Plain English with Derek Thompson - Why Is It So Hard to Predict the Next Great Quarterback?

Episode Date: April 29, 2022

This week we have the NFL draft, which is an annual exercise in failure. Every year, some NFL team makes a disastrous quarterback decision but also overlooks a potential star. Why is it so damn hard to predict QB play in football? Are scouts stupid, or is the future just unknowable, or is hiring fundamentally chaotic, or is there something specific about quarterbacking that makes it uniquely difficult to forecast? The economist David Berri joins to share his research on why scouts are terrible at evaluating quarterbacks. His ideas shed light on larger questions like "What is talent, exactly?" and "Does anybody know what they're doing when they're hiring somebody for a new role?" Host: Derek Thompson Guest: David Berri Producer: Devon Manze Learn more about your ad choices. Visit podcastchoices.com/adchoices

Hey, it's Sean Fennessey. We've got something special cooking on the Prestige TV podcast. I'll be recapping one of my favorite shows, HBO's Barry, every Sunday night with the writer-director star of the show, The Great Bill Hater. We'll talk about the show's wild twists and turns, its special brand of dark comedy, and how it all came together. So on Sunday nights, immediately after a new episode airs, you can hear Bill and I break it all down on the Prestige TV pod. Subscribe on Spotify or wherever you get your podcasts. Today's subject is football. Football and the mystery of talent.

The NFL draft is this week. That's where executives get together to select the college players that will join their team. I'm a huge fan of the NFL draft for a couple reasons. First, I love football. Second, I think talent and potential

are such fascinating topics. I think we all do. We all read articles about, you know, the next great pop star, the next great movie star, Next Great Product, Company, Technology. Like, next great thing is just a really fun concept. But there's something about the topic of potential in the NFL specifically that doesn't

just fascinate me. It kind of haunts me. And that is the question, why is it so damn hard to predict the next great quarterback? And to be clear, it is incredibly hard to predict the next great quarterback. Like, here's a slice of recent NFL draft history. In last year's draft, 2021, the top four quarterbacks drafted, in their first year in the NFL, threw more interceptions than touchdowns. That's really bad.

Only two rookie QBs threw for more than 12 touchdowns. They were Mack Jones, fifth QB selected, and Davis Mills, eighth quarterback selected. In 2018, the number one pick, Baker Mayfield, has now essentially been ghosted by his team. The next quarterback drafted was Sam Darnold. He was bad. Then he got traded.

Then he was still bad. The fourth drafted quarterback was Josh Rosen. He was even worse. Just a disaster. The only quarterback from this draft to win MVP was picked fifth. Lamar Jackson. The number three picked Josh Allen also total stud. 2017 draft produced Pat Mahomes, maybe the best quarterback in football.

But he wasn't picked first in his own draft. So, you know, an innocent person is going to think, well, the guy picked ahead of the best quarterback in football must be pretty darn good. Nope. Mitchell Trubisky, who is bad. Do you see a pattern here? The NFL is a multi-billion dollar business that pays scouts and other brainiacs millions and millions of dollars to predict

who can succeed as quarterback, and just about every year, they find some way to screw up massively. And today's guest has a theory about why. His name is David Berry. He is an economist at Southern Utah University who researches sports. and player efficiency.

So why is the NFL so bad at predicting the next great quarterback? And what does that tell us about the challenge of predicting talent anywhere? I'm Derek Thompson. This is plain English. David Barry, welcome to the podcast. Thanks for having me. So you and I are speaking near hours before the NFL draft. We cannot therefore comment directly on who's going to draft to,

but we can say with, I think, a high degree of certainty that many mistakes we made, before we get into your research about why drafting NFL quarterbacks is so damn hard, I would like you to tell me why is an economist answering questions like,

why is drafting an NFL quarterback so damn hard? I wrote it with Rob Simmons, who's a co-author of mine in England, and I don't remember what led us to do this. It is part of a broader conversation I've been having in my research about how people in sports make decisions. And that's interesting to economists

because one of the core assumptions in economics for very, very long time is that human beings are perfectly rational and know what they're doing. And I have never bought that story. I've always found that to be ridiculous because I've actually met human beings and therefore I didn't believe

that they were perfectly rational. And so I have spent a long time investigating decision making in sports. And at times, they do make decisions in a way that you would think as an economist fits the story that they're rational. And at other times, they do not. And the NFL draft is a great example of, no, they do not. So let's jump right into it. What is the relationship between a quarterback's draft position and his subsequent performance in the NFL?

