SPIN Production Rankings: How POP and Proof Work

Production over potential

Recruiting rankings measure what a player might become. Ours measure what he did. Those are different questions, and the second one gets asked far less often than it should.

College coaches are paid to project. They look at frame, speed, and how a 16-year-old might fill out by his third year in a weight room, and they are right to. But a star rating is a bet on a future, and a lot of Friday nights get lost inside it. The receiver who put up 1,200 yards against the best defenses in his state and the receiver who ran a fast forty in shorts are not the same player, and only one of them has proven anything yet. SPIN ranks the proof.

No recruiting rank, star rating, offer list or college destination is an input to these boards. A player's rating is competition-adjusted production: his season, adjusted for who he played it against. During the season we call it POP, for Production Over Potential. When the season is final, it becomes his Proof.

Competition level is the whole argument

Every statistic on a high school box score was earned against somebody, and the somebody varies more at this level than at any other in football.

Picture a college conference that had Georgia, Kansas State, Boise State and a Missouri Valley FCS program in the same division, playing each other every year for a title. Nobody would compare their quarterbacks' yardage straight across. That is a normal high school region. In late September, one Utah 6A region had a program rated inside the top 100 of the roughly 14,000 eleven-man teams in the country and a program rated past 4,000th. They are in the same region. They play each other in October. A defensive back's four interceptions against the first team and four against the second are not the same four interceptions.

It gets wider across classifications. Put all six classifications on one page and the range of opponents a Utah player might face covers almost the entire national spectrum, from the best programs in the country to programs that would struggle in a JV league two counties over.

So we adjust for it, game by game. Every opponent a player faces carries a national rating, and we use the most current one: the final rating for a finished season, the latest rating we have during a live one, because the question is how good the opponent turned out to be, not how good it looked in August. Production against an elite opponent is worth more; production against a badly overmatched one is worth less. That opponent-by-opponent adjustment does most of the work, and the region example above is the reason it has to.

Where the game was played counts too. A home game gets no adjustment. A road game is credited as if the opponent were a little better than its rating, by the home-field edge we measure in our own scores, and a game at a neutral site gets half of that. Going on the road against a good team is the hardest job in the sport, and the rating treats it that way. The edge is small, so the adjustment is small, and it is the same for every program.

A smaller adjustment sits on top of it for the general level of a player's classification. This one is a judgment, and we will say so plainly: a full season in a state's most competitive classification is a different weekly environment than the opponent ratings alone describe, and we apply a modest credit for it. We tested the boards with and without it and against several sizes of it. The order at the top did not depend on the choice, and we kept it because it made the boards more stable from one season to the next. We also tested adjusting for the side of the ball a player actually faced, the opposing defense for an offensive player and the opposing offense for a defender, and it did not improve the boards over the whole-team rating.

The result is that a quarterback with 4,000 yards in a smaller classification and one with 3,000 in the state's most competitive can land next to each other on the same board, and both of them earned the spot.

This is what makes one statewide board possible. We do not think a running back in a state's smallest classification and one in its most competitive are playing the same game. We think we can measure the difference, and once you can measure it you can rank them together.

One more judgment sits above the state level. National team ratings are good at sorting a region's own teams and worse at sorting one region against another, and our editors watch these programs play each other. So each season SPIN ranks the regions it covers, and that ranking makes a slight adjustment when players from different regions are compared. It never changes the order inside a region; a Utah board is a Utah board. It only matters when a Utah player and a California player are on the same page, and there it leans the way the games have leaned. Where a program's schedule is national rather than regional, its players are ranked against the region it actually plays, not the one it is zoned in.

California has no statewide classification, so for California we do the same job a classification does by grouping leagues into five SPIN competition tiers by the strength of the opposition their teams actually faced, drawn once from the previous full season and frozen for the year. The tiers are ours, not the CIF's, and they change nothing about a player's rating. They decide which Top 10 he appears on.

How deep the boards go

Every player with a stat line in at least five games carries a rating, from the best quarterback in the largest classification to a sophomore getting his first carries in the smallest. The boards print the top of that list: a Top 25 at each position statewide, and a Top 10 for each classification as the state's governing body draws it, or, in California, for each CIF section and each SPIN competition tier. We check every classification and section against the governing body before a board publishes. A group's Top 10 at a position publishes only when enough players in that group are rated to make it a real list; when there are too few, those players appear on the statewide board only, and the group page says so.

This is version one. We are starting with the top of each list on purpose, so the boards earn trust before they go deeper, and we will expand toward every rated player in a state as we go. When the depth changes, this page will say so. The 2026 boards start with Utah and Northern California. Arizona, Southern California, Nevada, Colorado, Hawaii and the Pacific Northwest come next, followed by a national board.

What each position is graded on

Every stat below is a season of individual game lines, not a season total, and every game is weighted for the opponent before it counts. Every play is converted into a single internal unit of production, so different kinds of plays can be added together. A touchdown is worth more than the gain that produced it, a turnover given away costs a great deal, a turnover taken away is worth even more, and a defensive touchdown most of all. The values are SPIN's, set to how much each play moves a game; play-by-play data, which would let anyone measure that directly, does not exist at this level.

Two things about the defensive side. An assisted tackle counts as a share of a tackle, not a whole one, because that is how the play happened. And defensive backs are graded on two tracks, coverage production and run support, with the stronger track carrying more of the rating, because most rosters list "DB" without saying corner or safety and the two jobs produce different stat lines. Letting the stronger track lead lifts every defensive back's number a little, and it lifts them all by about the same amount, so the order between them is what it would be under an even split.

