I have just spent the last 10 minutes trying to think of a way to describe for the generations after me how outlandish Jim Everett as a Hall of Famer sounds to those who followed his career. This would be the hottest hot take ever offered on ESPN and a true tribute to free speech were you allowed to make the argument. Except for his third and fourth seasons as a starter, he was not even regarded as a good player, and I think Henry Ellard and Willie Anderson got a lot of the credit for those good seasons. He could throw the deep ball, I'll give you that. As you know, he was considered a laughingstock afterwards, with the Jim Rome thing commanding attention. One can certainly see how that formed an unfair narrative. But Everett as a Hall of Famer?
I think what it really comes down to is what statistics you use and just how you choose to balance the good and bad seasons you get from those statistics. You have your favorite statistics and your way of balancing peak and career, and you follow this method to the end of the earth. Applied to historical players, when the statistics traditionaly haven't been, they lead to particularly (ahem) unconventional ratings. But when there is such a gulf between, say, the quarterback's won-loss record (64-89) and his number of Pro Bowls (1) and your rating, I think you have a responsibility to be extra analytical and self-critical and try to see what if anything you may be missing. If different perspectives don't validate your particular method, that is troubling. Do common sense and your argument fit together?
My major project 15 years ago was rating all quarterbacks historically by their teams' yards per drive in years when they were the starter. Everett was absolutely unremarkable in that. I also have his composite league interception rate as compared to the league's as mediocre. It seems unlikely to me that the running game would move the needle enough to have cost his offenses so much in yards per drive. Even after 1989, Everett did put up some yards, but it is easy to do that when you are losing most games and throwing a lot. He never averaged better than 7.2 yards an attempt after '89. Maybe there is more to be learned about the deficiencies in the particular statistics you are using than about Jim Everett from them? If you scramble the deck, try something else somewhat reasonable and still have Everett as a standout, then we can talk some more. But you're a talented guy, and I do worry for your reputation when you say something that is just regarded as crazy, and you don't seem to have any notion that it is taken that way, and that you have no intellectual company in the opinion.
I would rate Rivers, Roethlisberger, and Manning in the same order that you do. Where Manning was better than you give him credit for was in his durability. He never missed a game because of injury. It didn't seem to be luck, not that that matters. It was amazing the hits he got up from. The other thing the Giants' coaches said was that he was expert in audibling to running plays, and so deserves some of the credit for how well the Giants ran through 2008 or 2009. Not that it adds up to a Hall of Fame career, but just to make a full accounting. To me, rather than being like a Bumgarner, he was like an old-time pitcher who was 250-240 for his career, or something.
You know I've ever been one to shy away from a hot take David, as long as it's positive about a player. This is no different.
I have had that same internal debate about Jim Everett with myself multiple times. In essence, it comes down to how much QB volume is worth in the years before QBs touching the ball 600 times became standard. It's sort of like the worth of a volume goaltender in ice hockey. I'm not sure if you watch or care about ice hockey, but it's the age old question. What's better, 45 starts at an 80 GA%-, or 70 starts at 100? I suppose an imperfect baseball analogy could be whether you'd prefer 23 starts of 120 ERA+ or 33 starts of 100.
In football, it's the exact same thing. Would you prefer 498 plays of being 39% better than the average QB (1989 Joe Montana) or 602 plays of being 28% better than average (1989 Jim Everett)? Modern statistical understanding likes Jim Everett better. This is why Jim Everett is number one on my QB ranking list for 1989, and not Joe Montana. That sounds like blasphemy to those who lived through it, but I'm not trying to gauge the opinions of those who lived through it. I didn't live through it. It would only be a guess on my part, so I elect not to even try.
It's not meant to say that Jim Everett was perceived as being better than Joe Montana in 1989. It's to say that if 1989 happened in 2024, modern football people would like Jim Everett's season more, even if they didn't like it more then, which they clearly didn't because Joe won the MVP. I believe when I was talking about Jim Everett I was free from the constraints of the word regarded in my formulation, so I felt free to discuss 1989 as if it happened yesterday, unconstrained by the views of the time. This is much like how with modern statistical understanding Dave Stieb is dramatically better than Jack Morris. I can say that comfortably. Nevertheless, Jack is in the Hall of Fame, while Dave is not. If you interpret that fact as a mistake (which I do, and a lot of others do as well), it's an implicit acknowledgement that modern statistical understanding has a place in debates like this, at least in some capacity.
I entirely understand that Jim made the Pro Bowl just one time, but it's got more to do with his place and time than anything. For an example of the opposite, look no further than Donovan McNabb, who I don't think even yourself would assert is anywhere near as good as Jim Everett. Nevertheless, he made five Pro Bowls, due to being right in the middle of perhaps the weakest conference in NFL history, which is the NFC in the early-mid 2000s. This is how Donovan can share a league with somebody like Trent Green, and make the Pro Bowl three times more. It's not a fair fight.
The late 80s-early 90s NFC was a bloodbath. Mix this in with galling decisions like somehow Randall Cunningham going over him in both 1988 and 1989, despite having worse individual numbers at the head of a worse team (shown beautifully when Jim and the Rams spanked the Eagles in the 1989 playoff game they played), and you can see why I respect Pro Bowl count so little. It's both uneven across history and incorrect in its own place and time, both due to conference barriers and horrifying decisions like the Randall Cunningham one.
As far as team record, it's not Jim's fault that he spent the whole of his post-1989 career stranded on some of the worst rosters in the entire NFL. When I get myself into Tom Brady-Peyton Manning debates, and I bring up the fact that Tom had so much more defensive help than Peyton did, what do I hear constantly? 'Players don't build the roster.' Now, I'm invoking that argument here. It was not Jim's choice that his Rams deteriorated into the worst roster in the league right as he was (theoretically) entering his prime. He did not choose to get traded to the NO Saints conveniently right after their roster had fallen apart, and not in time to perhaps save it.
Just look at the 1991 Rams as an example. Behind all the time, theoretically constantly in disadvantageous positions due to always having to pass the ball, this is still a pass offence that ranked eighth in the league in net yards per attempt. Eighth is not spectacular, but it's eighth. Mix this in with his propensity to turn the ball over a little bit, you still get that he was the 11th best QB in 1991, according to me. 3-13 record or not. I don't know why they scored so few points. There's no PBP data for me to cross reference, but yardage wise, this was still a passing offence to be reckoned with. 1992 is similar, with slightly fewer yards, slightly fewer turnovers, slightly more points, etc..
You are right that most of Jim's career was spent dragging doomed teams to above average passing offences. This entails a lot of volume, and a lot of volume on 17% DVOA play (1992 Jim Everett) makes you a really valuable player. What would you say about a pitcher who has the fourth most innings in the league, and finishes sixth in ERA+, but somehow goes 8-10 on the year? You would call him a great player, who needs a better opportunity to let his talent shine. That's what I'm saying about Jim Everett, for basically every season of his career. This even continues as late as 1995, where Jim's DVOA is an astounding 19%, on the fourth most touches in the league. Absolutely nobody was talking about Jim Everett in 1995, but with stats like this they should've been, and would've been if 1995 were the season that's just finished, with modern statistical understanding.
Do take note that the league leader in EPA/Play in 1995 was Erik Kramer with his 0.166. Offence was just not the same thing back then as it was today. The best QB in the NFL back then is about as good as Kyler Murray now. The goalposts have moved, so I'm not saying that you can drop Jim Everett (or anybody whose prime was in the late 80s) into the league now and he would thrive, but relative to his era, I like him better than anybody except Joe Montana, Steve Young, and Dan Marino (and Dan Fouts, who he had an extremely limited crossing of paths with I believe).
Also, I don't find my rankings to be that conventional. Applying the same methodology to all of NFL history since the merger reveals a top ten QBs of all time list featuring the following names (unordered): Drew Brees, Aaron Rodgers, Peyton Manning, Dan Marino, Steve Young, Joe Montana, Fran Tarkenton, Roger Staubach, Patrick Mahomes, Dan Fouts. I don't believe this is that unconventional of a ranking. The one thing that people disagree with is that it puts all of its weight on scoring and moving the football, and none of its weight on winning (hence no Tom Brady or John Elway), but it's not a greatest list. It's a best list. Greatest lists are all about team accomplishments. Best lists are all about the individual, and hyper-focusing on individual abilities, Jim Everett was better than 15% DVOA for a lot of years, more years than John Elway (six for Jim, five for John) in a shorter career.
We can disagree on the merits of DVOA (in a similar way that we would disagree on the merits of WAR, for instance), but that doesn't change that it's the best all-encompassing QB stat that exists for 1989. I have to work with what I have. I would love if I could calculate Jim Everett's career EPA/Play, and see whether or not it's higher than John Elway's, but such a capability does not exist.
However, I've never been one to refuse a challenge, so let's try another approach that seems reasonable. The two most important skills a QB can have are passing accuracy, and pocket presence. These are the two skills most transferrable to a general situation. They're the two that historically keep the best when a QB switches teams, which in a game where so much of what happens is dictated by circumstance, is what we ought to look at, so let's look at how Jim compares in the two most important skills against John Elway again. No hate on John Elway. He's just the worst player that's still generally considered a 'certain Hall of Famer' type from Jim's era, so he has to deal with all this.
My CPOE projection has Jim at 0.1 for his career. It has John at -0.6. If we don't trust my projection, we can use football reference's cmp%+ (103 for Jim for a career, 101 for John),. Both sing the same song that Jim is the more accurate passer, but by approximately half a percent in the completion percentage column. They're so close we can basically call this a saw-off. Then, we move to sack rate, which shows what I've been looking to get to this entire time. Jim is prodigious with his 112 for a whole career, never (not even in his rookie year) falling below 104. Football Reference has no leaderboards for this particular statistic, but Eli Manning has the 17th lowest raw sack rate of all time, and has a career sk%+ of 111, so putting Jim as roughly the 15th best QB all time at avoiding sacks does not seem unreasonable to me. Contrast this with John Elway's career 103 (meaning his career is worse than Jim Everett's worst season at controlling the pocket) makes clear who the more elusive man in the pocket is.