Are the top quarterbacks drafted the best NFL quarterbacks? Is the answer yes, no, or hell no? If you are trying to predict who's going to play, then the quarterback's taken first will play more than the quarterback's taken later. That is definitely true. If you're trying to predict who is going to play better on a per play basis, not a chance in hell. They do not know that. There is no statistical relationship between per play performance and where they're actually drafted. So we do know that if you're taken earlier, you're going to get a lot of chances.

And if you're taken later, you are not going to get a lot of chances. And but of those who end up getting chances, their performance and their draft position are not actually related. So draft position predicts opportunity, but it does not predict performance. Let's go right through this paper. For those following it home, the name of the paper is catching a draft on the process of selecting quarterbacks in the National Football League Amateur Draft. So in a nutshell, David, your thesis is that where people draft quarterbacks, basically is unrelated, unrelated to future NFL performance. Take me through everything that you looked at in order to come to that conclusion.

Yeah, we looked at a wide variety of performance measures, some that I've created. So I've gone through and looked at the data and connected it to wins and looked at how many wins a quarterback produces. Some of it is the measures that the NFL has created, such as the quarterback rating, some is completion. And then the individual stats, completion percentage, touchdowns for attempt, interceptions per attempt, fumbles per rushing attempt. And there just simply is nothing that's very predictable.

You can't predict any of this stuff knowing where they're drafted. Once you see them play, there's no difference between them. And so there are quarterbacks who do play better than others. It's not related to where they were drafted, though. So this is a quote from your paper. On a per play basis, quarterbacks chosen with picks 11 through 15, as well as picks 51 through 90 outperform quarterbacks chosen in the top 10. Top 10 quarterbacks really don't offer more.

They just get to play more. One explanation of this, one explanation of the fact that the top drafted quarterbacks don't end up being the most efficient on a per play basis is that this is a reverse order draft. That means that the quarterbacks chosen in the top 10 tend to be going to really sucky teams. So, like, maybe the quality of the teams that are hiring the top 10 picks is lowering their numbers. You looked into that potential explanation. Is that it? Is it because of the reverse order draft that we're seeing this phenomenon? Part of that is the story, right? It is the fact that

quarterback play is entirely depends on the context of what you put around them. So when we look at a top quarterback, we often, this is the mythology the media creates. A top quarterback makes their team so amazingly great that it doesn't make a difference what team they play on. They're just winners. They're all winners. That's the story. We got to draft a winner. And that's ridiculous. That's not how it works. You are, the quarterback's play depends on who you put around them. You give them better receivers, a better offensive line, a better rushing attack, a better offensive coordinator. Remember, the quarterback's not calling the plays. The offensive coordinator calls the place. What I think so interesting is that the cult of the quarterback sort of sees it

and sees the role of this position in two very different ways. In the one hand, you put your finger on this, the quarterback is, I think, rightly considered the most important part of a football team. But at the same time, the quarterback is in many ways the most contextual position in professional sports. Like, the best basketball players can run their own plays. They can dribble the ball right up the court and drive right to the hoop. In baseball, baseball is just basically a series of duels, right? If you're at whatever, Aaron Judge, like, you walk up to the plate and you face the pitcher, It doesn't meant, like, you know, having team chemistry, who cares? You're up against one pitcher.

It's just a duel. The pitcher pitches the pitch and you swing. But quarterbacks are both unbelievably significant to the outcome of the game and absurdly contextual. The running plays, they didn't necessarily choose. Behind a line they didn't draft, throwing two-odd receivers who are not them. They simply cannot do what a lot of stars in other sports do. I want to go back to your paper because there's a couple subtle points that you make. I think are really interesting.

You looked at all the stuff they measure at the combine. Body mass index, and for those who only vaguely know what we're talking about, sort of vaguely the two ways that you can draft a quarterback or scout a quarterback, excuse me, you can look at their performance in college, and then you can look at them at the combine. At the combine, you get all these stats, body mass index. There's a wonder lick test, which is a very dubious test of intelligence. You timed their 40-yard dash.

And your results found that basically none of this is related to either a quarterback's college performance or their professional performance. I mean, am I oversimplifying by saying that, according to you, the Combine is basically one enormous red herring that throws this like mountain of statistics at scouts, almost none of which is actually useful for the quarterback position? Exactly. That's true.

I think this is a fallacy that people often make in decision making is they think, I want as much information as possible. but the question has to be is the information you're gathering related to the decision that you're making. And so they collect a lot of information on quarterbacks. And they give them the Wonderlich test a great example. Has no football questions on it. So it's not relevant to the job. There's a bunch of questions not related to anything you're going to be doing here.

