Sports Analytics Weekly by kubeia.io - 3/2026

The tools bookmakers use to block data-savvy gamblers, and how to get round them

Every season, clubs produce thousands of pages of match reports.

Shots, possession, passes, duels, sprints, xG, field tilt — all carefully tracked, ranked, compared.

Yet the main question remains: which of these actually matter for winning football matches?

This paper tackles that question head-on by doing something most analyses avoid:

throwing almost everything into the model and letting the data decide.

Using player-conditioned GPT models to evaluate transfer fit through counterfactual match sequences.

Goalkeeper evaluation has always been one of the thorniest challenges in soccer analytics. The position is defined by small sample sizes, high variance, and context-driven outcomes. In Major League Soccer, where roster rules magnify the impact of every marginal dollar, understanding goalkeeper value is especially important. With clubs often operating near budget ceilings, a single overperformance or underperformance in goal can shift playoff probability, alter roster-building timelines, or change the financial implications of a season.

However, PFNs are not just one more algorithm in scikit-learn, but they turn the way we model tabular data upside down.

It’s time we talk about why tabular ML is (maybe) getting weird.

How many times do I need to see something to believe it?