Quarterbacks are the one position graded on efficiency as well as volume. A quarterback's production per attempt, adjusted for the defenses he threw against, is weighed alongside his total production, and his completion rate counts for or against him against a typical completion rate, lightly, because completion rate rewards a screen game as much as it rewards accuracy. Rushing yards and touchdowns count for a quarterback at full value in the production half of his rating and at a reduced value in the per-attempt half, where they have no attempts to divide by.

A player is rated on a side of the ball only when he produced on that side in a real share of his games. A roster listing by itself does not make a receiver or a defensive back. Players who clear that bar on both sides also appear on a Two-Way board, where the rating is the average of his two sides.

What is not graded: return yards, kicking and punting, blocked kicks, and anything on the offensive line.

How a season becomes a rating

Every game a player recorded a line in is valued on its own, then the games are combined. The scale is logarithmic, so the biggest nights earn diminishing credit and one huge game cannot carry a season, with a floor so a quiet night against a good defense does not zero out a month. Production in games that were lopsided before kickoff, judged by the two teams' ratings, is discounted by a measured amount, because garbage time is not the same football.

A player's per-game figure is then pulled toward the average for his position in proportion to how few games it rests on. A player with five games is trusted less than one with thirteen, which is what keeps a hot month from outranking a full season. Players need at least five games with a recorded line to appear on a board, and quarterbacks need a minimum number of attempts.

Showing up counts. A player who missed a month is not charged for it as if he had four empty nights, but a player who produced for eleven games has produced more than one who did it for eight, and the rating credits games played on top of the per-game figure. Durability and availability are part of production.

We rate per game rather than per carry or per target on purpose. Touches and attempts are entered by hand and can be edited afterward; games are on the public schedule. And rate statistics on small workloads swing wildly on two long runs. Quarterback is the exception, where production per attempt is part of the rating, because passing is the one place the box score gives us a clean denominator.

Verified statistics

SPIN production ratings use the most official statistics available for each player. Some are publicly available, some are brought to us by programs and verified, and some come from official organizations. Every line on a published board goes through our checks before it counts, and when a line could not be verified we say so. Where a public record and a verified record disagree, the verified record is the one we use, and it stays used. A program that publishes no statistics has no rated players. Some programs post only part of their season. Where too little of a program's season is posted, its players are held off the boards until it is, and where we know a player's line is missing games, the board marks it.

Public statistics at this level are entered by coaches, volunteers and parents. They are imperfect, and defensive statistics are the least consistent of all: one stat crew hands out two assists a play and another hands out none. Garbage in, garbage out: if a stat crew credits a player with plays he did not make, the rating believes them until somebody shows us otherwise. If you are a coach, a player or a parent and a number here is wrong, send us what you have. If it checks out against film or an official record, we fix the input, we log who told us and when, and the correction carries into every board that number touches. The process is the same for every program.

Reading the rating

POP and Proof. The rating has two names, one for each stage of a season. During the season it is POP, Production Over Potential, and it moves every week as games come in. When the season is over, its box scores have been read one last time and its season totals checked, every POP becomes that player's Proof: the final record of what he did that year. A Proof can be corrected if a verified record turns up later, and a corrected Proof carries a dated note.

Both run from 25 to 99 on one scale, and a number means the same thing at every position and in every state we rate. The scale is calibrated to national percentiles: each position is calibrated separately, and the number says how rare that season of production would be among starters at the position across the country. We estimate that from every region SPIN rates, pooled and set against the national ratings of every eleven-man program, so a 90 in Sacramento is the same 90 in Salt Lake City, and the estimate firms up as SPIN adds states. The scale is mapped so that it reads the way people read a grade.

A rating is the same number wherever it appears. A player's rating on his classification board is his rating on the overall board. The class board is the same list, filtered.

Next to each rating we print the schedule a player faced as Heavy, Standard or Light against the national field, his games played, and his full season stat line for the position. The schedule label is a summary for the reader; the calculation uses every opponent's rating, not the label. The tiers give the context; the number gives the rank. A 91 and a 90 produced seasons of similar quality, and the order between them is still the order: on a deep board a single point can separate several places, and the player on top earned the point.

How we know it works

A production model is easy to make look reasonable and hard to make right, so we test it against things we did not build. Our 2025 Utah boards agree with the all-state teams, which we had no part in selecting, at about two of every three spots, judged within each classification. When we built the engine, the same boards separated players who drew major-college recruiting interest from those who did not about nine times in ten. In testing, both figures were higher with the competition adjustment than without it. We also ran the boards under every alternative we seriously considered, several sizes of competition adjustment, several quarterback blends, different values for defensive plays, and the top of every board held. Where a choice did not change the order, we picked the one that was most stable from one season to the next. When we change the engine, we update this page.

What the box score cannot see

A box score records things that went right. Nobody at this level tracks the missed tackle, the dropped pass, the throw that was late, or the corner nobody threw at because he was the corner nobody threw at. A production rating measures results, and it measures them more carefully than a stat sheet does. It does not measure process. Decision-making, timing, accuracy and the ability to make something out of a broken play show up in a box score only through what they produced. If SPIN adds a grade for those traits, it will come from film or a live evaluation, as a separate number, and it will not be folded into this one.

We do not rank offensive linemen. There is no box score for the position and no honest way to build one.

The first 2026 boards: Utah and Northern California.