Talk to the data. Ask any football question and get an intelligent answer — grounded in live match stats, recent results, and the publication's tactical corpus. Built on an LLM with genuine context, not a chat wrapper.
Most football stat sites show you leaderboards. Most football AI is a chatbot with no idea what a half-space is. The Lab sits in between — fed by live data and the publication's own editorial archive. Here's what reader conversations will look like:
Bukayo Saka (Arsenal) is 4.8 goals above his 9.2 xG — the league's largest over-performance among players with 8+ goals. Context: Saka's shot profile has shifted outside the box compared to last season, meaning he's finishing harder chances rather than tap-ins. Historically this pattern regresses, but his underlying creation (xA) is also up 22%, so the finishing gains aren't entirely random.
Arsenal used a 4-3-3 with Rice pushed higher and Ødegaard leading a vertical press on Real's build-up. Their first line (Martinelli-Havertz-Saka) was 4.6m narrower than their league average — forcing Real's CBs wide onto the fullbacks, who were then pressed by Arsenal's own fullbacks stepping up. Real's pass completion in their own third dropped to 68% (their season average is 87%).
Sevilla have dropped 11 points from winning positions since January 1 — the most in La Liga. Six of those 11 came in the final 15 minutes, suggesting a late-game game-management issue rather than bad luck. Their opponents' xG concedes in the 75+ minute window is 0.42 higher than in the first 75 minutes combined.
A non-obvious statistic from the most recent round of matches, with the tactical context that explains it. Automated generation, human editorial voice.
Bayer Leverkusen have attempted 118 passes into the final third per 90 this season — the highest of any side in Europe's top 5 leagues.
Xabi Alonso's successor has kept the vertical identity; what's changed is the source. Under Alonso, 42% of those passes came from Wirtz. Without him: Schick and Grimaldo account for 37% between them, and the team is less reliant on a single number-ten.
Liverpool have conceded 0.68 xG per 90 from opposition counter-attacks — the lowest in the league, down from 1.24 last season.
The change isn't personnel (same back four). It's structural: the rest-defence shape has 0.4 fewer players ahead of the ball per sequence. They're attacking with fewer bodies, and it's showing in transition.
The Lab is rolling out in phases. Here's the full picture of what it will be:
The Lab opens to readers in Phase 2 (live data integration). Drop your email to get access the day it goes live.