The Justin Fields Conundrum
Finding Fields' potential, comparing him to Caleb Williams, and understanding why it's more complicated than that for Ryan Poles.
The Chicago Bears are once again on the clock.
Before the 2023 NFL draft, Bears general manager Ryan Poles traded the #1 overall pick to the Carolina Panthers. The Bears got a massive return, including Carolina’s 2023 pick at #9 overall, two second-round picks, star receiver DJ Moore, and Carolina’s 2024 first-rounder.
Poles traded away the first pick after deciding he was confident in the development of quarterback Justin Fields, whom the Bears had selected at #11 overall in the 2021 draft. Poles was now building the roster to give Fields every chance to succeed in 2023; the trade brought Fields a dominant receiver in Moore, and the Bears drafted offensive lineman Darnell Wright to provide Fields with additional protection.
Once the season began, however, expectations changed quickly. The Bears stumbled out to a 1-4 start, and it was looking like another lost season in Chicago. Things went from bad to worse when Fields dislocated his thumb in a week six loss to the Vikings. The one real bright spot for Chicago was that Carolina, who owed the Bears their 2024 first-round pick, had yet to win a game, setting that pick up to be an early selection on draft day.
After missing five games due to his thumb injury, Fields played some of his best football down the stretch and helped the Bears win four of their last six games. The Panthers, on the other hand, finished last in the league, handing the Bears the #1 overall pick.
So now, for the second consecutive season, Poles faces the same question, one with massive stakes for the franchise, Justin Fields, and Poles himself: What should the Bears do with the first pick?
Field(s) of Dreams?
Poles has two options*: Stick with Fields and trade #1 overall for additional reinforcements, or pick a quarterback at #1 and trade Fields.
Evaluating these options can be reduced to one question: How good is Justin Fields compared to the quarterback the Bears can draft at #1 overall - likely USC’s Caleb Williams?
Poles bought himself a valuable data point by sticking with Fields last season. We can now look at Fields’ first three seasons and see how players with similar output over that stretch produce long-term.
The metric we’ll use here is Adjusted EPA (Expected Points Added) per play. Adjusted EPA/play is a stat that measures every quarterback action in a game - passes, rushes, sacks, scrambles, or penalties - in its expected impact on the final score. Adj. EPA/play is a useful metric for quarterback evaluation because it’s relatively agnostic to style. Fields creates a lot of value with his legs, but we can use Adj. EPA/play to quantify that impact, combine it with his passing, and compare that to all other quarterbacks.
Fields posted Adj. EPA/play of -.104 (if you have plays that hurt your team, you get negative EPA) in his rookie season, followed by .04 and .013 over the next two seasons. To better understand Fields’ potential, we can find the Adj. EPA/play production curves for the five quarterbacks with the most similar** first three seasons to Fields. We’ll also include a projected fourth season for Fields based on a linear regression run across all quarterbacks who’ve played 4+ seasons since 2012, along with the 25th, 50th, and 75th percentile marks 2023 Adj. EPA/play.
The top comparison for Fields was Blake Bortles, Jacksonville’s #3 overall pick in the 2014 NFL draft. Other top comparables are Geno Smith, Jay Cutler, Ryan Tannehill, and Daniel Jones. It’s an archetype of serviceable, although not outstanding, quarterbacks. Tannehill, Smith, and Cutler all made the Pro Bowl once, and most fans wouldn’t regard them, Blake Bortles, or Daniel Jones as total failures. At the same time, they are quarterbacks that you would probably pass on if you knew how they would progress.
From Ryan Poles’ perspective, the biggest concern would be how little this group won, particularly in the postseason. Bortles was benched in his fifth season with the Jaguars, and only had one season where he won more than five games. These five quarterbacks only made the playoffs seven times in the nineteen combined seasons between them, and none of them reached the Super Bowl. Although most of this group saw increased Adj. EPA/play going into their fourth season, if Fields follows these career arcs, he’ll likely never become a quarterback that can turn the Bears into a true contender.
Another factor to consider is what Fields’ next contract would look like and how that would impact the rest of the roster. The Daniel Jones comparison is also relevant because it might give us an idea of what it would cost Chicago to stick with Fields, given that they’ve had similar production over their first few seasons. Jones signed a four-year, $160M contract before this season, making him the tenth highest-paid quarterback in the league, tied with Matthew Stafford and Dak Prescott. If the Bears decide keep Fields, their next action would be extending his contract, and they’ll likely need to pay him in the top half of quarterbacks.
Quantifying Caleb
But the analysis of Fields’ value is only relevant when compared to what the Bears could have at the #1 overall pick.
Projecting Caleb Williams’ NFL career is trickier than trying to project Fields’ future output because our only data points for Williams come from his time in the NCAA.
One way we can attempt to project Williams’ output is to work backwards. Williams is a lock to be drafted at the beginning of the first round, so to project his early career production, we can look at other first-round quarterbacks’ early-career Adj. EPA/play output. Those quarterbacks provide us a range of early-career outcomes: elite options like Patrick Mahomes, good contributors like Justin Herbert, and busts like Mitchell Trubisky. By placing first-round quarterbacks taken since 2012 into these tiers*** and plotting their average production over their first four seasons, we’ve created a range of projections for Williams.
Unless Williams ends up a poor early-career performer - comparable to round-one quarterbacks like Trubisky, Sam Darnold, and Marcus Mariota - he has a decent chance of posting better numbers than Fields in year one. Based on his body of work at USC, where Williams won the Heisman and has been among the best quarterbacks in college football for two seasons, Williams looks much more like a prospect in the elite or good tiers. That would put Williams on a career arc similar to Josh Allen, Patrick Mahomes, or Justin Herbert.
