Football stats still start with goals and assists. Those numbers travel badly across leagues. A 12-goal season in one league is not the same season in another. Live 0–10 ratings are useful for a single night. They are a poor way to rank a campaign across England, Spain, Germany, Italy, and France.
Baseball solved a version of this with WAR: compare the player to a replacement-level one, in context. NVC is that idea applied to the Big 5.
Replacement level here means a realistic fringe or rotation player, not a theoretical zero. The model takes attacking output, build-up value, defensive contribution, and availability, then adjusts for league strength, opponent quality, and team style. The result is scaled 0–100 so the table is readable.
That is why the current leaderboard does not look like a Ballon d’Or shortlist. Mbappé and Yamal at the top are unsurprising. The system is asking “how much did this player add over a replacement in his context?” not “who is most famous?”
NVC will miss things event data misses: leadership, pressing intelligence that isn’t tagged well, set-piece specialism, locker-room value. Use it as the quantitative layer, then watch the tape.