How well do you do on that? It's like, why don't you go have them take an organic chemistry class? Wouldn't that be just as good? That has nothing to do with anything either. So it's just, it's a silly thing they do. And then they do things like, we're going to measure your height. And they sit there and they say, well, this guy's 6-4 and this guy's 6-2. And they think that's different.

I'm sorry, that's the same. That doesn't make any difference. Height predicts where they're drafted, but it has nothing to do with how they subsequently perform. One of the best ones is hand size. They look at it. They obsess on hand size. And they sit there and they measure their hands.

And they did this with Pickett in this draft. His hand was too small. And they're like, I wonder. if he'll be able to hold the ball in a place like Pittsburgh, where he played college football for four years, holding the ball. Yes, he can hold a football. He's already established he can do that. Measuring his hand is pointless.

He's already done that. So that would only be relevant if he never did it before. Then you would measure his hand. Of all the things that you looked at, and it's really, the paper is really phenomenal because you go through all the various college stats that could be predictive of professional performance. You obviously go through the combined stats.

I should be clear that there is one statistic, one statistic that your analysis found had a statistically significant and positive impact on performance at the professional level. And that statistic was college completion percentage, the share of passes thrown in college that are actually completed. Just tell me a little bit about how we should interpret that part of the paper, that of all these things you can look at, there is a little bit of a positive relationship here in completion percentage. Okay. So this is the point I make to students, and this is what statistical significance means that what you found is not zero. So that does it mean,

and we say this over and over again in class, that doesn't mean it's important. Not zero is not the same thing as important. So saying, hey, I found something that's not zero. Great. Does it actually help me make better decisions? Yeah, probably not because it's not a very big effect. And there's all sorts of examples of quarterbacks. You know, Josh Allen is a great example of a quarterback who couldn't throw in college and they taught them how to throw in the NFL. And there are other quarterbacks who thought through really well in college but can't throw it in the end of because the problem you have is that these quarterbacks are playing college football with non-NFL players and non-NFL coordinators against non-NFL defenses. And then you're trying to

project in your mind what that's going to mean when they play. with NFL players against NFL defenses and NFL coordinators. And I'm sorry, those things are not really related to each other. So we don't know what that means. If you're a quarterback at a top school like Alabama or Clemson, and you're throwing to receivers who are immensely better than the defensive backs that are covering them, and they're open by 10, 15 yards down the field, and you manage to complete a pass, you're not playing quarterback.

You're playing catch. And that doesn't prove anything. because in the NFL, they're all going to get covered. They're all going to be covered by other NFL players. And you're going to discover when you get in the NFL, oh, that guy's not open by 10 yards anymore. Now the guy's right next to him, what should I do? And the court is going to throw the ball anyways,

but it might get picked off. It might not. We don't know. You got to throw it because the defensive back's not going anywhere. He's going to stand right there by him. They're not going to be that open anymore. You just got to throw it, and you hope for the best. I can imagine that some people listen to your argument,

are worried that you're representing a point of view that they might think of as scouting nihilism. That is the idea that absolutely nothing predicts anything when it comes to quarterbacks. So if you want to draft a quarterback, you might as well just write their names and little ping pong balls and put it in one of those little things with the wheel

and spin it and spin it until it spits out a number like you're playing bingo or something. You might as well utterly randomize the process. And I cannot possibly believe that the process can or should be randomized. I watch football. I love football. Pat Mahomes is better than Nathan Peterman. He just is.

Like, Joe Burrow is better than some undrafted quarterback who just sucked in college. And scouts saw that. So the way that I interpret your research is something like this. And you tell me how you disagree with this interpretation. Collectively, NFL scouts are pretty crummy when it comes to determining within the set of pretty good quarterbacks, which are the absolute best, and which are the second or third best. But they're not worthless at creating that set. They're not worthless at identifying. These are roughly the five to ten best quarterbacks in the draft. But then once you've narrowed it down,

randomness and context and fit. And the fact that the most highly evaluated talent is typically going to the worst teams, these environmental factors essentially take over and that creates the relatively chaotic outcomes that you found in your paper. Tell me how my interpretation either fits or doesn't fit

with the facts as you've put them together. All right. So you mentioned Mahomes and you mentioned Burrell, but you could also come back and say, Well, yeah, Kurt Warner is also better than a whole lot of other quarterbacks, and he was working at a grocery store when they found him. Tom Brady was a six-round draft pick, and we all think he's the greatest quarterback ever. And, again, six-round draft pick.

