You do not draft the player he was in college. You draft the projection of who he becomes across the gate.
College and pro football are two different sports wearing the same name, joined by one hard gate with no return path, so a prospect's college number is not his pro number, they come from two engines that weight traits differently, and the gap between them is the projection. That projection is Mode 6, a separate, lower-confidence forecast run through the pro engine, where the rank order does not even have to hold. The board then reads each prospect by the slot he projects into, and prices the pick as the highest-surplus asset in the sport, the cost-controlled rookie deal.
Drafted once, from one closed system into another.
College and pro football are two different sports wearing the same name, different rules and a different developmental economy, joined by one hard draft gate with no return path. So his college number and his pro number are two different objects from two different engines, and neither is a discount of the other. The gap between them is the entire projection.
hard gate
The engine never shows a Mode 6 projection glued to the locked college number as co-equal: the college KR is present-tense value, the pro projection a forecast on a different engine. Treating the draft as a smooth continuation of a college career is the single modeling error the two-engine architecture exists to prevent. Two engines, one gate, a projection in the gap.
Illustrative on the real two-engine gate (the college engine as present-tense value and the pro engine as a projected value, the one-way one-time draft gate, the college and pro numbers as two distinct reads). Composite prospect, demonstration figures.
Run the traits through the pro engine, and the order can flip.
Mode 6 is the one operation that crosses the two engines, and it is the backbone of the draft. It is not a re-read of the same number at a different level, and it never multiplies a college KR by a coefficient, because there is none. It runs the prospect's traits through the pro engine's weights and produces a distinct, lower-confidence forecast.
The projection is on-demand and never modifies the college OVERALL (two different objects), and it is re-priced by where he lands (the developmental quality of the drafting team is a real term in the forecast, not a property of the player). Mode 6 also reports the bust-risk read and the developmental runway once he is in the pros. Do not translate the number, re-read the traits, and band the forecast honestly.
Illustrative on the real Mode 6 projection (the traits re-read through the pro engine into an Entry KR and a Peak, Floor, and Median, the rank-order flip, the translation decomposition and the wide position-conditioned bands widest at running back and quarterback). Composite prospects, demonstration figures.
The top buys the ceiling, the late rounds buy the role.
The engine determines the projected draft range and leads the output with the KR that matches it, because a top-five pick and a fifth-round pick are buying different things. Then it prices the pick, because a pick is priced not just for talent but for the cost-controlled surplus that talent delivers against its slotted cost.
A draft pick contains three auditable bets, each with its own confidence, so a miss is traceable to the term that broke rather than filed as a generic bust, and the same prospect is a different pick at a different team, because the drafting team's development and scheme fit re-price the projection. Lead with the number the slot is buying, and price the pick as the cost-controlled surplus bet it is.
Illustrative on the real draft-board layer (the board leading with the slot-appropriate projection, the rookie-scale surplus and the rookie-quarterback window, the three-bet decomposition, the team-specific re-pricing). Composite prospects and picks, demonstration figures.
The draft is the one hard, one-way gate in the whole architecture, and everything about it follows from that. College and pro football are two different sports and two different engines, so a prospect's college number is his present-tense value in a system he is leaving, not his value in the one he is entering, and the gap between the two is the entire projection. The engine crosses that gate with Mode 6, which never multiplies the college number by a coefficient, because there is none, it re-reads the prospect's traits through the pro engine and produces a distinct, lower-confidence forecast, an Entry KR and a banded Peak, Floor, and Median, in which the rank order can flip because the traits that win in college are not the traits that carry to the pros. The board then leads with the number the slot is actually buying, the ceiling at the top, the projection in the middle, the role and the floor at the end, and it prices the pick as the highest-surplus asset in the sport, the cost-controlled rookie deal that makes the rookie-quarterback window the most valuable position in football. And every pick is three bets, the player, the projection, and the team that has to develop him, each with its own confidence, so a miss is traceable rather than mysterious. Draft the forecast, not the college season, band it honestly, and price the surplus the slot delivers.
Draft the projection, not the college season. Price the surplus the slot buys.
NFL Draft Intelligence reads a prospect across the hard gate, projects him through the pro engine with Mode 6 at an honest band, leads the board with the number the slot is buying, and prices the pick as the cost-controlled surplus bet it is.