Therefore, we have that the two are approximately equal on the pass accuracy side, and Jim laps John on the pocket presence side, giving Jim Everett a clear win over John Elway in terms of the two most important skills a QB can have in my opinion. Does this pass the test as being a second reasonable approach where Jim Everett looks like a standout?
As far as the main purpose of this article, I would say you're correct. Eli Manning's durability is legendary in an absolute sense, but this article was mostly about the comparison, so in a relative sense, up against this company, Eli's durability really isn't very special at all, because Philip Rivers missed zero career games too. Ben was also pretty good in this sense. He did only start seven more career games than Philp did, despite having a career three seasons longer when discounting Philip's time spent on the bench in 2004 and 2005, but most of this was an additive effect of several knicks and bruises. Miss two games here, three games there.
As far as always correctly checking to the run play, something like that is incredibly hard to account for in a consistent manner. There's a modern allegory in Lamar Jackson, who adds approximately 0.05 EPA/Play to his team's rush offence just by virtue of being who he is, but you cannot simply add this to his EPA/Play figure, because then you'd have to do it for every player across the league, and the sophisticated analysis has been done for Lamar, but it hasn't been done for everybody, so that's just guesswork. I'm not saying that there's not real value there, but I am saying that it's value that's probably best to leave unaccounted for, for statistical consistency reasons.
Thank you for the comment as always David. I hope I did at least something to sway you a little closer to the Jim Everett team. I hate that Jim Rome and all the other nonsense happened to the man who I consider to be the most underrated QB in football history. It all gets in the way, and obfuscates what was a great career on the field individually, even if in a sport like football that values winning more than baseball does, the rosters around him were never good enough to get him into the Hall of Fame.
Thanks, Robbie! It will take quite a bit of time for me to go through this slowly enough to get something from it, but it gives me an idea of where you and your partners in crime are coming from. While I have liitle doubt that Kaepernick 2012 and the early Roethlisberger were not as good as their yards per attempt would suggest, I am surprised that volume of pass attempts is considered quite as harmful as elevating Don Majkowski, 1989, and Everett to that degree suggests. Certainly, those offenses did not rank with the 49ers!
One gets more mileage out of Elway than perhaps one should because his statistics in most years were absolutely divorced from his "elite" reputation. He was my bete noire in 1986, but then in 1987, he really did have a standout all-around year. This his performance plummeted for years, but he was vindicated somewhat in his opinion that the Reeves' offense was holding him back when his performance improved greatly under Mike Shanahan.
So I have always come at things from an adjusted yards per pass attempt framework, the adjutment from interceptions. That's how I know the players and have filed them away in my mind. From your genericaly offered rankings, there is certainly a high correlation in that and in the CPOE, EPA per Play perspective, but sometimes, like with Everett, there is a great divergence.
Your Rivers, Everett comparison I think needs a reality check, if we go by Pro Football Reference's "approximate value." 208 for Rivers' career, 108 for Everett's. Rivers in double digits every year he started, Everett in double digits six times. I don't know how they calculate that, and I imagine it could be easily dismissed and of no incremental value. But I do urge you to at least consult it to get that standard take on a player, should you ever submit your work to someone who is considering you for a paying, prestigious position. I would consult so it so you can understand who your audience is. Maybe those lists that have Trent Greene wherever they have him are an equally good reference, I don't know. I have a "Quarterback Abstract" book at my mother's house I should dig up, which has entries on just about every quarterback.
Supposedly, it was patently obvious that Everett became afraid some point early in his career, and could not keep his eyes downfield. Because of that, he was not a quarterback people respected. I was a fan then, but frankly, the Rams were completely irrelevant, so I don't remember noting this myself. It's interesting that his sack totals always seemed to be fairly low. I actually would think that if you were afraid, it would work against you, and you'd end up getting sacked more. I see that happening a bit with the "scarred" Daniel Jones. You tend to have happen what you're afraid of happening.
Your last paragraph caught my eye immediately. We are in absolute agreement. I tend to think Jim Everett's reputation of being afraid of contact is nothing but trash, for the exact reason you said. I'm not saying it's wrong, but it's clear and obvious that Tom Brady and Patrick Mahomes and Peyton Manning and everybody else that we routinely saw go down without contact were afraid of getting hit too. I'm big on consistency, so if it's okay for Patrick Mahomes to be afraid of contact (which he is, clearly), it must be okay for Jim Everett to be afraid of contact too. If not, our opinion of NFL history needs a serious revision, because fear of contact (or an unwillingness to stand in for risk-reward reasons, which has exactly the same on-field effect) is quite common amongst the all-time greats.
As far as keeping his eyes down the field, Jim's throw depth was never below league average until he got to the Saints. Consistently longer than Joe Montana's. They played in different systems. I understand that, but it's another thing that doesn't seem to make the reputation make sense. When you combine the fact that his throw depths were consistently good, and the fact that he was fantastic at avoiding sacks (which most QBs who 'see ghosts' are typically awful at), it all adds up to me not buying it very much at all. He may very well have been afraid of contact, but suffered no on-field effects from it, so I don't really care.
The thing with QB volume is that the extra attempts are the opposite of low hanging fruit. If we stick to 1989, Joe Montana's 498 total touches are the 498 low hanging fruit touches. The 498 touches that were most likely to succeed, because the team didn't need offence from their QB particularly often. The hypothesis is that Jim Everett or the Majik Man git these same 498 low hanging fruit touches, but had to also add more difficult ones because their team needed them, about 100 in Jim's case and an extra 250 in Don's. Just look at Josh Allen's rate stats from 2022 and 2023, and compare them to 2024. You will see the benefit a player's rate stats get from cutting 100 plays off the sample.
Jim and Don are not quite the same, as Jim was 28% better than league average (per-play) on 602 touches, meanwhile Don was 4% better on 751. Even an extra 150 plays (and 250 plays to Joe Montana) cannot make up this difference, but being above league average on a per-play basis on 750 plays worth of volume was not something that'd been seen before 1989. Taking the statistical hypothesis that his 250 extra touches were on average more difficult than the 500 low hanging fruit that every QB gets, he gets a pass for having a little bit lower per-play value, a big enough pass that my fancy statistics say he was the fifth best QB that year, despite 4% being truly unimpressive on the surface of it. Apply the same logic to Jim Everett's 100 extra plays than Joe and per-play value just 11 percent worse, and I don't mind putting Jim up there.
Quite frankly, it's tough to detach a QB's individual value from an offence's value as a whole in the days when the play distribution (pass-run) used to be a lot closer to 50/50. Great QB play used to correlate much less with great offensive results than it does these days, but the Rams were second in net yards per pass attempt in 1989, behind only the 49ers, and every one of the Rams' dropbacks except five were taken by Jim Everett. Joe gets to share over 100 of the 49ers' with Steve Young, who put up even better per-play results than Joe did.
I don't disagree at all with your opinion on John Elway, but the position I take is that you either accrued value or you didn't, and John finished a 15+ year career with five truly great individual seasons, meanwhile Jim Everett had four ('88, '89, '90, '94) in much fewer tries, and we can quibble about 1992, and even 1995 if you're a really big fan of volume. You saw the comparison I made in terms of individual skill, and I can do one against Jim Kelly too, although it's much more of a preference thing comparing Jim Everett's 0.6 CPOE, 114 sk%+ to Jim Kelly's career 2.6, 104 than the strict preference I had for Jim's skills over John Elway's. If you think sack avoidance is the more important skill (which I do), Jim Everett still looks good here, but there is room for opinion in this comparison, where there wasn't against Elway.
Which takes me to ANY/A. I like that stat a lot as well, but the main problem I have with it is that it weights the negative plays incorrectly. Plugging the average loss on a sack of about six yards into the formula, it weights INTs as being about seven times worse than a sack, when in actuality (we can know this because EPA exists now) INTs are only about 2.5 times worse on average than a sack is. In essence, ANY/A does not weigh negative plays correctly in my opinion. It's got the right idea, but if it took a bit of the weight off turning the ball over it'd be slightly more accurate. This overrating of the importance of turnovers, and underrating of the importance of sack avoidance work in tandem to hurt people whose biggest strength was their sack avoidance, and main weakness was a tendency to turn the ball over, which (not coincidentally) applies to both Trent Green and Jim Everett, and pretty much every other QB who's found themselves underrated throughout history.
I don't want to nakedly accuse people from the past of not valuing sack avoidance enough, because there were surely smart people back then too, but 1989 is the perfect example of the dichotomy. Because of Joe Montana's sk%+ of just 95, the 49ers had to deal with a lot of negative plays that the Rams (with Jim's 112) didn't. This difference was not enough to make Jim the better player on a rate basis, but it was enough to bring Jim's per-play value within 11 percent of Joe's, which I think I would've been spit on had I said at the time, which in my opinion is due to the lack of retroactive importance placed on the skill of avoiding sacks. The flaw in the ANY/A formula is a big culprit for this in my opinion.
The one thing that I will concede is that my lack of respect for career longevity at the QB position is a bit radical, which is why my disrespect for players whose main strength is longevity, and are a bit lacking in per-play value (John Elway, Tom Brady being the big examples) is a bit radical, but in my opinion, the football Hall of Fame brought this upon themselves when they allowed 113 career AV Kurt Warner into the club. Kurt qualified for leaderboards in just nine seasons, so if he can make it, why can't Jim Everett, who (on a per-play basis) is basically the same player as Philip Rivers?
Plus, I would hazard a guess that AV does not properly value sack avoidance either.
Perhaps I can be overly elaborate for the purpose of believing what I want to believe, but teams really didn't throw in place of running the way they do today. Some plays today aren't ambitious. They don't ask a lot of the quarterback, and they can't achieve a lot, unless the play is particualry well designed and catches the opponent offguard. That wasn't the case when people really valued the running game. Then, a pass was a pass. So I don't know how that should change how volume of passing is adjusted for by era, the changing variability of real quarterback result by play. You seemed to be suggesting that EPA/play is higher on short passes than long, so it might work against a Montana, actually. I guess it's also a matter of which passes get added on when volume is increased (long or short, high EPA or low EPA). I don't necessarily see why high passing teams can't add both in equal measure, or why they necessarily have to be adding the ambitious throws to the exclusion of the short ones.
