarks TabArena, BeyondArena, TALENT and ScoringBench.
🌩 Forecasting
You Should Always Go For It
argmin.net - Ben Recht
Football season is well underway, and my favorite absurd sports-analytical caricature, Seth Walder, has been tweeting out his hilarious takes. Walder is an ESPN NFL analyst who built a model that he claims can predict the probability of winning a game for every play to three digits of precision.
Though he maintains his sauce must remain secret, Walder claims he can determine the optimal move for any NFL play at any point in any game. His sophisticated model accounts for all aspects of the game, including teams’ strengths, the nuances of route trees, and the fluid dynamics of the air. Apparently, by looking at vast amounts of historical data and using the most cutting-edge probabilistic calculations, he can compute the probability that a team wins given their next play call.