They had no idea. Nobody knew Tom Brady was going to be good. Nobody. The Patriots didn't know either. The Patriots have said that. If we thought he was good, we would have picked him in the first round. We picked him in the sixth round because we didn't think he was any good. And so if you're saying that the suggestion is randomly pick the quarterbacks,

and that's what the research saying? We don't know. We're not saying that because we don't know that either. So there's no, what we, what we're saying here is we don't know how to rank the quarterbacks and maybe you ought to think a little bit harder in building your offense that there's more involved in this than who you're picking at quarterback. Maybe you ought to be thinking about who your offensive line is and who your wide receivers are and who you're running back and who's calling the plays.

And what you maybe want to do is just find quarterbacks who can do the things. who can do the things that you need to have done. Make the throws that you need to make, make the decisions that you need to make. And that should be the criteria and not get this idea that if I get this magical quarterback, it's going to solve all my problems. The reason why we give this halo effect on these drafted quarterbacks, because you haven't seen them be bad yet. And so you're imagining in your head, well, they're just going to be amazing.

I spent time talking to them and they're just leaders. And they just, it's like, okay, they haven't done it yet. Wait till they fail a few times and then tell me how great you think they are. There's one other discovery that you made in your research. This is with Brian Burke from Advanced Football Analytics.com. I think bears very directly on this conversation. So you and Brian looked at the correlation between NFL quarterback play season to season. So this isn't college versus.

professional. This is professional season to professional season. And the results indicated that even a veteran quarterback will often have a lot of variability season to season. You found the correlation between a quarterback's interception per attempt, season to season, was basically zero for touchdown per attempt. It was only a little bit higher. I think this is so interesting and so important because, first of all, it cashes, it matches my experience watching sports. Like, you know, Russell Wilson is electric at the first half of two years ago and then suddenly kind of falls off a cliff. Cam Newton, extraordinary MVP winning season.

Next year, basically nothing. Derek Carr is amazing, and then he's not very good. And then he's amazing again. He's like, you know, top 10 again. And the draft offers this kind of an illusion of a focusing point. It's like, can we predict the future? today. But what your observation here suggests is that if on any given day we try to make predictions about future quarterback performance, it's really, really hard. Because once again, the paradox of

the quarterback is that the most important position in football is also among the most contextual in football. He is at the mercy of all these forces around him that he doesn't control. And that's why you have this enormous variation in performance year to year, even among all these veterans. Exactly. Exactly. Yeah. So Jared Goff in the first eight or ten games this last season was the worst quarterback in the NFL. And he had no receivers to throw two. And then suddenly they went and got one receiver that he used to throw two with the Rams, Josh Reynolds, won. And then they changed offensive coordinators. And in the last, you know, four or five games, he was among the top five quarterbacks in the league. And you're like, really, that's all it took? You just give him.

one guy to throw two? And suddenly he's like, oh, I know how to do this. And that's, that's quarterbacking in the NFL. It's, you can change things every, ever so little. And suddenly their performance, because it all depends on what plays are you calling? Who are they throwing to? Who's blocking for them? And so it's, the quarterback is not out there by themselves. We, we place so much importance on this one position. And it just isn't realistic to think that that quarterback is going to transform by themselves a team that is, you know, two and 15 into a team that's 15 and 2 because all these other things have to happen as well. And so you, so as a decision maker, you have to take a step back and say, look, I'm not going to be able to find magical

quarterback to solve my problem. I got to build a team and I got to think about all of the different units on the team and what they're doing and how do I maximize all of that rather than just, hey, quarterback, win the game for me all by yourself, because that's not realistic. And this is something that also people should think about when they think about quarterbacks winning games at the end. John Elway used to do this all the time, win games at the end. And a few of my friends, including my wife, who was a Broncos fan, would always point this out.

The only reason why John Elway has to win the game at the end because he couldn't play well the first three quarters. That's what's happening here. If he knew how to play the first three quarters, we wouldn't need him at the end. And it's like, I don't need you to do that. Why don't you play the first three quarters? I want to make sure that I get you in this point because I don't want to at least represent myself as thinking

that I don't believe talent exists, that I think that quarterback talent is something that is purely contextual. I think the two things are true. I think that quarterbacking talent absolutely exists. I think it's a thing. I think it's a thing with vast distribution. I think there's some people that are terrible

at playing quarterback, and there's some people that are just unbelievably gifted at throwing, at being aware in the pocket, at making decisions in three and a half seconds, at understanding context, at understanding the likely defense that he's going to face play after play. I think all those things exist.