Looking at his approximate value, I don't think the Kurt Warner Hall of Fame selection holds up. I hope the football community doesn't proceed from this precedent. It was particularly egregious because, particularly at the time, even his brilliant peak didn't clearly reflect his abilities. He really didn't get the respect at the time that his stats suggested, even if he won MVP. It was grudging respect. He didn't have obvious physical abilities, and he had so much speed and quickness around him. He sort of duplicated the success in Arizona (although, again, with plenty of help), and from that people just went totally in the other direction, to thinking he was 0% fluke. A complete logical fallacy, it seems to me. And his record in Arizona really doesn't seem nearly as impressive as his record in St. Louis, anyway.
By the way, do you plan on moving your intensive study to other positions?
A thought question is if the Jim Everett high first-round pick was even considered a good one. I think teams stuck with their quarterbacks longer then? Quarterbacks certainly rode the bench longer, typically. They remained the starter longer, I think, because there was no free agency, and free agency still matters to length of run, even if quarterbacks almost never leave in a simple shift. So I don't think Everett's starting 105 games for the Rams means he was in good stead. I know you don't care, I just think it's interesting that most Rams fans might tell you he was a bust and wasn't even worth a first-round draft pick. I don't think he had nearly the enthusiasm and optimism around him, after 1989 anyway, that Pennington, Green, and even your dreaded Jay Cutler had.
It's an interesting idea as far as changing the meaning of volume as we go back in history. It seems to me like you're indicating that every pass was harder back then, which is of course true. That's why old numbers can't compete with modern ones. If 1989 Jim Everett were on my 2024 ranking list, he would be 18th, but back then he was one of the best in the world. We've been talking for days about whether or not he was the very best. It was just a different world these guys were playing in. We agree on that, but people like Jim Everett were trendsetters.
While even the greats like Joe Montana and Steve Young were still touching the ball just 500 times and calling it a full season, it was people like Jim Everett and Don Majkowski showing the league where it was getting ready to go, with a full QB season being 650-750 touches these days.
I have actually thought about this a lot in the context of the 1970s, where 350 total touches used to be considered a full workload for a QB. That was only in 14 games though, so it's about 400 by modern standards (25 touches per game), but still. 350 was about where it was at. In this context, what do we think about players like Fran Tarkenton, who routinely got up to 450, and sometimes even 500? If we assume that passing was more valuable than running back then too, that's 100-150 plays of great value that the Giants/Vikings got that other teams didn't every year. To me, this is an immense part of Fran's overall value. His ability to sustain his great per-play performance over a bigger sample than his contemporaries were able to.
Whether his team needed these extra plays, or whether they just did this because Fran Tarkenton could, it's tough to say in retrospect, although Jack Concannon got to 481 total touches in 1970, and I can guarantee that was not for performance reasons. He was really bad, but touched the ball a lot anyway. That's the life of a QB sometimes, especially one that's always behind like Jack Concannon.
This really is the battle with QB volume. Getting great per-play value over a larger sample is obviously better in a vacuum than in a smaller one, but QB volume tends to come when a team is losing, but this is not always true, because people like Fran Tarkenton (and Jim Everett) often got inordinate amounts of touches on teams that were good. It's a tough balance to strike.
The general assumption is not that extra throws are more difficult from an aDoT sense. The general assumption is that a lot of these extra throws come in situations where the opponent knows you're going to throw, which (in general) makes things more difficult. I say the word general so many times because this changes on a player to player basis, and there is no PBP data to go back and verify in the eras that we're speaking of.
As far as Kurt Warner, lord knows I'm a high peak guy. I think Kurt Warner is a Hall of Fame guy, and Jim Everett is a Hall of Fame guy, and Roger Staubach is a Hall of Fame guy. High peaks for short time spans get more respect from me than the more consistent accumulator types (i.e. John Elway, Tom Brady). As such, I can't argue with Kurt Warner being in the Hall of Fame. It would be disingenuous of me to do that.
The tough thing with Kurt is that it's hard to tell if the St Louis talent lifted him up that high, or if he just never fully recovered from the broken throwing hand that forced him out of town in the first place. Trent Green left St Louis and got better, so it's not like the Rams had a uniformly positive effect on everybody they touched. You could also clearly see the dip when Kurt left and Marc Bulger came in, and I really like Marc Bulger. To me, this means Kurt deserved his two MVP awards. They never got anywhere near the championship again without him, with the same great receivers and some pretty darn good QB play, but I think there was something special there.
Passing accuracy doesn't look flashy, but that's what Kurt had. That's what he did better than anybody else. It's similar to how Peyton Manning never looked especially flashy either. He just dropped back and threw the ball accurately all the time. No heroics required.
As far as Jim Everett being worth it to the Rams, it's a bit unfair to him to even ask the question in this manner. I take the position that it's not his fault the team traded so much for him. If it were just the one first round pick (instead of the two firsts plus a fifth that the Rams traded the Oilers for him), I think he would've got off a lot lighter in this regard, sort of like how nobody talks about Andrew Luck being a failure of a draft pick. If the Colts traded two firsts plus a fifth for Andrew Luck, instead of him just being the one draft pick, I don't think he would be let off the hook so easily either, even if I change absolutely nothing about how he played, and Andrew Luck was never anywhere near the player Jim Everett was. I would take just 1988 and 1989 over Andrew Luck's whole career.
So was he worth what the Rams paid for him? In 1988 and 1989 he was, but afterwards? Probably not, but I think if Jim were just a straight up third overall pick, he would get a lot less flak. He would have the Andrew Luck reputation of being far from a success as a top three pick QB (only a championship can make you a success, in the eyes of most, and Jim got closer to one than Andrew Luck ever did), but not necessarily a failure either.
Fans tend to hold things like this against players, but Jim Everett didn't make the trade. He doesn't deserve to get blamed for it. Bryce Young is in a similar spot right now. He's coming along nicely in my opinion, but Carolina fans are frustrated anyways, because he did not cost one pick. He cost several. This is the same spot Jim was put into.
I'm getting there as far as spreading my wings to cover other offensive positions. I'm looking deeper and deeper every day. I will never study defence this intensively though, because individual contributions are so difficult to separate from those of the team as a whole, which makes for both less accurate statistical understanding and a less interesting story. I prefer to study the defence of teams as a whole.
Any article of mine typically takes multiple afternoons of work.
My writing pace is about one hour to 800 words, which is quite fast from what I understand, but on something like this that's 11K words long, it still eats up a lot of time. The only thing that takes slightly longer than that are the lists, which take longer because I have to learn the basics of 20 different stories instead of just one.
It's really bad for business, because if I took one of those afternoons every week and dedicated it to promotion instead of the finer details of whatever I'm writing, my subscriber count could be triple what it is now, but I'm not here for business. I'm here to write quality stuff. My interest is not really whether anybody reads it or not.
So yes. That's the answer. About an hour to 800 words, which for how long most of my things are means multiple afternoons worth of writing every week.
I think going through the questions is a good way to help determine who should be in the Hall and is something I've seen only one other time. While I agree with your conclusions, I don't think the voters will. I think Ben being in the good good stats, lots of winning category will get him in and I think the mythos created around Eli is too strong to overcome for Rivers. I think if Rivers would have won an MVP and won in 06 and 07 against the Patriots, he would be in.
I would also be interested in a further dive into Jim Everett, because even based on NY/A and ANY/A, he has only 2 seasons in the top 5 and 3 other seasons in the top 10. In your formula, what is giving him this boost compared to other numbers?
I truly think Philip (and not Eli, key point) would've been the one to stop the 2007 Patriots, if he'd been given a real chance at it. He'd just played perhaps the best playoff game of all time the week before, but then tears his ACL in the fourth quarter. You can't play the QB position moving one leg at a time. The Chargers should've just started Billy Volek, and I'm not entirely sure he wouldn't have beaten the Patriots. He certainly was a better QB than the -0.05 EPA/Play, -5 CPOE Eli Manning in 2007.
What I'm trying to get at is that the Chargers lost, and Philip Rivers was the reason they lost, in both 2006 and 2007, but there are extenuating circumstances (worst luck of all time one season, trying to play on a torn ACL the next) that ought to be taken into account there. This is the Philip Rivers what-if. We've talked about the Chad Pennington what-if before, and this is the Philip Rivers version.
In retrospect, I can quite confidently say the 2007 Patriots were dead. It took everything they had to beat the Jaguars in the second round game harder than anybody remembers it being, and after working so hard to finish an undefeated regular season they were just out of gas, particularly on offence. They couldn't move the ball at all against either the Chargers or the Giants, and it was the luck of fools and Patriots that they ran into a hobbled (to put it lightly) Philip Rivers and one of the worst QBs in the league in Eli Manning, allowing them to nearly pull it off anyway. Like I said in the article, I believe any of the top 11 QBs in the NFL (from Donovan McNabb upwards) would've beaten the Patriots in those 2007 playoffs, some by multiple scores.
What if Philip Rivers makes it through the Colt game unscathed? That's the what-if. It's purely supposition on my part, but the Patriots were limping, looking for somebody to lose to. The real Philip Rivers would've proved more than good enough for that, and with that win on his (and the Chargers') resume, where do they go from there? It almost surely gets them out of this playoff loser mentality, that they still remain in today, because they almost surely defeat the New York Giants in the SB with that defence they had in 2007. I'm not sure if the Giants would've scored a point that day. They only put up 17 on a pretty mid 2007 Patriot defence.
With the Chargers as 2007 Super Bowl champions (a reality that I think was closer than anybody thinks), the roster still takes a pretty big nosedive from there, but had enough changed in the interim to make things different in SD? The team would almost certainly still be in SD, and Philip Rivers would be the cause of that. He would certainly be in the HoF.
