TransferGPT

How it works

How TransferGPT evaluates players

Every answer is a scouting verdict built on data. Here is what goes into it.

What we cover

More than 90 leagues. Transfer history and player scouting reports go back 10 years; match statistics cover the last three seasons. We combine public match data with professional scouting databases. Profiles and contracts are refreshed every week.

Player level

Each player is compared with players in the same role, league and season, per 90 minutes. Defensive actions are adjusted for how much of the ball the team has. League strength is built in: the best player in a smaller league is not rated the same as the best player in a bigger one.

Match data and scouting view

Match performance is combined with a scouting assessment of technical, mental and physical qualities. The more minutes a player has, the more the match data counts. For young players and players who have just changed league, the scouting view weighs more.

Club fit

For each club we look at the squad in that position: ages, contracts, injuries and playing time. We check whether the player would start, rotate or not get in, and whether the deal fits what the club usually pays in fees and wages.

Price

Our model's estimate is based on real fees paid for comparable players: same age band, position and level within their league. It is adjusted for contract length, so a player whose contract ends soon costs less. It is not a market value.

Wages

Always shown per year. Taken from club financial reports where available, otherwise estimated.

What it is not

Not a rating to copy, but a verdict with reasons. Our figures are estimates, not guarantees, and recent transfers or newly covered leagues can lag by a few days.