Valuation engine
xV Valuation Model
Forecasting a player’s market value two years forward.
A proprietary ensemble that projects where a player’s value is going - not just where it is - validated to 89.4% directional accuracy on young players.
What it does
xV takes a player’s development trajectory and projects their market value forward on a two-year horizon. It powers the ranked shortlists, squad valuations, and the value projections in every deep-dive report. The output is not a single number but a trajectory with a direction and a magnitude, validated against observable career outcomes - transfers completed, leagues reached, and minutes accumulated.
Used in: Position shortlists, squad valuations, player deep-dive projections.
Why it’s defensible
What makes this proprietary.
A trajectory model, not a snapshot
Most valuation tools price a player as they are today. xV is trained to predict change - the direction and size of a value move over two years. That is a fundamentally harder target, and it is where the competitive signal lives: identifying the player whose value is about to move before the market reprices them.
Calibrated per league, not pooled
A model trained on one market does not transfer cleanly to another. We calibrate independently across 21 leagues and publish the per-league accuracy so the reliability is visible market-by-market. Building and maintaining that per-league calibration is a years-long data and engineering investment, not a switch you flip.
An ensemble, not a single estimator
The production model is a proprietary ensemble of gradient-boosted estimators selected and weighted through extensive backtesting. The architecture is deliberately resistant to the failure mode of single-model approaches, which over-fit the leagues and eras they were trained on.
How we validate it
Every published number comes from strict train–test separation: the model is trained on transfers up to a cutoff date and scored on observable outcomes after it that it never saw - completed transfers, league-tier moves, minutes accumulated, and international call-ups. Directional accuracy is measured against what actually happened to the player, not against another set of estimates.
What we don’t publish
The feature engineering, the trajectory representation, the ensemble weighting, and the data pipeline behind xV are proprietary and not disclosed. What we publish is the evidence - the accuracy, the confidence intervals, and the per-league breakdown - so the result can be trusted without exposing how it is produced.
The underlying data and models are proprietary. We show the validated results, not the inputs that produce them.
See it applied to your shortlist.
Every report shows the model’s output with the per-league accuracy behind it.
Order a Report