All of that gone, as a result of just one knee injury. Onto Jim Everett.
As with all my arguments, Jim's relies on peak performance. Let's go season by season.
Jim's prime begins in 1988, where despite a rush offence not worth using that got used A LOT, putting him in a lot of bad positions, Jim posted 1293 football outsiders DYAR, which puts him in the mix with Dan Marino and Boomer Esiason for the most in the league, all almost 500 ahead of anybody else. His rate stats are fantastic also, finishing fourth in DVOA and fifth in ANY/A despite an astonishing (for 1988) 609 touches. Mix this in with Boomer touching the ball only 491 times in 1988, you find that Jim is second on my QB tier list that year.
Onto 1989, where the rush offence is even worse, and gets even less touches, leaving Jim to pick up even more of the offensive workload. He parlays this into being the only QB in the NFL even close enough to see Joe Montana's dust in DYAR, and being much closer to the pack but still second place in DVOA and ANY/A. Combining this with the fact that he played all 16, compared to Joe Montana missing three starts in 1989, leads me to put Jim number one that season. He did not even make the Pro Bowl this season by the way, but he was without doubt the best QB who started all 16.
1990. Jim's only good teams that he will ever play on are already behind him, but he's still fantastic. The rush offence is horrendous. The defence is horrendous. His receivers are not what they used to be, but Jim is still chucking. Ninth in ANY/A and ninth in DVOA are not great on a rate basis, but on a volume basis on a team that DESPERATELY needed the volume, Jim was top five in DYAR again, and became the first player in NFL history not named Dan Marino to put up 600 touches three seasons in a row, a level of durability that's very difficult to maintain in this era. Combine this with a lot of the names above him on the rate lists being guys like Jay Schroeder (430 touches) and Jim Kelly (418 touches), and Jim finds the top five for the third season in a row.
Perhaps all this work got to him, because in 1991 he's an uninspiring 11th, even despite being about 16th on a per-play basis, due to the same 'all the guys better than him in the rate stats have 400 plays to his 600, so what would they have done on the horrendous Rams?' dichotomy.
He's trying Marc. He's really trying. By the time we get to 1992, Henry Ellard and Flipper Anderson are not what they used to be, and as the Rams are truly beginning to fall into the toilet as a franchise, there are no other offensive players you've ever heard of. Nevertheless, 6th in DYAR, 6th in DVOA on that typical Jim Everett volume land him 6th on my 1992 tier list.
He finally misses time in 1993, and is horrendous when he does play, so finally the Rams (who had not drafted a QB throughout this whole process of four losing seasons in a row, an important acknowledgement both contemporaneously and in retrospect that QB was not the problem in LA) decide to make a move, trading Jim Everett for a seventh round draft pick (LOL). Immediately after this, that franchise would fall into the pits of despair, moving to St Louis for their trouble.
Unfortunately, Jim got to New Orleans just as they were falling into the pits of despair. Challenge Marc. Name me one New Orleans Saint (other than Jim Everett) from 1994 to 1996. The rush offence was not worth using. The defence was horrendous, and Jim carried all of this to a 7-9 record with his 1101 DYAR (3rd), and 0.193 DVOA (5th) on that typical Jim Everett volume, landing him in the top five of my tier list again.
1995 is much the same story, with the same Saints roster, the same 7-9 record, but this time with 1168 DYAR (5th) and 0.190 DVOA (8th). With improvement leaguewide, roughly the same numbers only got him 7th place this time. He tried it again in 1996, but it was over, and you didn't hear much about Jim Everett after that.
That's eight prime years, and his tier list finishes in those eight years are 2, 1, 5, 11, 6, 29 (in the injury-shortened 1993), 4, 7, almost all of it on awful rosters, perpetually putting him in bad positions. Four top five finishes plus two additional top ten finishes (that was almost three) in an eight year prime has very rarely been done. He's not Peyton Manning (top five every year from 2003-2010), but look at the prime of Hall of Famer Jim Kelly, who played around the same time as Jim: 11, 9, 11, 9, 7, 3, 5, 9.
That's not Jim Everett calibre.
ANY/A hates Jim Everett, because he turned the ball over a lot, and that stat really overrates the importance of preventing turnovers, but stats like DYAR love his volume, and stats like DVOA back up that it wasn't all volume. He was top five on a rate basis as well several times, which gets exponentially harder when your awful rosters consistently ask you to crank out so much volume. Ask Josh Allen in 2024 the statistical benefit of decreasing the volume a little bit.
Adding all of this together, we find Jim Everett's position on my all-time QB ranking, which we all know is peak-performance based. If I had to have a QB for a four season stretch, I would take Jim Everett with my 15th pick. Behind Tom Brady (12th), Kurt Warner (13th), and Philip Rivers (14th), but ahead of Brett Favre (19th), Warren Moon (29th), Boomer Esiason (31st), and John Elway (33rd).
People tend to laugh when I bring this up, and point out that Jim's career record is 64-89, as if I don't know that. He only won two career playoff games. He only played in four of them, but for a man who spent a whole career having to deal with such trash rosters, he turned a lot of it into chicken salad. This is not basketball. Jim was not putting up great stats on bad teams just because he could like he was Zach LaVine or something. Football doesn't work like that. He was put into horrendous positions over and over and over again, and he showed his mettle.
I often wonder what could've been if Jim Kelly came into the league on the 1986 Rams, and Jim Everett came in on the 1986 Bills. Would it still be the case that Kelly is in the Hall of Fame, while Everett isn't? In my opinion, absolutely not, but we can only guess about questions like that.
So there. That's a surface level look into the career of Jim Everett. NY/A doesn't like him because it discriminates against bad rosters without a lot of RAC. ANY/A doesn't like him because he turned the ball over a lot, but for somebody like me who takes pains to discriminate against neither bad rosters nor people who turn the ball over a lot, I think Jim Everett is the 15th best QB of all time.
Thanks for that explanation on Jim. In addition to DVOA and DYAR liking him, his teams passing first downs are pretty high which I think helps explains the gap between his yards based numbers and his DVOA based numbers. Also, what led you to choosing 4-years instead of 3 or 5?
It's not actually a strict four. It's a minimum of four seasons that MUST be included. Beyond that, up to three seasons are allowed to be excluded entirely, for falloff or rookie reasons. However, three seasons do not have to be excluded if keeping them will improve the score. This generally leaves room to ignore the falloff of a player's career, or a few bad seasons to start it, but not both of these things.
It's basically a peak performance metric, with very little sympathy for extended falloffs or taking a long time to get going. The perfect player according to my formula is somebody like Peyton Manning or Dan Marino, who came into the league a finished product, and fell off quickly. This makes sense though in my opinion, because that's very few bad plays in a career.
The same does not apply to somebody like Tom Brady, because even once you exclude 2001, 2002, and 2019, you're still not out of sub-tenth place seasons. Therefore, he fares worse than a lot of people think he would, mostly because he took so long to get going. On a career-based ranking like mine, this hurts, and it should hurt in my opinion.
Because of this mechanic of punishing long lead-in times, and punishing long falloffs, it's not a true prime performance metric. It's a career performance metric, but peak performance is basically what it captures. It basically just captures peak performance, with punishment built in for having a lot of off years, be they in the front or on the end or in the middle, my formula doesn't care. This is in contrast to the public eye, which generally tends to just ignore off-years in the middle as if they didn't happen at all.
The impetus for the cutoff being four is actually Kurt Warner, who made the Hall of Fame with just four truly great seasons. Since you can be a Hall of Famer now with just four great years, I built a formula that doesn't discriminate between four great seasons and twelve of them.
With my excluding seasons method, Kurt's ranking on this list doesn't have to deal with his 2002, 2004, or 2005, instead using only his 1999, 2000, 2001, 2007, 2008, and 2009 seasons to calculate his performance level, and come to his ranking of tenth place, just above Tom Brady. At their peaks, they were about the same, but Kurt came into the league finished already, compared to Tom taking a few years to get there. That's the only difference.
If we look at somebody like Jim Everett, the formula sees him as his 1988, 1989, 1990, 1991, 1992, 1994, and 1995 seasons, entirely ignoring the lead-in year in 1987, the injury shortened 1993, and the falloff in 1996. If there were more than three bad qualified seasons, I would've had to punish him for them, but there weren't.
You may say this punishes newer players, who often are forced onto the field sooner, and therefore have a lot more bad plays than some of the older guys, and I say that's a feature. Not a bug. The newer guys who play bad as rookies and second year players just have more bad plays in their careers. I don't feel bad for my performance formula punishing them for that, even if it's not their fault since they shouldn't be playing in the first place.
There are also some other things about the aggregation method that put extreme value on the top-end performance while perhaps nullifying some of the bad, that I don't think I'm able to explain in a Substack comment box. One part I can explain is that there's quite a hefty buff a career gets for ever having been the best QB in the NFL in any season, which one can argue with, because it's relative. For instance, Carson Palmer gets it but Drew Brees does not, because Drew's prime basically coincides with the best parts of Brady and Manning's careers, and Carson wrangled his way into the top spot in a rather weak QB season in 2015. The strength of this number one position basically carries Carson on its own to 53rd all-time, a position he would've been nowhere near without the first place buff.
Anyways. I've gotten way into the weeds. Basically, it's an augmented peak performance ranking that punishes careers with a lot of bad plays in them (i.e. Tom Brady). I chose four seasons because that's how many top ten seasons Kurt Warner had, and he is currently in the Hall of Fame. Two of those four seasons are first places, which drives my formula crazy and pushes him all the way up to tenth place, but you know what? First place twice is the same number that Tom Brady had, so I think it's right to reward it that much. My method is designed to answer questions like this, and it's also designed to answer questions like Daunte Culpepper, who had a fantastic season in 2000, then two bad ones in 2001 and 2002, then two more great ones in 2003 and 2004, and a whole bunch more bad ones after that. Most formulas (like Kevin Cole's, which was my main problem with his ranking methodology) have trouble with stop-start careers like this, but mine does not.
