How we read players.

Three things about every player

We describe a player in up to three beats. The first is always there; the other two only appear when the numbers earn them, which for most players they don’t.

Most players have no twist line. That’s the design: when a twist appears, it means something.

What the labels are

Every label is cohort-relative and box-score-honest. We take a player’s per-season statistics, compare them only against other players at the same position that season, and let the shape of those numbers pick the archetype. A “top-1%” on the site always means top-1% among FBS players at that position that year — not all-time, not across positions.

The labels describe how a player produced, from what the box score records. They are not scouting grades, not a measure of talent, and not a projection. They begin when the college data does.

The CFB Index Score

Every scored player also gets a single 0–100 number, the CFB Index Score — how good his season has been, built from real production and an opponent-adjusted read on the plays themselves. It is only ever compared against other players at the same position, in the same season — a 70 always means the same thing: solidly above an average starter at that position that year, nothing more.

A player needs to clear a minimum number of games before his score counts toward any ranked board. Below that floor we still show the number on his own page — we just label it plainly as a small, still-building sample rather than a settled read, and it’s left out of any “#N of M” ranking until it qualifies.

We don’t publish the formula’s internal weights or coefficients — the score is meant to be read, not reverse-engineered. What we do publish is the rank itself: every qualified score comes with its plain position in that season’s field, e.g. “#4 of 132 quarterbacks.”

Today the score covers offensive skill positions only — quarterback, running back, wide receiver, tight end. Defensive scoring is being built and held back until it clears the same accuracy bar the offensive score already has.

“The rare part” — and how rare it is

An archetype credits a player for one signature strength — the thing that defined his label. The rare-part star fires when he is also elite at a second thing his label never mentions: top-tier in the credited trait, and top-5% in an uncredited one. A Pocket Surgeon is labelled for accuracy; the star says he can also push it deep and find the end zone.

It is genuinely scarce — it fires for well under 1% of player-seasons, and only when the player also cleared a real workload (enough attempts, carries, or snaps that the number isn’t a small-sample fluke). The wider list of these players lives on the two-way elite board.

The accolade ladder

Standing is a seventeen-rung ladder, from walk-on up to national player of the year. The rungs come from two independent places: real recognition (watch-list picks, all-conference and all-America selections, player-of-the-year honors) and, below the honors tiers, a production rung from usage and output. It measures how much the sport has recognized a player — a different question from how he produced.

“Ahead of the ballot”

The most interesting gap is a player who is elite on the tape but still low on the recognition ladder — the rare-part star fires while he’s below the watch-list. We call it ahead of the ballot: the production is there, the accolades haven’t caught up yet. You’ll see it noted directly on a player’s own page when it applies.

The flaw note, and what licenses it

Occasionally a player is elite at his signature and, at real volume, has one genuinely weak area his label doesn’t already warn about. When that happens we write a single note, and it lives under strict rules:

• The number is the whole claim — “a fumble every 31 touches, bottom 4% among backs with his workload,” never a pejorative adjective like “fumble-prone.”
• It’s always tagged ~ our read — our reading of the box score, not a settled fact.
• Volume floors apply — a flaw can only fire on a near-full season of work, never a couple of bad games.
• It only appears in a player’s own context, once a player is a locked-in starter — a real hole in a prominent player’s game is analysis; nitpicking a deep reserve is not. There is deliberately no list, hub, or leaderboard of weaknesses.

What we can’t see

This is the honest part. Everything above is built from the box score, and the box score is blind to a lot of football:

• Drops, catchable balls, and route-running. We see catches and yards, not the throw a receiver should have caught or the separation he did or didn’t create.
• Blocking. Offensive-line play barely registers in the box score, so we make almost no measured claims about it — and a blocking tight end’s best work is invisible to us.
• Coverage. A shutdown cornerback’s reward is that nobody throws his way — which reads as an absence of stats, not a strength. When the box score can’t see why the voters are right, we don’t second-guess them.
• The defensive rung gap. Below the honors tiers, the production rung is built from offensive usage; there is no equivalent for defenders yet. So a genuinely elite defender who hasn’t been voted an honor can only reach recognition through a ballot — and an “ahead of the ballot” defender who stays quiet in December is often bumping the ceiling of what the ladder can measure, not a bad read.
• Situation and scheme. Strength of the surrounding cast, play-calling, and game script all shape the numbers, and we only partly adjust for opponent.

When something is uncertain, we say so — that’s what the ~ our read tag is for. We’d rather name a limit than paper over it.