Your AI football analyst.
Grounded in real data.
Ask a grounded football analyst any question, or open a full player or team report — position-aware, sample-checked, built directly from updated data across the Premier League, La Liga and Egyptian Premier League.
Football intelligence, grounded in real data.
An AI analyst you can talk to, and a team-intelligence engine underneath it — every answer traces back to real event data, never a guess.
AI Analyst
A grounded football analyst you can talk to — retrieves real numbers from the event database first, then explains what they mean tactically. Never invents a stat.
Team Intelligence
Weighted spatial maps, pressing behaviour, ball-loss zones and league-contextual style archetypes — the mechanism behind the results.
Style-Space Similarity
Find comparable players and teams by playing style, not raw counts — position-relative, with the archetype and the "why" behind every match.
Seven leagues on the roadmap — three live today.
The same intelligence engine runs across every competition. Premier League, La Liga and the Egyptian Premier League are fully loaded now; Serie A, Bundesliga, Ligue 1 and the Saudi Pro League are next.
Real reports, built from real data.
Every card opens a full, on-demand intelligence report — position-aware maps, Vision AI scoring, percentiles vs peers. Generated live from event data. Scroll for more.
Player Reports
scroll →Notes from building the model.
How the analysis actually works — methodology, mistakes we caught, and what changed.
خمسة أسماء تحت 23 سنة من الدوري المصري الممتاز تستحق المتابعة
خمسة أسماء تحت 23 سنة من الدوري المصري الممتاز تستحق المتابعة
2026-08How Vision AI Actually Rates a Player
No fixed weights, no single magic formula nobody can explain. Here's exactly how the Vision AI Score is built — one tactical question at a time — and why it says "not enough data yet" instead of guessing.
2026-08Why We Rebuilt Player Similarity From Scratch
Raw-metric clustering let Federico Valverde and Frenkie de Jong land in the same bucket. That's not a fluke — it's what happens when you cluster event counts instead of playing style.
2026-07One player, one breakdown — analysis, video and data together.
Hand-picked players get their own page: written analysis, match video and photos, a quick visual dashboard, and a link to their full data report — all in one place.
Tactical detection for coaches and analysts — read from every moment of the match.
The work an analyst does by hand — watching a match back, finding the recurring patterns, clipping the evidence — done across the whole game automatically. Every phase of play is checked, so nothing is missed because there was not time to reach it.
- ▸For coaches and analysts — how an opponent builds, which side they overload and where their block breaks first, with the whole match tagged automatically rather than the handful of moments there was time to clip by hand.
- ▸Nothing missed — every phase of play is examined, so a pattern is caught whether it happens once or twenty times.
- ▸Checkable — each detection names the players, the minute and a confidence score, so any claim can be traced back to what actually happened.