I have just spent the last 10 minutes trying to think of a way to describe for the generations after me how outlandish Jim Everett as a Hall of Famer sounds to those who followed his career. This would be the hottest hot take ever offered on ESPN and a true tribute to free speech were you allowed to make the argument. Except for his third and fourth seasons as a starter, he was not even regarded as a good player, and I think Henry Ellard and Willie Anderson got a lot of the credit for those good seasons. He could throw the deep ball, I'll give you that. As you know, he was considered a laughingstock afterwards, with the Jim Rome thing commanding attention. One can certainly see how that formed an unfair narrative. But Everett as a Hall of Famer?
I think what it really comes down to is what statistics you use and just how you choose to balance the good and bad seasons you get from those statistics. You have your favorite statistics and your way of balancing peak and career, and you follow this method to the end of the earth. Applied to historical players, when the statistics traditionaly haven't been, they lead to particularly (ahem) unconventional ratings. But when there is such a gulf between, say, the quarterback's won-loss record (64-89) and his number of Pro Bowls (1) and your rating, I think you have a responsibility to be extra analytical and self-critical and try to see what if anything you may be missing. If different perspectives don't validate your particular method, that is troubling. Do common sense and your argument fit together?
My major project 15 years ago was rating all quarterbacks historically by their teams' yards per drive in years when they were the starter. Everett was absolutely unremarkable in that. I also have his composite league interception rate as compared to the league's as mediocre. It seems unlikely to me that the running game would move the needle enough to have cost his offenses so much in yards per drive. Even after 1989, Everett did put up some yards, but it is easy to do that when you are losing most games and throwing a lot. He never averaged better than 7.2 yards an attempt after '89. Maybe there is more to be learned about the deficiencies in the particular statistics you are using than about Jim Everett from them? If you scramble the deck, try something else somewhat reasonable and still have Everett as a standout, then we can talk some more. But you're a talented guy, and I do worry for your reputation when you say something that is just regarded as crazy, and you don't seem to have any notion that it is taken that way, and that you have no intellectual company in the opinion.
I would rate Rivers, Roethlisberger, and Manning in the same order that you do. Where Manning was better than you give him credit for was in his durability. He never missed a game because of injury. It didn't seem to be luck, not that that matters. It was amazing the hits he got up from. The other thing the Giants' coaches said was that he was expert in audibling to running plays, and so deserves some of the credit for how well the Giants ran through 2008 or 2009. Not that it adds up to a Hall of Fame career, but just to make a full accounting. To me, rather than being like a Bumgarner, he was like an old-time pitcher who was 250-240 for his career, or something.
You know I've ever been one to shy away from a hot take David, as long as it's positive about a player. This is no different.
I have had that same internal debate about Jim Everett with myself multiple times. In essence, it comes down to how much QB volume is worth in the years before QBs touching the ball 600 times became standard. It's sort of like the worth of a volume goaltender in ice hockey. I'm not sure if you watch or care about ice hockey, but it's the age old question. What's better, 45 starts at an 80 GA%-, or 70 starts at 100? I suppose an imperfect baseball analogy could be whether you'd prefer 23 starts of 120 ERA+ or 33 starts of 100.
In football, it's the exact same thing. Would you prefer 498 plays of being 39% better than the average QB (1989 Joe Montana) or 602 plays of being 28% better than average (1989 Jim Everett)? Modern statistical understanding likes Jim Everett better. This is why Jim Everett is number one on my QB ranking list for 1989, and not Joe Montana. That sounds like blasphemy to those who lived through it, but I'm not trying to gauge the opinions of those who lived through it. I didn't live through it. It would only be a guess on my part, so I elect not to even try.
It's not meant to say that Jim Everett was perceived as being better than Joe Montana in 1989. It's to say that if 1989 happened in 2024, modern football people would like Jim Everett's season more, even if they didn't like it more then, which they clearly didn't because Joe won the MVP. I believe when I was talking about Jim Everett I was free from the constraints of the word regarded in my formulation, so I felt free to discuss 1989 as if it happened yesterday, unconstrained by the views of the time. This is much like how with modern statistical understanding Dave Stieb is dramatically better than Jack Morris. I can say that comfortably. Nevertheless, Jack is in the Hall of Fame, while Dave is not. If you interpret that fact as a mistake (which I do, and a lot of others do as well), it's an implicit acknowledgement that modern statistical understanding has a place in debates like this, at least in some capacity.
I entirely understand that Jim made the Pro Bowl just one time, but it's got more to do with his place and time than anything. For an example of the opposite, look no further than Donovan McNabb, who I don't think even yourself would assert is anywhere near as good as Jim Everett. Nevertheless, he made five Pro Bowls, due to being right in the middle of perhaps the weakest conference in NFL history, which is the NFC in the early-mid 2000s. This is how Donovan can share a league with somebody like Trent Green, and make the Pro Bowl three times more. It's not a fair fight.
The late 80s-early 90s NFC was a bloodbath. Mix this in with galling decisions like somehow Randall Cunningham going over him in both 1988 and 1989, despite having worse individual numbers at the head of a worse team (shown beautifully when Jim and the Rams spanked the Eagles in the 1989 playoff game they played), and you can see why I respect Pro Bowl count so little. It's both uneven across history and incorrect in its own place and time, both due to conference barriers and horrifying decisions like the Randall Cunningham one.
As far as team record, it's not Jim's fault that he spent the whole of his post-1989 career stranded on some of the worst rosters in the entire NFL. When I get myself into Tom Brady-Peyton Manning debates, and I bring up the fact that Tom had so much more defensive help than Peyton did, what do I hear constantly? 'Players don't build the roster.' Now, I'm invoking that argument here. It was not Jim's choice that his Rams deteriorated into the worst roster in the league right as he was (theoretically) entering his prime. He did not choose to get traded to the NO Saints conveniently right after their roster had fallen apart, and not in time to perhaps save it.
Just look at the 1991 Rams as an example. Behind all the time, theoretically constantly in disadvantageous positions due to always having to pass the ball, this is still a pass offence that ranked eighth in the league in net yards per attempt. Eighth is not spectacular, but it's eighth. Mix this in with his propensity to turn the ball over a little bit, you still get that he was the 11th best QB in 1991, according to me. 3-13 record or not. I don't know why they scored so few points. There's no PBP data for me to cross reference, but yardage wise, this was still a passing offence to be reckoned with. 1992 is similar, with slightly fewer yards, slightly fewer turnovers, slightly more points, etc..
You are right that most of Jim's career was spent dragging doomed teams to above average passing offences. This entails a lot of volume, and a lot of volume on 17% DVOA play (1992 Jim Everett) makes you a really valuable player. What would you say about a pitcher who has the fourth most innings in the league, and finishes sixth in ERA+, but somehow goes 8-10 on the year? You would call him a great player, who needs a better opportunity to let his talent shine. That's what I'm saying about Jim Everett, for basically every season of his career. This even continues as late as 1995, where Jim's DVOA is an astounding 19%, on the fourth most touches in the league. Absolutely nobody was talking about Jim Everett in 1995, but with stats like this they should've been, and would've been if 1995 were the season that's just finished, with modern statistical understanding.
Do take note that the league leader in EPA/Play in 1995 was Erik Kramer with his 0.166. Offence was just not the same thing back then as it was today. The best QB in the NFL back then is about as good as Kyler Murray now. The goalposts have moved, so I'm not saying that you can drop Jim Everett (or anybody whose prime was in the late 80s) into the league now and he would thrive, but relative to his era, I like him better than anybody except Joe Montana, Steve Young, and Dan Marino (and Dan Fouts, who he had an extremely limited crossing of paths with I believe).
Also, I don't find my rankings to be that conventional. Applying the same methodology to all of NFL history since the merger reveals a top ten QBs of all time list featuring the following names (unordered): Drew Brees, Aaron Rodgers, Peyton Manning, Dan Marino, Steve Young, Joe Montana, Fran Tarkenton, Roger Staubach, Patrick Mahomes, Dan Fouts. I don't believe this is that unconventional of a ranking. The one thing that people disagree with is that it puts all of its weight on scoring and moving the football, and none of its weight on winning (hence no Tom Brady or John Elway), but it's not a greatest list. It's a best list. Greatest lists are all about team accomplishments. Best lists are all about the individual, and hyper-focusing on individual abilities, Jim Everett was better than 15% DVOA for a lot of years, more years than John Elway (six for Jim, five for John) in a shorter career.
We can disagree on the merits of DVOA (in a similar way that we would disagree on the merits of WAR, for instance), but that doesn't change that it's the best all-encompassing QB stat that exists for 1989. I have to work with what I have. I would love if I could calculate Jim Everett's career EPA/Play, and see whether or not it's higher than John Elway's, but such a capability does not exist.
However, I've never been one to refuse a challenge, so let's try another approach that seems reasonable. The two most important skills a QB can have are passing accuracy, and pocket presence. These are the two skills most transferrable to a general situation. They're the two that historically keep the best when a QB switches teams, which in a game where so much of what happens is dictated by circumstance, is what we ought to look at, so let's look at how Jim compares in the two most important skills against John Elway again. No hate on John Elway. He's just the worst player that's still generally considered a 'certain Hall of Famer' type from Jim's era, so he has to deal with all this.
My CPOE projection has Jim at 0.1 for his career. It has John at -0.6. If we don't trust my projection, we can use football reference's cmp%+ (103 for Jim for a career, 101 for John),. Both sing the same song that Jim is the more accurate passer, but by approximately half a percent in the completion percentage column. They're so close we can basically call this a saw-off. Then, we move to sack rate, which shows what I've been looking to get to this entire time. Jim is prodigious with his 112 for a whole career, never (not even in his rookie year) falling below 104. Football Reference has no leaderboards for this particular statistic, but Eli Manning has the 17th lowest raw sack rate of all time, and has a career sk%+ of 111, so putting Jim as roughly the 15th best QB all time at avoiding sacks does not seem unreasonable to me. Contrast this with John Elway's career 103 (meaning his career is worse than Jim Everett's worst season at controlling the pocket) makes clear who the more elusive man in the pocket is.
