Signing Accuracy — Jupiler Pro League 2024/25

Club assessment — the universe of signings

The report considers exclusively genuine signings for the first-team squad, identified by the player's effective incorporation into the group between seasons. Excluded from the universe are: academy promotions, loan returns, internal administrative moves and transfers recorded but never completed. The aim is to measure the quality of market decisions — and those alone.

Player Impact Score (PIS)

Each evaluated signing receives a score from 0 to 100 that combines four observable technical dimensions: individual performance (average match ratings over the season), output (goals, assists and decisive actions adjusted for game time), collective performance (starting status and correlation with the club's results) and availability (volume of matches, adjusted for interruptions). High-impact players with reduced game time are promoted through an alternative efficiency criterion.

Outcome classification

The rating tiers are: Excellent and Good (successful signings — a full point in the index), Fair (intermediate performance — half a point) and Below expectations / Did not deliver (unsuccessful signings). Two categories sit outside the calculation, for distinct reasons: Injury (a serious injury prevented a fair assessment) and Investment (an under-23 player with low game time — the return depends on development and cannot be counted for or against the club in this cycle). Signings that did not reach relevant game time in the season are classified as unsuccessful, reflecting the sporting cost of the decision.

SigningLab — projection at the moment of signing

The SigningLab column shows the accuracy the club would have achieved had it followed the projections of the SigningLab model trained exclusively on the 2021-2023 seasons — therefore prior to all the signings evaluated here. The success projection is generated at the moment of signing — before the player takes the field for the new club — using exclusively information available up to that date: career history, physical-technical profile, adaptation pattern in similar leagues and the signing club's context. It is precisely because it is a forward-looking projection, and not a retrospective reading, that the model's accuracy holds such value: it measures the ability to predict the outcome of a market decision before that decision produces any effect.

The model is trained on more than 2 million individual matches and 423 thousand transfers across a decade, with strict temporal validation. Fair-performance assessments count as a hit in either direction of the recommendation.

A note on scope

This is SigningLab's public, simplified model, developed for transparent retrospective market assessment. SigningLab's commercial solutions for partner clubs use a substantially larger set of variables and dedicated predictive mechanisms, and are not fully reflected in this document.