I just also think that context in the quarterback position is so important that it might explain why results like yours find that there's so much variance season to season in quarterback performance and why there's so little effectiveness in predicting quarterback performance when even immensely talented people are making this transition from one team

in an amateur league to another team in a professional league. Is that an okay summary of where you live? That's exactly right. That's exactly what we're saying. Is that, yeah, it does depend. It could be the case that some players are way better at this than others. Again, we don't know that, but it could be true.

but it is the case that who you put around them is going to make a huge difference. And also I think we underestimate it matters who you're playing and it matters what that defense is doing. The defense is doing stuff. You know, Steve Mariucci said this years ago after a loss. He said, you know, the other team tries too.

I mean, I'm not doing this by myself. Again, you're not watching a play where we're scripting everything on our own. They're playing too against us and they're doing stuff and they're trying their hardest. and sometimes they just try better than we do. And that's life.

That's how it works. And so that's, and I think people think they dismiss that as, as, you know, when a quarterback throws an interception, sometimes it's because the defensive back simply made a play. It also could be the receiver ran the wrong route. It could be the line didn't block right.

I mean, there's a lot of things that go into that. And we blame all these outcomes on just the quarterback. And it's like, you know, I tell students this, how many times when you watch a quarterback's own interception, they run down in the field and they have a long, intense conversation, a receiver, which appears to be, you know, you were supposed to look at the defense and the route goes in, not out. I don't know why you don't know that.

And the receiver's like, oh, yeah, you're right. Yeah, safety was, okay, yeah, I got that. Yeah, yeah, thanks. You just threw me an intercept. No, but it's an important point that the receiver doesn't get the interception, right? No, they don't. Like statisticians could give the receiver the interception, but we always. always give it to the quarterback, which goes to this difficulty of predicting season to season

because you might get a wide receiver that is more prone to essentially allowing interceptions on hot reps, he understands. And this I've never tested, but I have this hypothesis and maybe someone could test us sometime. That intercept, that defensive backs have a lot of interceptions earlier in their career, and it's because they're clueless. Because they don't know where they're supposed to be. And so therefore, they end up in places the quarterback doesn't think they're going to be there. So the quarterback throws it, throws it right to.

the defensive back, defense the back looks like a genius. Look, I was right there. And the quarterback's like going, I've seen your defense. What are you doing? I actually don't know what I'm doing. I have no idea what I'm doing. I don't even know what defense were right. What it suggests, however, if you're right, I'm not entirely sure that I, that I buy that thesis, but to the extent the thesis might be correct, what Bill Belichick should do is inject like randomness into his defenses. Like every single play, one defender shouldn't know what they're supposed to do. So they like walk around randomly. So it's harder, therefore, for the sky. outs on the other team are essentially for the offensive coordinators of the other team

to predict the defense because there was always one agent of randomness that is never doing something consistently game of the game. That is essentially, I think, what Matt Patricia's defense was with the Patriots. Unfortunately, when he got to the Lions, his idea was that all positions could be interchangeable, therefore you couldn't predict what anybody was doing. But he always, he forgot the fact that some of those players actually have to be good. The term should be optimal randomness. That defenses should achieve optimal randomness.

It's great thing you could switch roles, but no good if you can't do any of them right. I want to conclude here on a big picture thought about what it is that we're talking about, because what we're talking about is how do you predict human achievement? How do you predict talent? And the model in my head from your research goes something like this. There's like three factors that determines success in a sport like football. There's hard ability, which is height, wingspan, vertical leap. These things are relatively easy to measure.

there's soft ability. It's like intelligence, decision making, I don't know, grit. Those things are much harder to measure, much harder to predict. And then there's context, which is basically impossible to predict because it doesn't exist yet. You can only guess you are drafting players into a context that does not exist. You're taking the flower that blooms in Australia and trying to plant it into California, you have no idea what's going to happen. So hard ability, soft ability, and context into me to be the variables that we're playing with.

And that might explain why in other sports, scouts are a little bit more accurate. So, for example, in the NBA, every MVP, other than Nicola Yokic, ironically, has been drafted among the first 15 players. So collectively, teams seem pretty good at identifying the best players draft after draft. They're not great at it. It's top 15. It's not perfect at all.

