There are many ways to evaluate quarterback performance. Some are good, some are bad, and some end up with John freaking Elway at or near the top of the list (usually the ones that are... bad). There are a few """advanced""" black-box metrics like QBR and I guess now also EPA/pass, but I'm still a fan of the old standard, passer rating. Passer rating is a weighted (and, unfortunately, capped) aggregation of four crucial statistics representing passing performance: completion percentage, yards per pass, touchdowns per pass, and interceptions per pass. Essentially, a good quarterback should complete more of their passes for longer gains, more touchdowns, and fewer interceptions. It seems simple, and is, but it's still quite a good metric, even in comparison to the proprietary (and possibly overtuned) metrics made by corporations of various flavors. It also has the notable advantage of being applicable even for quarterbacks who played many decades ago, long before the advent of advanced stats. We have no idea what Sammy Baugh's QBR or EPA/pass is, but we can calculate his passer rating with perfect accuracy.
Passer rating is a very good metric that effectively measures how well a quarterback performs, over the course of a game, a season, or a career. It does, however, have a few shortcomings when it comes to making historical comparisons. Most relevant to our purposes, passer rating has broadly increased over time. In 1975, the average passer rating was 65.8; in 2025, it was 91.4. This gap gets bigger as you go further back in time. There are many causes for this phenomenon. The naive explanation is that quarterbacks have simply gotten better, which is certainly true on average, but far from the whole story. The NFL has changed dramatically over the years, both due to rule changes (which have largely favored the passer) and changes in how coaches and teams assess the value of various aspects of the game (put simply, modern teams value passing efficiency -- and volume -- far more than teams of the past). Advancements in coaching have also played a major role; for instance, Joe Montana, who played in Bill Walsh's legendary West Coast Offense, thrived due to the scheme's emphasis on short, precisely timed routes, rather than deep vertical throws, which his relatively weak arm strength precluded. Modern quarterbacks benefit from all of these developments, while older quarterbacks had access to none of them.
This leads to some strange phenomena, such as the indisputably mediocre Jimmy Garoppolo -- who has bounced around to four teams across a 12-year career, largely as a backup -- having a higher career passer rating than, among others, Tom Brady, Steve Young, Peyton Manning, Kurt Warner, and Joe Montana. Is Garoppolo "better" than them in any meaningful sense? No -- not even in an inflationary sense. He's not bad (he's slightly above average for his career, with a few flashes of very good play), but he's not an all-time great, and it would be hilariously incorrect to consider him one. So a flat comparison based on passer rating is obviously insufficient. Rather, we must assess how good a player was relative to their era.
Fortunately, there is a metric that does precisely that: Passer Rating Index (or PRI). PRI scores a player based on how well they performed, in terms of passer rating, relative to the league they played in. A score of 100 is average, and every 15 points above or below the mean indicates one standard deviation. In short, a 110 PRI season is good, a 120 PRI season is great, a 130 PRI season is MVP-worthy, and a 140 PRI season is one of the best ever. In any given year, the best season tends to range from about 120 to 150, with an average league-leading season landing at around 133. But these numbers are deceptively far apart. There have only been 25 seasons in NFL history with a PRI of 140 or higher (think: 1984 Marino, 2013 Foles, 2024 Jackson, and 2007 Brady), whereas the 120 PRI range includes lots of good-but-not-great players, like Matt Hasselbeck (122), Jimmy Garoppolo (121), Andy Dalton (125), and of course John Elway (121, his career best. This is not the last shot I will be taking at Elway).
Pro-Football-Reference doesn't have a "career PRI" number, unfortunately (I mean, they do on each player's individual page, but I'm not doing that), but even if they did, it wouldn't be perfect. For one thing, players can get punished for underperforming early or late in their careers; for another, an average rating would fail to capture peak, while punishing players who play for many good-but-not-great seasons. So I've come up with a better approach (and one that is, you know, possible with the data I have access to). Here it is.
