Most player systems stop at what happened. KaNeXT keeps going until it can explain what is real.
One player, one universal KR scale, read on one of two engines (college or pro) and placed on the ladder by KLVN and confidence, not by forking the number. The headline is a single locked OVERALL KR, the player read through his team's actual scheme (a defensive front by a coverage shell, or an offensive family by a personnel grouping), and it never arrives as a naked number: it arrives at a stated confidence and it opens. Under it sit the three components, the nineteen clusters, and the ninety-two traits, each carrying its own tier of data. And it is honest about the football that box scores lie about, the sack and the reception and the tackle that scheme and teammates create as much as the player.
A KR you can open until you hit the evidence.
The OVERALL is one number, but it is never a closed one. It opens into the three components the player is made of, then into the clusters, then into the individual traits, and every trait carries the tier of data behind it, so you can press the headline all the way down to what the engine actually saw.
The OVERALL is 84.2 and every layer under it is legible: the components, the clusters, the traits, and the tier of data behind each. Every one of the 19 clusters is on screen: the ones a receiver lives on (Receiving, Speed and Explosion, Agility, Offensive IQ) opened to their traits, and the ones he is not judged on (Pass Rush, Coverage, the blocking and defensive families) weighted to near zero by the OPF, because it is one union library read through his position. His tools are MEASURED off tracking and testing, his separation MEASURED off charting, his coverage recognition INFERRED and held at lower confidence because the feed cannot see it. A rating you can press to the cluster, the trait, and the tier is a rating you can argue with, which is the only kind worth having.
Illustrative engine read on the real KR hierarchy (OVERALL, the three components Skill / Athletic / Processing, the 19 clusters and 92 traits, each on its own MEASURED / PROXY / INFERRED / UNSCORED tier). Composite wide receiver, demonstration figures.
The counting stat is a starting signal. The charted rate is the measurement.
Football box scores are the most misleading in sports, because the counts are outcomes that scheme, teammates, and opportunity create as much as the player. A sack, a tackle, a reception, an interception: each is a real event, and none isolates the man. So the engine reads a box count as PROXY, a starting signal, and reserves MEASURED for the charted and tracked feeds that actually isolate the player.
Total tackles, sacks, interceptions.
Shaped by targets, scheme, and the players around him.
A starting signal, corroborated within band, never flipped on a hunch.
Pressures, forced incompletions, yards over expected.
Tracked speed and the isolated win.
The feed that measures the player, not the outcome around him.
This is the football form of grade-the-player-not-the-box, and it doubles as the confidence story: the same reception total that reads MEASURED at the NFL or a Power conference (full charting and tracking) reads PROXY at a Division II program off the box alone, so the same production carries a tighter or a wider band by how much of it the engine can actually isolate. One scale, and the level lives in the confidence, not in a fudged number. Read the player the box cannot show, and price what you cannot see honestly.
Illustrative engine read on the real football box-tier rule (box counts at PROXY, charted and tracked feeds at MEASURED) and the coverage-driven confidence gradient. Composite receivers, demonstration figures.
One OVERALL, read through the scheme he actually plays. The rest is on demand.
Production is always read through a scheme, so there is no scheme-agnostic base number. The OVERALL is fixed to the offense or defense the player actually plays in, badges baked in and his special-teams contribution folded, and it never moves. The what-ifs, how he would read in another scheme, and his best-case ceiling, are on-demand views that never pose as the headline.
Move him into a vertical Air Raid and the fit and the scheme-read KR shift while the locked OVERALL never moves; his ceiling is a real 86.4 offered only when asked. The default output is the true one, what he is where he actually plays, because that is what is real. And the whole number is anchored to production first, with the component math confirming within a bounded window, never inventing a bigger number than the field supports. Believable over impressive: a real 81 percent beats a manufactured 95.
Illustrative engine read on the real locked-OVERALL, Scheme Explorer, and on-demand ceiling structure (offensive scheme = family by personnel; a single scheme fit; the special-teams contribution term folded). Composite receiver, demonstration figures.
A player rating is a bet on an unhappened future, and the honest version is never a naked number. It is a number you can open down to the trait and the tier of data behind it. It is a number that knows what the box score cannot show, so the sack and the reception and the tackle that scheme and teammates create are read as the starting signals they are, and the charted rate that isolates the player is what earns MEASURED. It is a number that carries its own confidence, so the same production read at the NFL and at a small college is told apart by its band and not flattened into false equality. And it is a number locked to what is actually true, the player in his real scheme, with the flattering what-ifs kept on demand where they cannot pose as the headline. A manufactured 95 impresses once and fails the first time reality tests it. A real 81 survives.
Press the number. See what is real.
Player Intelligence gives you a single locked KR you can open to the evidence, priced with the confidence that sizes the bet, honest about the football the box score hides.