Williams will also be significantly cheaper than Fields for at least the next four seasons. Because of the NFL’s rookie pay scale, which caps what new players make when they enter the league, Williams is likely to get paid $40M total for a four-year deal, roughly a quarter of what Fields would be worth if he got a contract similar to Daniel Jones. That money the Bears save at quarterback could help them shore up positions across the roster and give Williams a better shot to succeed.
The upshot of all this analysis is that Williams is a much likelier savior for the Bears than Justin Fields. We don’t know how either of them will play next year, but based on how players like them have progressed in the past, players like Williams - highly-touted first-rounders - tend to have better career outcomes than players like Fields, and also a better chance at delivering a Super Bowl for the Bears.
So we know what the Bears should do with the first overall pick, but what they will do might be an even more interesting question.
Pressure on Poles
Ask the Bears' owners or fans what their goal is, and you’ll get the same answer: To win a Super Bowl as soon as possible.
Poles’ incentives as general manager, however, are much more complicated.
A poor 2024 season will almost certainly cost Poles his job. In a piece by Kevin Cole at Unexpected Points covering the misaligned incentives of NFL general managers, Cole notes that historically, NFL management regimes typically have three seasons to produce a winning record and a playoff appearance before ownership fires them. Poles is entering year three and has neither a winning season nor a playoff appearance. Whether he drafts a quarterback or sticks with Fields, this will likely be his last chance to show enough progress in Chicago to keep himself employed.
This mismatch in incentives has a name: the principal-agent problem. It describes a situation in which the best choice for an organization doesn’t align with the best choice for the entity making the decisions for that organization. It’s very common amongst NFL teams, where coaches and GMs have a short time frame to produce championship-level teams, which often take longer to cultivate than the time they are afforded to build them.
This is where the Fields versus Williams debate gets tricky. In the same piece, Cole notes that the principal-agent problem creates an aversion to drafting quarterbacks when their jobs are on the line. The reason is that, in general, veteran quarterbacks tend to produce better single-season outcomes than rookie quarterbacks do in year one.
Now, we’ve already gone through an analysis that suggests that there is a good chance that Caleb Williams will do as well or better than Justin Fields on day one. Most Tier-1 and Tier-2 quarterbacks had better rookie seasons than Fields’ fourth-year projection. Still, if you’re Poles, you could certainly make a case that Fields is a safer bet for the 2024 season, especially if he believes Fields' solid finish to 2023 was more than a flash in the pan. Imagine if Poles had decided to move on from Fields last season and taken Bryce Young in the first round. Young had the second worst Adj. EPA/play in the league in year one with the Panthers. A similar rookie season for Williams in Chicago could be lights out for Poles.
There are certainly a range of outcomes that would be good for the Bears’ long-term Super Bowl chances but bad for Poles’ short-term chances of getting through the season. For example, Josh Allen turned out to be an elite quarterback, but posted poor Adj. EPA/play numbers in his first year that, in turn, resulted in five wins for Allen’s Bills. If Williams is the next Josh Allen, and he posts a .015 EPA/play like Allen did as a rookie, there is a good chance Poles won’t survive it. Poles might need Williams to be a good long-term solution and a reliable producer in year one who leads the Bears to the playoffs.
Because of clashing time horizons putting the Bears’ long-term Super Bowl aspirations at odds with Poles’ short-term employment aspirations, what the Bears should do isn’t necessarily what the Bears will do. The real question in Chicago before draft day might be which quarterback Ryan Poles thinks will get him through 2024.
Edited Excellently by Greta Gruber
Pre-reads by Aidan King, Trey Zamora, Joey Schreder, Tommy Jones, and Richard Pente
Footnotes:
*There are other options, but they have some major drawbacks. Picking anything other than a quarterback at #1 wouldn’t make any sense for the Bears, as they could send the pick to a quarterback-needy team for more draft capital like they did last season. Drafting a quarterback at #1 and keeping Fields, putting them into open competition for the starting job, would make for a must-watch offseason, but it is also unlikely. Trading Fields would return valuable draft assets - some NFL GMs project a second-round pick or more - that would help the new starter in Chicago.
**To find the quarterbacks most similar to Fields, I went through a database of every qualifying (>304 dropbacks) season since 2012, found quarterbacks who played 4+ seasons, and then used a weighted sum of the difference between each quarterback’s output in each season with that of Fields.
***The tiers were made by looking at the third and fourth seasons of quarterbacks taken in the first-round quarterbacks. That sample introduces selection bias because really bad quarterbacks are not starting for four seasons (Johnny Manziel, Zach Wilson, etc). The tiers are also slightly different than you would consider these QBs now, as I based them on weighted EPA/play in years 3(.40) and 4(.60) (Guys like Baker Mayfield and Kyler Murray are worse here, and Jameis Winston was surprisingly good). I manually adjusted Winston and Teddy Bridgewater down to tier 2, Lamar Jackson up to tier 1, and Justin Herbert up to tier 2 (get Harbaugh to L.A ASAP) to better reflect how we see these quarterbacks now. I also dropped Joe Burrow, who barely qualified this season and struggled with injury in his starts.
Elite: Patrick Mahomes, Josh Allen, Lamar Jackson, Tua Tagovailoa
Good: Jameis Winston, Justin Herbert, Carson Wentz, Baker Mayfield, T.Bridgewater
Poor: B.Bortles, Daniel Jones, Marcus Mariota, Kyler Murray, Sam Darnold, M.Trubisky
The Bears should not try to build a team to “carry” Justin Fields. Trade him.
Good article but on a side note, I’ve never really understood why people prefer adjusted EPA/play to raw EPA/play.