I pulled all the seasons from players with a Passer Rating Index of 120 or higher (actually 121, because that's how far the PFR leaderboard goes and I'm never ever ever ever paying for Stathead). Then I assigned players a point value based on how many seasons they had with a PRI of 120+, 130+, and 140+. Essentially, a 120-PRI season is worth 1 point, a 130-PRI season is worth 2 points, and a 140-PRI season is worth 3 points. For a tie-breaker, I went with their highest PRI season (rewarding players who peaked higher), followed by how recently they played. This latter tiebreak is here for two reasons: first, because I have an intuitive bias towards thinking modern players are better than older ones (they are, but I'm also just biased here); and second, because modern players face much higher levels of competition, and it makes sense to reward them for succeeding even in the face of such high standards. Anyway, I did some very complicated Excel magic (I even got it to produce a ranking system that accounts for ties) and produced a list of the greatest quarterbacks of all time. Here's the top 20 (really 23, because ties; they're listed by their rank, but sorted within ranks by the tiebreakers I mentioned):
(Most of this should be self-explanatory. "Max Score" is obviously max PRI, "S>###" indicates how many seasons they have above that PRI threshold, "Score" is the sum of these seasons, and "AvgYear" roughly indicates when their prime took place, by averaging the years of their 120+ PRI seasons.)
Most likely, the first thing you noticed when reading this list (aside from my exceptional data viz skills) is how many Old Guys are here. And it's true; in that top 20 (23), 8 of them played all or most of their peak seasons before 1970, and four of them have peak seasons from the '30s and '40s. (Five, if you count Y.A. Tittle's legendary 1948 season that you had definitely heard of before right now.) But the next thing you should notice is how good this list is. Yes, it has some weird old guys in it, but it also does a really good job of aggregating the best quarterbacks ever up at the top, without caring much about our era-related biases. And indeed, the average year of quarterbacks in my top 23 is 1981, just barely more than halfway (1978.5) between the first pass attempt ever recorded on PFR, in 1932, and the modern age in 2025. One player is from the 30s, two are from the 40s, two from the 50s, three from the 60s, four from the 70s, two from the 80s, one from the 90s, four from the 00s, three from the 10s, and one from the 20s. Whatever else you might say about my list, it is representative.
It is also, in my opinion, very accurate. Shockingly accurate, for a scoring system this simplistic. All of the players I consider serious contenders for the top 10 are in the top 23 (you may whine and cry about Mahomes, but he is tied for 24th, which is really not bad considering how short his career has been and how he has fallen off somewhat in the past few years). And the players at the top have seriously earned it. Yes, it's strange to see a list starting with both Sammy Baugh and Otto Graham, but they really were outstanding quarterbacks for their era. Remember how I said passer rating rises over time? Well, Baugh's 1945 passer rating of 109.9 would, today, place him #2 in the league, and Graham's 1947 109.2 rating would tie him with Stafford for #2. And while, yes, the league is much more competitive now, and they would surely struggle to adapt, it's also true that they were doing this in a league where the average passer rating was, like, 60. (In 1945 it was 47.6.) What's more, you can't get this high on the list from one great season; the peak for all-time one-hit wonders (ft. Nick Foles and Brock Purdy and no one else) is a tie for #30 all-time, with the likes of Milt Plum and Daunte Culpepper; solid players, but not contending for a top 10 spot. That is to say, Baugh and Graham were exceptionally good consistently. Baugh had 7 seasons of 120 PRI or higher, while Graham had 6, and both had 3 seasons of 140 PRI or better (no one else has ever done this). Was it easier to do that back then? Sure. But they were still outstandingly, unimaginably great relative to their competition. What else could we possibly ask of them?
Looking now at the more modern players, we get a stretch of Steve Young (one of the greatest quarterbacks of all time, and yet somehow underrated), Aaron Rodgers, Peyton Manning, Joe Montana, and Drew Brees (who characteristically has a bunch of excellent seasons and no all-time great ones). These guys are in everyone's top ten. Then we dip back into the past with Roger Staubach (my single biggest surprise here, no joke) and Len Dawson, who's been one of my pet "most underrated QB ever" picks for years (him and Young -- although after this, that list will have to be expanded to include Baugh and Graham as well). Then we get Tom Brady, Bart Starr, and some other Old Guys. (Shoutout to Ed Danowski with the all-time single-season PRI record of 234, WAY beyond #2's 198, both of which are absurd outliers. A PRI of 234 puts Danowski, what, 9 standard deviations above the mean, which is just silly. All six seasons of 150+ PRI are from guys in the 30s and 40s, with the exception of '04 Peyton Manning, because of course.) Most of these #11-23 guys are not hugely surprising, although Stabler and Anderson are a lot better than I thought (so's Tarkenton, actually), and I'm pleased to see Warner and Favre sneak in technically just outside the top 20.