But clearly they're better than at football. And that makes sense because basketball is a little bit less contextual. an individual point guard or shooting guard or LeBron James, some plays every position, can truly take over a game late and just essentially play offense by themselves in a way that no quarterback can even remotely do. Do you buy this idea that essentially as the context element goes down,

as context becomes less important, it becomes easier for scouts to predict human performance because what they're predicting is more transferable across time. Yeah, so that's a really big issue we think about Cisco analysis, is that what we're trying to do in measuring performance is, are you measuring a skill or are you measuring luck or context? And so when it comes to football measures and going back to that paper, I wrote with Brian Burke,

we went through a whole bunch of different measures of performance. None of them were very consistent across time. And it tells you that really you can't really measure quarterback performance very well. It does depend a lot of context. The way to illustrate the context argument is think about this. Take a top player like LeBron James or James Hardin or whoever, take them off the team they're on right now, put them on another team tomorrow night. Do you expect their performance to be any different at all? And the answer is no.

It's going to be exactly the same, even though they're playing with totally different teammates with a different coach who apparently has some ideas on offense that nobody's listening to. So right, they all know that, right? I think you're exaggerating with exactly the same. But I take your point that it's going to be much more similar than taking the quarterback and moving a team to team. The consistency is really, really high. In baseball, we expect it to be exactly the same, right? Baseball, we expect also to be very similar, right? But you take a quarterback on Sunday off that team and next Sunday put them on another team

with a different coach, different receivers, you're going to see disaster. They're not even going to know what the plays are. They've had no time with these receivers. They have no idea what they're doing, and it doesn't make a difference who the quarterback is. You can't trade a quarterback in the middle of season and expect that performance is going to stay basically the same because their performance depends on who's around them and what plays they're calling and what we're doing here.

And you just can't teach that in a week. And so that is, that's the issue, is that that's the problem you have here in predicting performance. And this is one of the going back to, you know, do the undrafted quarterbacks. Are they any good? Well, we could never know because you never give them a chance to play. For all we know, if you stuck that undrafted quarterback

and you gave them a bunch of practices and you let them play, you know, two or three seasons, maybe they would be just as good. I don't know. You don't know either. And so we don't know how many Kurt Warner's are working in grocery stores right now. It seems unbelievable. That was the only one.

So Kerr Warner was the only grocery store employee who could turn out to be a Hall of Fame quarterback. It could be that there are others who could do exactly the same thing, but they were never given the chance. I have a co-author right now who used to be a, he was in an NFL camp years ago, And he was, he got taken in as a camp. And he was told the very first day, you are not making this team. We already have three quarterbacks. You're just here to help us run practice.

And no matter how well you run those practices, it will make no difference. You are not making the team. I already got my three quarterbacks. That's, that's what an undirected quarterback faces. You're not, they bring them in, but they're just practice fodder. They're not there to make the team. And they're not going to do that because the coach is like, I'm not going to my whole team on an undrafted guy.

Nobody picked this guy. I don't know that they're any good. Very last point that I want to make because this is draft night or if people are listening in the morning, draft night was last night. You do make the point in the paper that to the extent that there are NFL positions that can be profitably predicted, offensive linemen is actually one of them. Not that it's the easiest thing in the world to predict great offensive linemen. There are, of course, busts.

But when you think about what it is that offensive linemen do, yes, they work within a phalanx of players, but they're also dueling. They are up against one, typically, sometimes two, defensive linemen that are trying to get past them. And so the job is more transferable. The work is more transferable and less context-based than quarterback play. And that might explain why, in your findings, it is easier to predict great offensive lineman play than it is to predict great quarterback play. It is essentially about the context question. Exactly, exactly. And the same thing is probably true for wide receivers as well. I mean, wide receivers, a lot of their,

what they do, and defensive backs, a lot of what they do is based on physical ability that is predictable. It does make a difference if a defensive back runs a 4-740, that is not the same thing as a 4-440. That's a totally different thing. And so, and you can't predict that. If your quarterback runs a 4-740 versus a 4-4-40, that may not make any difference because I'm never going to ask

him to run 40 yards, so I don't need you to do that. Peyton Manning never ran more than two yards in his life. And thank God, Because that was the ugliest parts of his career that I had the displeasure of watching. I often tell my students the Tom Brady Combine footage, this is the way I described Tom Brady after his combine footage, least athletic male model in the history of male models. No lie perceived in that statement. David Barry, thank you very much for joining me.

I appreciate it, man. Good. Plain English with Derek Thompson is produced by Devin Manzi. If you like what you hear, please follow, rate, and review us. New episode drops on Tuesday. Have a great weekend.