Therefore, we have that the two are approximately equal on the pass accuracy side, and Jim laps John on the pocket presence side, giving Jim Everett a clear win over John Elway in terms of the two most important skills a QB can have in my opinion. Does this pass the test as being a second reasonable approach where Jim Everett looks like a standout?
As far as the main purpose of this article, I would say you're correct. Eli Manning's durability is legendary in an absolute sense, but this article was mostly about the comparison, so in a relative sense, up against this company, Eli's durability really isn't very special at all, because Philip Rivers missed zero career games too. Ben was also pretty good in this sense. He did only start seven more career games than Philp did, despite having a career three seasons longer when discounting Philip's time spent on the bench in 2004 and 2005, but most of this was an additive effect of several knicks and bruises. Miss two games here, three games there.
As far as always correctly checking to the run play, something like that is incredibly hard to account for in a consistent manner. There's a modern allegory in Lamar Jackson, who adds approximately 0.05 EPA/Play to his team's rush offence just by virtue of being who he is, but you cannot simply add this to his EPA/Play figure, because then you'd have to do it for every player across the league, and the sophisticated analysis has been done for Lamar, but it hasn't been done for everybody, so that's just guesswork. I'm not saying that there's not real value there, but I am saying that it's value that's probably best to leave unaccounted for, for statistical consistency reasons.
Thank you for the comment as always David. I hope I did at least something to sway you a little closer to the Jim Everett team. I hate that Jim Rome and all the other nonsense happened to the man who I consider to be the most underrated QB in football history. It all gets in the way, and obfuscates what was a great career on the field individually, even if in a sport like football that values winning more than baseball does, the rosters around him were never good enough to get him into the Hall of Fame.
Thanks, Robbie! It will take quite a bit of time for me to go through this slowly enough to get something from it, but it gives me an idea of where you and your partners in crime are coming from. While I have liitle doubt that Kaepernick 2012 and the early Roethlisberger were not as good as their yards per attempt would suggest, I am surprised that volume of pass attempts is considered quite as harmful as elevating Don Majkowski, 1989, and Everett to that degree suggests. Certainly, those offenses did not rank with the 49ers!
One gets more mileage out of Elway than perhaps one should because his statistics in most years were absolutely divorced from his "elite" reputation. He was my bete noire in 1986, but then in 1987, he really did have a standout all-around year. This his performance plummeted for years, but he was vindicated somewhat in his opinion that the Reeves' offense was holding him back when his performance improved greatly under Mike Shanahan.
So I have always come at things from an adjusted yards per pass attempt framework, the adjutment from interceptions. That's how I know the players and have filed them away in my mind. From your genericaly offered rankings, there is certainly a high correlation in that and in the CPOE, EPA per Play perspective, but sometimes, like with Everett, there is a great divergence.
Your Rivers, Everett comparison I think needs a reality check, if we go by Pro Football Reference's "approximate value." 208 for Rivers' career, 108 for Everett's. Rivers in double digits every year he started, Everett in double digits six times. I don't know how they calculate that, and I imagine it could be easily dismissed and of no incremental value. But I do urge you to at least consult it to get that standard take on a player, should you ever submit your work to someone who is considering you for a paying, prestigious position. I would consult so it so you can understand who your audience is. Maybe those lists that have Trent Greene wherever they have him are an equally good reference, I don't know. I have a "Quarterback Abstract" book at my mother's house I should dig up, which has entries on just about every quarterback.
Supposedly, it was patently obvious that Everett became afraid some point early in his career, and could not keep his eyes downfield. Because of that, he was not a quarterback people respected. I was a fan then, but frankly, the Rams were completely irrelevant, so I don't remember noting this myself. It's interesting that his sack totals always seemed to be fairly low. I actually would think that if you were afraid, it would work against you, and you'd end up getting sacked more. I see that happening a bit with the "scarred" Daniel Jones. You tend to have happen what you're afraid of happening.
Your last paragraph caught my eye immediately. We are in absolute agreement. I tend to think Jim Everett's reputation of being afraid of contact is nothing but trash, for the exact reason you said. I'm not saying it's wrong, but it's clear and obvious that Tom Brady and Patrick Mahomes and Peyton Manning and everybody else that we routinely saw go down without contact were afraid of getting hit too. I'm big on consistency, so if it's okay for Patrick Mahomes to be afraid of contact (which he is, clearly), it must be okay for Jim Everett to be afraid of contact too. If not, our opinion of NFL history needs a serious revision, because fear of contact (or an unwillingness to stand in for risk-reward reasons, which has exactly the same on-field effect) is quite common amongst the all-time greats.
As far as keeping his eyes down the field, Jim's throw depth was never below league average until he got to the Saints. Consistently longer than Joe Montana's. They played in different systems. I understand that, but it's another thing that doesn't seem to make the reputation make sense. When you combine the fact that his throw depths were consistently good, and the fact that he was fantastic at avoiding sacks (which most QBs who 'see ghosts' are typically awful at), it all adds up to me not buying it very much at all. He may very well have been afraid of contact, but suffered no on-field effects from it, so I don't really care.
The thing with QB volume is that the extra attempts are the opposite of low hanging fruit. If we stick to 1989, Joe Montana's 498 total touches are the 498 low hanging fruit touches. The 498 touches that were most likely to succeed, because the team didn't need offence from their QB particularly often. The hypothesis is that Jim Everett or the Majik Man git these same 498 low hanging fruit touches, but had to also add more difficult ones because their team needed them, about 100 in Jim's case and an extra 250 in Don's. Just look at Josh Allen's rate stats from 2022 and 2023, and compare them to 2024. You will see the benefit a player's rate stats get from cutting 100 plays off the sample.
Jim and Don are not quite the same, as Jim was 28% better than league average (per-play) on 602 touches, meanwhile Don was 4% better on 751. Even an extra 150 plays (and 250 plays to Joe Montana) cannot make up this difference, but being above league average on a per-play basis on 750 plays worth of volume was not something that'd been seen before 1989. Taking the statistical hypothesis that his 250 extra touches were on average more difficult than the 500 low hanging fruit that every QB gets, he gets a pass for having a little bit lower per-play value, a big enough pass that my fancy statistics say he was the fifth best QB that year, despite 4% being truly unimpressive on the surface of it. Apply the same logic to Jim Everett's 100 extra plays than Joe and per-play value just 11 percent worse, and I don't mind putting Jim up there.
Quite frankly, it's tough to detach a QB's individual value from an offence's value as a whole in the days when the play distribution (pass-run) used to be a lot closer to 50/50. Great QB play used to correlate much less with great offensive results than it does these days, but the Rams were second in net yards per pass attempt in 1989, behind only the 49ers, and every one of the Rams' dropbacks except five were taken by Jim Everett. Joe gets to share over 100 of the 49ers' with Steve Young, who put up even better per-play results than Joe did.
I don't disagree at all with your opinion on John Elway, but the position I take is that you either accrued value or you didn't, and John finished a 15+ year career with five truly great individual seasons, meanwhile Jim Everett had four ('88, '89, '90, '94) in much fewer tries, and we can quibble about 1992, and even 1995 if you're a really big fan of volume. You saw the comparison I made in terms of individual skill, and I can do one against Jim Kelly too, although it's much more of a preference thing comparing Jim Everett's 0.6 CPOE, 114 sk%+ to Jim Kelly's career 2.6, 104 than the strict preference I had for Jim's skills over John Elway's. If you think sack avoidance is the more important skill (which I do), Jim Everett still looks good here, but there is room for opinion in this comparison, where there wasn't against Elway.
Which takes me to ANY/A. I like that stat a lot as well, but the main problem I have with it is that it weights the negative plays incorrectly. Plugging the average loss on a sack of about six yards into the formula, it weights INTs as being about seven times worse than a sack, when in actuality (we can know this because EPA exists now) INTs are only about 2.5 times worse on average than a sack is. In essence, ANY/A does not weigh negative plays correctly in my opinion. It's got the right idea, but if it took a bit of the weight off turning the ball over it'd be slightly more accurate. This overrating of the importance of turnovers, and underrating of the importance of sack avoidance work in tandem to hurt people whose biggest strength was their sack avoidance, and main weakness was a tendency to turn the ball over, which (not coincidentally) applies to both Trent Green and Jim Everett, and pretty much every other QB who's found themselves underrated throughout history.
I don't want to nakedly accuse people from the past of not valuing sack avoidance enough, because there were surely smart people back then too, but 1989 is the perfect example of the dichotomy. Because of Joe Montana's sk%+ of just 95, the 49ers had to deal with a lot of negative plays that the Rams (with Jim's 112) didn't. This difference was not enough to make Jim the better player on a rate basis, but it was enough to bring Jim's per-play value within 11 percent of Joe's, which I think I would've been spit on had I said at the time, which in my opinion is due to the lack of retroactive importance placed on the skill of avoiding sacks. The flaw in the ANY/A formula is a big culprit for this in my opinion.
The one thing that I will concede is that my lack of respect for career longevity at the QB position is a bit radical, which is why my disrespect for players whose main strength is longevity, and are a bit lacking in per-play value (John Elway, Tom Brady being the big examples) is a bit radical, but in my opinion, the football Hall of Fame brought this upon themselves when they allowed 113 career AV Kurt Warner into the club. Kurt qualified for leaderboards in just nine seasons, so if he can make it, why can't Jim Everett, who (on a per-play basis) is basically the same player as Philip Rivers?
Plus, I would hazard a guess that AV does not properly value sack avoidance either.
Worthwhile analysis.
Perhaps I can be overly elaborate for the purpose of believing what I want to believe, but teams really didn't throw in place of running the way they do today. Some plays today aren't ambitious. They don't ask a lot of the quarterback, and they can't achieve a lot, unless the play is particualry well designed and catches the opponent offguard. That wasn't the case when people really valued the running game. Then, a pass was a pass. So I don't know how that should change how volume of passing is adjusted for by era, the changing variability of real quarterback result by play. You seemed to be suggesting that EPA/play is higher on short passes than long, so it might work against a Montana, actually. I guess it's also a matter of which passes get added on when volume is increased (long or short, high EPA or low EPA). I don't necessarily see why high passing teams can't add both in equal measure, or why they necessarily have to be adding the ambitious throws to the exclusion of the short ones.