I should probably take a second to talk about Brady. I know he's the consensus GOAT at this point, and I really don't mind it; his '07 season is the best I've ever seen with my own two eyes, and he's obviously won a stupid amount of important games in his career. His 10th-place finish might seem incongruous with this. But it's pretty easy to explain: for one thing, Brady's legacy is largely built on the back of his rings, which this list doesn't care about at all. His four qualifying seasons and one all-time great one ('07, of course) are excellent, and rightly earn him his top 10 slot, but it's also just a fact that he didn't have all that many statistically outstanding seasons. He had a lot of very good seasons, but only four above a 103 passer rating, and all four are on this list. He's not actually being underrated here. And I say that as, despite everything, sigh, a fan of his. We aren't judging career accomplishments here, we're judging statistical outstandingness, and those numbers do not lie. And they spell DISASTER for Tommy in Washington. SAMMY BAUGH >>> TOM """SHADY""", PATRIOTS BTFO, BAUGH WOW AND HOW!!!1!
Ahem. Anyway. The list goes on, well beyond my top 23 -- in total there are 130 quarterbacks on it. Here are a few interesting or amusing notes:
- Bob and Brian Griese both make the list (Bob at #24, Brian at #30). And, to answer your next question, Peyton and Archie are both here too, at #5 and #79. And -- your next question -- no, Eli is not.
- Jim Kelly is at #30, tied (but ahead by tiebreaks) with Russell Wilson and Philip Rivers. Rivers, to my surprise, is somewhat behind his contemporaries of Ryan and Romo, who are tied for 24th.
- There's Chad Pennington here, at #51, and like Trent Green (#51) and stuff. Ryan Tannehill is here (also #51 -- there's a lot of ties, okay?). It's a weird list.
- Terry Bradshaw is #51, just behind Troy Aikman (#51).
- Matt Hasselbeck is here! At #79. What's more, Dave Krieg is here, and he's not even bad! He's tied for #30, losing on tiebreak to, well, everyone, but tied nevertheless with Russell Wilson, Jim Kelly, Daunte Culpepper, Philip Rivers, and Boomer Esiason, among others.
- Quick note, tying back to my earlier comments -- Jimmy Garoppolo does in fact make this list! He's tied for #79, #121 after tiebreaks, one spot ahead of Josh Allen (lol).
- Out of the 130 quarterbacks on the list, John Elway finishes tied for last (#79), but sixth-to-last by tiebreakers (so, #125 after tiebreaks). I will say here, though, that his finish is not at all an insult; making this list at all is a notable accomplishment. Elway's 121 PRI is a genuinely good score, and would have led the league (checks notes) eight times in the last... 90 years. Well. Led or tied. The takeaway is not so much that we should laugh at Elway for only being 21 points above the mean, but rather that his career high water mark fails to reach the highs of the actual all-time greats, and indeed falls somewhat short of virtually every other quarterback ever included in so much as an MVP conversation...
- Except Cam Newton, who didn't make it. And yes, he is the only MVP quarterback in history to miss this list, I checked. (To be clear, not every MVP quarterback made it in their MVP season -- most recently, and appallingly, in 2024 Lamar Jackson's all-time-great 143 PRI season, the 13th best season of all time by PRI, lost out to Josh Allen's 116 PRI -- but they all made it sometime.) His 115 PRI falls somewhat shy of the career bests of Sam Darnold and Geno Smith, ties the career best of Tarvaris Jackson (RIP), and only barely beats out the best effort of Jim Zorn. It does, however, clear the 112 PRI of John Elway's MVP year; Elway's qualifying 121 PRI season came six years later.
No comments:
Post a Comment