01
Curated data
A structured dataset is used to test the full player-comparison pipeline with consistency and traceability in mind.
The Performance Lab
Gate 02Player intelligence
See the playerbehind the player.
The Performance Lab turns football data into useful questions. Start by finding players who perform the same role in a similar way.
Start the experiment
The active experiment
Player similarity.
Search for a player, then compare their role-specific statistical profile against the right peers.
Input
Choose a player
Search for a player...
Peer scope
Role + season
Minimum sample
500 mins
Engine
Cosine score
The console is ready
Pick a player. Find their footballing relatives.
The tool compares players inside the right role and season context, rather than chasing surface-level statistics.
Try: Leão · Pulisic · Scamacca · Koopmeiners
How it works
Data should make the game clearer.
This is a starting point, not a final verdict. The value is in the question that follows the comparison.
02
Role context
Statistics become percentiles within a player's role and season, avoiding comparisons that flatten football into one position.
03
Similarity score
The engine compares those role-based profiles using cosine similarity, then ranks the closest matches.
The Director's Office
Have an experimentin mind?
Bring a football question, a data idea or a collaboration worth exploring. We can build something more useful together.
Bring it to the table