Looking at his approximate value, I don't think the Kurt Warner Hall of Fame selection holds up. I hope the football community doesn't proceed from this precedent. It was particularly egregious because, particularly at the time, even his brilliant peak didn't clearly reflect his abilities. He really didn't get the respect at the time that his stats suggested, even if he won MVP. It was grudging respect. He didn't have obvious physical abilities, and he had so much speed and quickness around him. He sort of duplicated the success in Arizona (although, again, with plenty of help), and from that people just went totally in the other direction, to thinking he was 0% fluke. A complete logical fallacy, it seems to me. And his record in Arizona really doesn't seem nearly as impressive as his record in St. Louis, anyway.
By the way, do you plan on moving your intensive study to other positions?
A thought question is if the Jim Everett high first-round pick was even considered a good one. I think teams stuck with their quarterbacks longer then? Quarterbacks certainly rode the bench longer, typically. They remained the starter longer, I think, because there was no free agency, and free agency still matters to length of run, even if quarterbacks almost never leave in a simple shift. So I don't think Everett's starting 105 games for the Rams means he was in good stead. I know you don't care, I just think it's interesting that most Rams fans might tell you he was a bust and wasn't even worth a first-round draft pick. I don't think he had nearly the enthusiasm and optimism around him, after 1989 anyway, that Pennington, Green, and even your dreaded Jay Cutler had.
It's an interesting idea as far as changing the meaning of volume as we go back in history. It seems to me like you're indicating that every pass was harder back then, which is of course true. That's why old numbers can't compete with modern ones. If 1989 Jim Everett were on my 2024 ranking list, he would be 18th, but back then he was one of the best in the world. We've been talking for days about whether or not he was the very best. It was just a different world these guys were playing in. We agree on that, but people like Jim Everett were trendsetters.
While even the greats like Joe Montana and Steve Young were still touching the ball just 500 times and calling it a full season, it was people like Jim Everett and Don Majkowski showing the league where it was getting ready to go, with a full QB season being 650-750 touches these days.
I have actually thought about this a lot in the context of the 1970s, where 350 total touches used to be considered a full workload for a QB. That was only in 14 games though, so it's about 400 by modern standards (25 touches per game), but still. 350 was about where it was at. In this context, what do we think about players like Fran Tarkenton, who routinely got up to 450, and sometimes even 500? If we assume that passing was more valuable than running back then too, that's 100-150 plays of great value that the Giants/Vikings got that other teams didn't every year. To me, this is an immense part of Fran's overall value. His ability to sustain his great per-play performance over a bigger sample than his contemporaries were able to.
Whether his team needed these extra plays, or whether they just did this because Fran Tarkenton could, it's tough to say in retrospect, although Jack Concannon got to 481 total touches in 1970, and I can guarantee that was not for performance reasons. He was really bad, but touched the ball a lot anyway. That's the life of a QB sometimes, especially one that's always behind like Jack Concannon.
This really is the battle with QB volume. Getting great per-play value over a larger sample is obviously better in a vacuum than in a smaller one, but QB volume tends to come when a team is losing, but this is not always true, because people like Fran Tarkenton (and Jim Everett) often got inordinate amounts of touches on teams that were good. It's a tough balance to strike.
The general assumption is not that extra throws are more difficult from an aDoT sense. The general assumption is that a lot of these extra throws come in situations where the opponent knows you're going to throw, which (in general) makes things more difficult. I say the word general so many times because this changes on a player to player basis, and there is no PBP data to go back and verify in the eras that we're speaking of.
As far as Kurt Warner, lord knows I'm a high peak guy. I think Kurt Warner is a Hall of Fame guy, and Jim Everett is a Hall of Fame guy, and Roger Staubach is a Hall of Fame guy. High peaks for short time spans get more respect from me than the more consistent accumulator types (i.e. John Elway, Tom Brady). As such, I can't argue with Kurt Warner being in the Hall of Fame. It would be disingenuous of me to do that.
The tough thing with Kurt is that it's hard to tell if the St Louis talent lifted him up that high, or if he just never fully recovered from the broken throwing hand that forced him out of town in the first place. Trent Green left St Louis and got better, so it's not like the Rams had a uniformly positive effect on everybody they touched. You could also clearly see the dip when Kurt left and Marc Bulger came in, and I really like Marc Bulger. To me, this means Kurt deserved his two MVP awards. They never got anywhere near the championship again without him, with the same great receivers and some pretty darn good QB play, but I think there was something special there.
Passing accuracy doesn't look flashy, but that's what Kurt had. That's what he did better than anybody else. It's similar to how Peyton Manning never looked especially flashy either. He just dropped back and threw the ball accurately all the time. No heroics required.
As far as Jim Everett being worth it to the Rams, it's a bit unfair to him to even ask the question in this manner. I take the position that it's not his fault the team traded so much for him. If it were just the one first round pick (instead of the two firsts plus a fifth that the Rams traded the Oilers for him), I think he would've got off a lot lighter in this regard, sort of like how nobody talks about Andrew Luck being a failure of a draft pick. If the Colts traded two firsts plus a fifth for Andrew Luck, instead of him just being the one draft pick, I don't think he would be let off the hook so easily either, even if I change absolutely nothing about how he played, and Andrew Luck was never anywhere near the player Jim Everett was. I would take just 1988 and 1989 over Andrew Luck's whole career.
So was he worth what the Rams paid for him? In 1988 and 1989 he was, but afterwards? Probably not, but I think if Jim were just a straight up third overall pick, he would get a lot less flak. He would have the Andrew Luck reputation of being far from a success as a top three pick QB (only a championship can make you a success, in the eyes of most, and Jim got closer to one than Andrew Luck ever did), but not necessarily a failure either.
Fans tend to hold things like this against players, but Jim Everett didn't make the trade. He doesn't deserve to get blamed for it. Bryce Young is in a similar spot right now. He's coming along nicely in my opinion, but Carolina fans are frustrated anyways, because he did not cost one pick. He cost several. This is the same spot Jim was put into.
I'm getting there as far as spreading my wings to cover other offensive positions. I'm looking deeper and deeper every day. I will never study defence this intensively though, because individual contributions are so difficult to separate from those of the team as a whole, which makes for both less accurate statistical understanding and a less interesting story. I prefer to study the defence of teams as a whole.
Loved the article Robbie.
I just want to know, though. How long does it take you to write this? Days?
Any article of mine typically takes multiple afternoons of work.
My writing pace is about one hour to 800 words, which is quite fast from what I understand, but on something like this that's 11K words long, it still eats up a lot of time. The only thing that takes slightly longer than that are the lists, which take longer because I have to learn the basics of 20 different stories instead of just one.
It's really bad for business, because if I took one of those afternoons every week and dedicated it to promotion instead of the finer details of whatever I'm writing, my subscriber count could be triple what it is now, but I'm not here for business. I'm here to write quality stuff. My interest is not really whether anybody reads it or not.
So yes. That's the answer. About an hour to 800 words, which for how long most of my things are means multiple afternoons worth of writing every week.
I think going through the questions is a good way to help determine who should be in the Hall and is something I've seen only one other time. While I agree with your conclusions, I don't think the voters will. I think Ben being in the good good stats, lots of winning category will get him in and I think the mythos created around Eli is too strong to overcome for Rivers. I think if Rivers would have won an MVP and won in 06 and 07 against the Patriots, he would be in.
I would also be interested in a further dive into Jim Everett, because even based on NY/A and ANY/A, he has only 2 seasons in the top 5 and 3 other seasons in the top 10. In your formula, what is giving him this boost compared to other numbers?
I truly think Philip (and not Eli, key point) would've been the one to stop the 2007 Patriots, if he'd been given a real chance at it. He'd just played perhaps the best playoff game of all time the week before, but then tears his ACL in the fourth quarter. You can't play the QB position moving one leg at a time. The Chargers should've just started Billy Volek, and I'm not entirely sure he wouldn't have beaten the Patriots. He certainly was a better QB than the -0.05 EPA/Play, -5 CPOE Eli Manning in 2007.
What I'm trying to get at is that the Chargers lost, and Philip Rivers was the reason they lost, in both 2006 and 2007, but there are extenuating circumstances (worst luck of all time one season, trying to play on a torn ACL the next) that ought to be taken into account there. This is the Philip Rivers what-if. We've talked about the Chad Pennington what-if before, and this is the Philip Rivers version.
In retrospect, I can quite confidently say the 2007 Patriots were dead. It took everything they had to beat the Jaguars in the second round game harder than anybody remembers it being, and after working so hard to finish an undefeated regular season they were just out of gas, particularly on offence. They couldn't move the ball at all against either the Chargers or the Giants, and it was the luck of fools and Patriots that they ran into a hobbled (to put it lightly) Philip Rivers and one of the worst QBs in the league in Eli Manning, allowing them to nearly pull it off anyway. Like I said in the article, I believe any of the top 11 QBs in the NFL (from Donovan McNabb upwards) would've beaten the Patriots in those 2007 playoffs, some by multiple scores.
What if Philip Rivers makes it through the Colt game unscathed? That's the what-if. It's purely supposition on my part, but the Patriots were limping, looking for somebody to lose to. The real Philip Rivers would've proved more than good enough for that, and with that win on his (and the Chargers') resume, where do they go from there? It almost surely gets them out of this playoff loser mentality, that they still remain in today, because they almost surely defeat the New York Giants in the SB with that defence they had in 2007. I'm not sure if the Giants would've scored a point that day. They only put up 17 on a pretty mid 2007 Patriot defence.
With the Chargers as 2007 Super Bowl champions (a reality that I think was closer than anybody thinks), the roster still takes a pretty big nosedive from there, but had enough changed in the interim to make things different in SD? The team would almost certainly still be in SD, and Philip Rivers would be the cause of that. He would certainly be in the HoF.
