15. Can recruitment models reliably compare players across leagues and identify role fit?

Require evidence for the model’s specific prediction and target population before relying on cross-league rankings. TransferLab, for example, advertises custom metric profiles and a club-strength adjustment model; that establishes a product claim, not validated accuracy for a particular MLS or NWSL recruitment question. (analyticsfc.co.uk)

Ask the supplier to distinguish:

• Describing past performance.
• Finding stylistically similar players.
• Adjusting for team or competition strength.
• Forecasting future performance after a move.
• Estimating market or salary value.

These should be tested as different tasks.

A proposed validation pack should include:

1. A stated outcome: what exactly is predicted, over what period?
2. A time-respecting test: only information available at the historical decision date.
3. Relevant populations: role, age, gender and source/destination markets.
4. A simple baseline: does the model improve on an agreed uncomplicated comparison?
5. Uncertainty: small samples, missing history and unsupported market combinations.
6. Failure cases: examples where the model was materially wrong.
7. Selection limits: which players are absent from the evaluation, and why?

For women’s recruitment, inspect the actual reference population and metric thresholds. SkillCorner announced dedicated women’s physical metrics on 11 June 2026, describing recalibrated thresholds and methodology. That is a concrete reason to ask for dataset and metric versions rather than assuming every physical comparison uses the same basis. (skillcorner.com)

Treat a role-fit score as evidence to investigate. Do not allow it to replace the football explanation of what the player must do in your team.