Passer rating was built for 1973 football
The NFL passer rating is the only major statistical formula in American sports that still appears unchanged on broadcast graphics fifty-three years after it was written. It was designed in 1973 by an NFL statistician named Don Smith, commissioned by commissioner Pete Rozelle, to solve a specific historical problem: how to compare quarterbacks whose era-adjusted completion percentages, yards per attempt, touchdown rates, and interception rates were wildly incompatible. The formula did that job. The game it was built to score is no longer the game being played, and the rating has become a number that points away from the questions modern football is actually asking.
What the formula does
Passer rating takes four inputs: completion percentage, yards per attempt, touchdown percentage, and interception percentage. Each input is normalized against a 1970s-era baseline, capped at a ceiling of 2.375 and a floor of zero, and the four normalized values are averaged and multiplied by 100/6 to produce a number on a scale that runs from 0 to 158.3. The maximum is not an accident of the math. It is a deliberate ceiling chosen so that no quarterback could be rewarded for stat-padding past a defined level of excellence.
That ceiling matters. A quarterback who completes 80% of passes for 13 yards per attempt with 12% touchdown rate and zero interceptions earns a 158.3. A quarterback who completes 80% of passes for 16 yards per attempt with 14% touchdown rate and zero interceptions also earns a 158.3. The formula cannot tell them apart, because both are above the cap. That was not a problem in 1973, when nobody was producing rates anywhere near the ceiling. It is a problem in a league where the top single-game performances routinely max out every input.
What it leaves out
Passer rating measures four things and ignores everything else a modern quarterback does. It does not include sacks. A quarterback who throws for 280 yards on 40 attempts and takes eight sacks for 65 yards has the same passer rating as one who throws for 280 yards on 40 attempts and never gets sacked. The 65 yards of lost field position and the eight possessions ended early do not appear in the calculation. Sack avoidance is one of the most important quarterback skills in the modern game. The flagship efficiency stat is blind to it.
It also ignores rushing. A quarterback who picks up 60 yards on six scrambles, including two third-down conversions, does not get credit for any of it in the rating. Lamar Jackson, Jalen Hurts, Josh Allen, and Kyler Murray have all produced seasons in which their scrambling contributed more EPA than several teammates' entire offensive output. None of that shows up. The 1973 game did not have quarterbacks who rushed for 1,000 yards a season. The 2025 game does. The rating still does not.
The yardage problem
The yards-per-attempt input treats every yard the same. A five-yard completion on third-and-four counts the same as a five-yard completion on third-and-eleven. A 40-yard touchdown counts the same in the rating as a 40-yard pass that ends in a turnover at the goal line, since the formula scores those possessions through different inputs and does not net them against each other. Down-and-distance context is not part of the math because down-and-distance data was not routinely available in 1973. It is now. The formula has not caught up.
This is also why dink-and-dunk offenses produce passer ratings that flatter the quarterback. A signal caller who completes 75% of his passes for 6.8 yards per attempt with modest touchdown and interception rates earns a rating in the high 90s, which sounds like good play. That same quarterback may have an EPA per dropback well below league average if most of those completions are checkdowns on third-and-long that do not move the chains. The rating reads as success. The drive sheet reads as failure.
What the modern alternatives capture
EPA per dropback, success rate, and ANY/A — adjusted net yards per attempt — were each designed to fix one or more of the rating's blind spots. ANY/A folds sack yardage and sack count directly into the per-attempt denominator, which means a quarterback who eats sacks gets a worse number, and the worse number is the right number. EPA per dropback weights each play by its actual impact on scoring probability, so the five-yard completion on third-and-eleven correctly scores as a near-zero play and the five-yard completion on third-and-four correctly scores as a large positive.
Success rate, meanwhile, is the simplest of the three and often the most predictive in season-to-season correlations. It tabulates the percentage of dropbacks that produced positive EPA, which captures the idea that consistency at moving the chains matters more than the occasional explosive. Every modern QB model used by front offices is built on some combination of these three inputs plus pressure-to-sack rate. None of them lean on passer rating.
Why the rating survives anyway
The passer rating has stuck around for the same reason RBIs stuck around in baseball for a century. It is on every scoreboard graphic, every box score, and every press release. Broadcasters can flash it without explanation. Fans recognize the number. The ceiling of 158.3 has acquired a mythology all its own. Anyone arguing for a replacement has to fight the inertia of half a century of convention, and the replacement options all require more explanation than the broadcast booth wants to give.
The NFL itself partly conceded the point in 2011 by introducing Total QBR, a proprietary ESPN stat that incorporates rushing, sacks, and play-by-play context. Total QBR is closer to a modern model but is not transparent about its weights, which makes it hard to scrutinize. The result is a status quo where the broadcast number is wrong in known ways and the replacement number is a black box. Front offices bypass both and build their own.
The honest read
The right way to read a passer rating is as a partial snapshot of the 1973 conception of quarterback play, applied to a 2026 game. It still tells you something about completion accuracy and interception avoidance, which are real skills. It tells you almost nothing about sack avoidance, rushing contribution, third-down conversion, or red-zone efficiency, which are also real skills and are increasingly the ones that separate starting quarterbacks from backups. The number that closes out the box score is not the number that decided the game. The two have been drifting apart since the formula was written, and the drift is accelerating.