All of that gone, as a result of just one knee injury. Onto Jim Everett.
As with all my arguments, Jim's relies on peak performance. Let's go season by season.
Jim's prime begins in 1988, where despite a rush offence not worth using that got used A LOT, putting him in a lot of bad positions, Jim posted 1293 football outsiders DYAR, which puts him in the mix with Dan Marino and Boomer Esiason for the most in the league, all almost 500 ahead of anybody else. His rate stats are fantastic also, finishing fourth in DVOA and fifth in ANY/A despite an astonishing (for 1988) 609 touches. Mix this in with Boomer touching the ball only 491 times in 1988, you find that Jim is second on my QB tier list that year.
Onto 1989, where the rush offence is even worse, and gets even less touches, leaving Jim to pick up even more of the offensive workload. He parlays this into being the only QB in the NFL even close enough to see Joe Montana's dust in DYAR, and being much closer to the pack but still second place in DVOA and ANY/A. Combining this with the fact that he played all 16, compared to Joe Montana missing three starts in 1989, leads me to put Jim number one that season. He did not even make the Pro Bowl this season by the way, but he was without doubt the best QB who started all 16.
1990. Jim's only good teams that he will ever play on are already behind him, but he's still fantastic. The rush offence is horrendous. The defence is horrendous. His receivers are not what they used to be, but Jim is still chucking. Ninth in ANY/A and ninth in DVOA are not great on a rate basis, but on a volume basis on a team that DESPERATELY needed the volume, Jim was top five in DYAR again, and became the first player in NFL history not named Dan Marino to put up 600 touches three seasons in a row, a level of durability that's very difficult to maintain in this era. Combine this with a lot of the names above him on the rate lists being guys like Jay Schroeder (430 touches) and Jim Kelly (418 touches), and Jim finds the top five for the third season in a row.
Perhaps all this work got to him, because in 1991 he's an uninspiring 11th, even despite being about 16th on a per-play basis, due to the same 'all the guys better than him in the rate stats have 400 plays to his 600, so what would they have done on the horrendous Rams?' dichotomy.
He's trying Marc. He's really trying. By the time we get to 1992, Henry Ellard and Flipper Anderson are not what they used to be, and as the Rams are truly beginning to fall into the toilet as a franchise, there are no other offensive players you've ever heard of. Nevertheless, 6th in DYAR, 6th in DVOA on that typical Jim Everett volume land him 6th on my 1992 tier list.
He finally misses time in 1993, and is horrendous when he does play, so finally the Rams (who had not drafted a QB throughout this whole process of four losing seasons in a row, an important acknowledgement both contemporaneously and in retrospect that QB was not the problem in LA) decide to make a move, trading Jim Everett for a seventh round draft pick (LOL). Immediately after this, that franchise would fall into the pits of despair, moving to St Louis for their trouble.
Unfortunately, Jim got to New Orleans just as they were falling into the pits of despair. Challenge Marc. Name me one New Orleans Saint (other than Jim Everett) from 1994 to 1996. The rush offence was not worth using. The defence was horrendous, and Jim carried all of this to a 7-9 record with his 1101 DYAR (3rd), and 0.193 DVOA (5th) on that typical Jim Everett volume, landing him in the top five of my tier list again.
1995 is much the same story, with the same Saints roster, the same 7-9 record, but this time with 1168 DYAR (5th) and 0.190 DVOA (8th). With improvement leaguewide, roughly the same numbers only got him 7th place this time. He tried it again in 1996, but it was over, and you didn't hear much about Jim Everett after that.
That's eight prime years, and his tier list finishes in those eight years are 2, 1, 5, 11, 6, 29 (in the injury-shortened 1993), 4, 7, almost all of it on awful rosters, perpetually putting him in bad positions. Four top five finishes plus two additional top ten finishes (that was almost three) in an eight year prime has very rarely been done. He's not Peyton Manning (top five every year from 2003-2010), but look at the prime of Hall of Famer Jim Kelly, who played around the same time as Jim: 11, 9, 11, 9, 7, 3, 5, 9.
That's not Jim Everett calibre.
ANY/A hates Jim Everett, because he turned the ball over a lot, and that stat really overrates the importance of preventing turnovers, but stats like DYAR love his volume, and stats like DVOA back up that it wasn't all volume. He was top five on a rate basis as well several times, which gets exponentially harder when your awful rosters consistently ask you to crank out so much volume. Ask Josh Allen in 2024 the statistical benefit of decreasing the volume a little bit.
Adding all of this together, we find Jim Everett's position on my all-time QB ranking, which we all know is peak-performance based. If I had to have a QB for a four season stretch, I would take Jim Everett with my 15th pick. Behind Tom Brady (12th), Kurt Warner (13th), and Philip Rivers (14th), but ahead of Brett Favre (19th), Warren Moon (29th), Boomer Esiason (31st), and John Elway (33rd).
People tend to laugh when I bring this up, and point out that Jim's career record is 64-89, as if I don't know that. He only won two career playoff games. He only played in four of them, but for a man who spent a whole career having to deal with such trash rosters, he turned a lot of it into chicken salad. This is not basketball. Jim was not putting up great stats on bad teams just because he could like he was Zach LaVine or something. Football doesn't work like that. He was put into horrendous positions over and over and over again, and he showed his mettle.
I often wonder what could've been if Jim Kelly came into the league on the 1986 Rams, and Jim Everett came in on the 1986 Bills. Would it still be the case that Kelly is in the Hall of Fame, while Everett isn't? In my opinion, absolutely not, but we can only guess about questions like that.
So there. That's a surface level look into the career of Jim Everett. NY/A doesn't like him because it discriminates against bad rosters without a lot of RAC. ANY/A doesn't like him because he turned the ball over a lot, but for somebody like me who takes pains to discriminate against neither bad rosters nor people who turn the ball over a lot, I think Jim Everett is the 15th best QB of all time.
Thanks for that explanation on Jim. In addition to DVOA and DYAR liking him, his teams passing first downs are pretty high which I think helps explains the gap between his yards based numbers and his DVOA based numbers. Also, what led you to choosing 4-years instead of 3 or 5?
It's not actually a strict four. It's a minimum of four seasons that MUST be included. Beyond that, up to three seasons are allowed to be excluded entirely, for falloff or rookie reasons. However, three seasons do not have to be excluded if keeping them will improve the score. This generally leaves room to ignore the falloff of a player's career, or a few bad seasons to start it, but not both of these things.
It's basically a peak performance metric, with very little sympathy for extended falloffs or taking a long time to get going. The perfect player according to my formula is somebody like Peyton Manning or Dan Marino, who came into the league a finished product, and fell off quickly. This makes sense though in my opinion, because that's very few bad plays in a career.
The same does not apply to somebody like Tom Brady, because even once you exclude 2001, 2002, and 2019, you're still not out of sub-tenth place seasons. Therefore, he fares worse than a lot of people think he would, mostly because he took so long to get going. On a career-based ranking like mine, this hurts, and it should hurt in my opinion.
Because of this mechanic of punishing long lead-in times, and punishing long falloffs, it's not a true prime performance metric. It's a career performance metric, but peak performance is basically what it captures. It basically just captures peak performance, with punishment built in for having a lot of off years, be they in the front or on the end or in the middle, my formula doesn't care. This is in contrast to the public eye, which generally tends to just ignore off-years in the middle as if they didn't happen at all.
The impetus for the cutoff being four is actually Kurt Warner, who made the Hall of Fame with just four truly great seasons. Since you can be a Hall of Famer now with just four great years, I built a formula that doesn't discriminate between four great seasons and twelve of them.
With my excluding seasons method, Kurt's ranking on this list doesn't have to deal with his 2002, 2004, or 2005, instead using only his 1999, 2000, 2001, 2007, 2008, and 2009 seasons to calculate his performance level, and come to his ranking of tenth place, just above Tom Brady. At their peaks, they were about the same, but Kurt came into the league finished already, compared to Tom taking a few years to get there. That's the only difference.
If we look at somebody like Jim Everett, the formula sees him as his 1988, 1989, 1990, 1991, 1992, 1994, and 1995 seasons, entirely ignoring the lead-in year in 1987, the injury shortened 1993, and the falloff in 1996. If there were more than three bad qualified seasons, I would've had to punish him for them, but there weren't.
You may say this punishes newer players, who often are forced onto the field sooner, and therefore have a lot more bad plays than some of the older guys, and I say that's a feature. Not a bug. The newer guys who play bad as rookies and second year players just have more bad plays in their careers. I don't feel bad for my performance formula punishing them for that, even if it's not their fault since they shouldn't be playing in the first place.
There are also some other things about the aggregation method that put extreme value on the top-end performance while perhaps nullifying some of the bad, that I don't think I'm able to explain in a Substack comment box. One part I can explain is that there's quite a hefty buff a career gets for ever having been the best QB in the NFL in any season, which one can argue with, because it's relative. For instance, Carson Palmer gets it but Drew Brees does not, because Drew's prime basically coincides with the best parts of Brady and Manning's careers, and Carson wrangled his way into the top spot in a rather weak QB season in 2015. The strength of this number one position basically carries Carson on its own to 53rd all-time, a position he would've been nowhere near without the first place buff.
Anyways. I've gotten way into the weeds. Basically, it's an augmented peak performance ranking that punishes careers with a lot of bad plays in them (i.e. Tom Brady). I chose four seasons because that's how many top ten seasons Kurt Warner had, and he is currently in the Hall of Fame. Two of those four seasons are first places, which drives my formula crazy and pushes him all the way up to tenth place, but you know what? First place twice is the same number that Tom Brady had, so I think it's right to reward it that much. My method is designed to answer questions like this, and it's also designed to answer questions like Daunte Culpepper, who had a fantastic season in 2000, then two bad ones in 2001 and 2002, then two more great ones in 2003 and 2004, and a whole bunch more bad ones after that. Most formulas (like Kevin Cole's, which was my main problem with his ranking methodology) have trouble with stop-start careers like this, but mine does not.