Price is not value. One is the odds, the other the wager.
The Value Engine takes the finalized footballing read from upstream and turns it into money: a market value, a value to a specific club, a wage, and a deal shape, each carried as a bet with its confidence. It never re-evaluates the player. It wraps the locked read in the structure of the transfer market and prices it, and it holds one hard distinction at the center: the market value is the engine's estimate of what the open market would pay, while the fee is the realized price of an actual deal, and the two are never merged. No value is ever a bare number, because a player has no single value at all, only a value in a context, priced with a band. It makes the odds. The human makes the wager.
The value is what the bet should be. The fee is the bet that got made.
Two numbers sit at the center of every deal, and the engine refuses to merge them. The market value is its estimate of what the open market would pay, reported as a band. The fee is the realized price of one actual transaction, set partly by forces outside the player, leverage, timing, a buyer's need, so it lands somewhere inside the band rather than on its midpoint, and the engine names the gap.
This is the same move the whole engine makes, the read and the wager kept apart, applied to money: the value is the honest estimate of what the bet should be, the fee is the one a club actually placed under the pressures of a real window. Treating the last fee paid as the truth is how markets overpay in cascades; the engine anchors to the value and reads the fee as a separate object. The value is the read. The fee is the deal. Never the same number.
Illustrative engine read on the real price-is-not-value doctrine (market value as the band, the fee as the realized price landing inside it, above = overpay and below = a buy, the gap named). Composite player, demonstration figures.
Every value carries its confidence, because a value without a band is a false precision.
Confidence is the unit here exactly as it is for the rating. No value is a bare figure; each carries the band its confidence implies, and the band is not a courtesy, it is empirically forced, because even strong market-value models leave most of the variance unexplained and break hardest on the highest-profile players. A high value at low confidence and a lower value at high confidence are different objects.
The read is the spine and the layers shape it, but they never override the comparable anchor on a hunch, and where the comparable set is thin the band widens and the engine says so. A believable 41M at high confidence is worth more to a real decision than an impressive 55M that is mostly a wide band off a single run. Believable over impressive holds for money exactly as it holds for the rating.
Illustrative engine read on the real market-value structure (the comparable-sale anchor as the spine, the five layers of age, contract, market context, marketability, and the confidence band, high-value-low-confidence versus lower-value-high-confidence as different objects, a recency spike flagged and mean-reverting). Composite players, demonstration figures.
The same quality is a different value in every context. The engine prices into one.
The engine is weightless here too: it never declares what a player is worth in the abstract, because that number does not exist. The same read is a different value under a different league economy, a different tax regime, and a different buyer, so the engine prices a player into a specific context, lays out how the value moves as the context moves, and then stops. The human supplies the context and chooses the deal.
This is why there is no single sticker price: the engine prices the player into a club, a league, a tax jurisdiction, and a deal shape, and reports the priced structure with its confidence, showing how the value slides as the context slides. Market Value and Value-to-Club are kept as two separate objects because a marquee club's commercial return can justify a price a smaller club could never recoup. The engine makes the odds. It hands the human the context and the choice. Price the player into a context. Then let the human make the wager.
Illustrative engine read on the real weightless doctrine (no single global value, the same read priced into different league economies, tax regimes, and buyers) and the two value objects (Market Value versus Value-to-Club), read-only on the KR. Composite player and contexts, demonstration figures.
The Value Engine wraps a locked footballing read in the structure of the transfer market and prices it, never re-evaluating the player, and it holds one distinction at the center: the market value is the engine's estimate of what the open market would pay, the fee is the realized price of one actual deal, and the two are never merged, because the fee carries leverage, timing, and need the value does not, and lands somewhere inside the band rather than on its center. Every value is a bet with a band, because even strong models leave most of the variance unexplained and a value without a band is a false precision, so a believable figure at high confidence beats an impressive one that is mostly width. And there is no single sticker price, because the same read is a different value under a different league economy, a different tax regime, and a different buyer, so the engine prices the player into a context, reports how the value moves as the context moves, and then stops. It makes the odds. The human supplies the context and makes the wager.
Make the odds. Let the human make the wager.
Market Value prices the locked read into a context, reports it as a band with its confidence, keeps the value and the fee as separate objects, distinguishes Market Value from Value-to-Club, and never re-rates the player.