Much Better But Still Not Very Good: The 2025 Caleb Williams Story

The Bears are on a hot streak. So why aren't we hearing Hosannas about Caleb Williams? (Studies stats.) (Watches film.) Oh. Ohhhhhhhh.

Somewhere between Drake Maye’s ascendence to the Pantheon of Elite Mount Rushmores and J.J. McCarthy’s descent to the bowels of the Vikings’ depth chart; between Jayden Daniels’ worrisome new injury and Bo Nix’s paint-by-numbers game management (until late in the fourth quarter, when he becomes El Greco); and between the Michael Penix who helped upset the Bills and the Penix seen lying on his back like a baby turtle on the freeway in the fourth quarter against the 49ers, there sits Caleb Williams: the most talked-about second-year quarterback that no one except Troy Aikman is talking much about.

The Chicago Bears are 4-2, having won four straight games. Granted, the Raiders and Saints were two of those opponents. But the Cowboys and Commanders, despite their flaws, aren’t exactly easy outs. And the Cowboys and Saints wins were rather lopsided.

Ben Johnson has clearly doused last year’s circus fire. That means Williams has been fixed and is on track to excellence. Aikman was just being a boomer when he criticized Williams on Monday Night Football: I, too, am getting close to if-you-miss-a-meeting-without-calling-you-are-dead-to-me years old.

Right? Right???

Well … Williams is undeniably a better quarterback than he was last year. But he was a disaster last year. With Johnson’s guidance, he has become non-disastrous. But non-disastrous is a long way from on track for anything.

It’s time to dive into the numbers to check on Williams’ progress and try to determine what’s next for Johnson’s Bears.

Caleb Williams 2025: the Basics

Let’s start with something deceptively simple and easy to overlook in our rush to scrutinize film and sift through splits: raw stats:

2024: 6.3 yards per attempt, 62.5% completion rate, 87.8 rating, 5.09 ANY/A, a 10.79% sack rate.

2025: 7.3 yards per attempt, 61.1% completion rate, 92.9 rating, 6.74 ANY/A, 5.61% sack rate.

Williams averaged 208.2 yards per game in 2024 and averages just 225.2 yards per game this year. Yardage totals are likely to go down a bit as a quarterback improves – garbage-time production gets replaced by handoffs and kill-the-clock tactics when a team starts winning – so the uptick in raw yardage is encouraging. Williams is also getting 10-15 yards of net offense back per game by avoiding sacks. His touchdown rate, not listed above, has gone up from 3.6 to 4.9.

Williams’ completion rate is down slightly this year. You and I may not care much about completion rate, but Ben Johnson does, and 61.1% remains below the NFL average of 65.5%.

Williams’ interception rate has risen from 1.1 to 1.6. Tiny fluctuations in interception rate can have an undue influence on the league’s efficiency rating, which only sees a slight improvement in Williams’ performance. We will talk more about interceptions in a moment.

Sacks are a quarterback stat, and Williams cutting his sack rate nearly in half is the big story here. A 16% increase in yards-per-attempt is also a big deal. Williams ranks 14th in the NFL in Adjusted Net Yards per Attempt this season. He ranked 31st last year. Eliminating about two sacks per game does wonders for an offense’s efficiency.

Williams ranks 19th in the NFL with a DVOA of -1.5. So DVOA sees a slightly below average passer. Williams ranked 28th in 2024 with at -12.0%. DVOA may tell the whole story I am trying to tell in 2,500 words with a pair of numbers. But the elaboration will be worth the journey.

Interrogating Caleb’s Raw Stats

FTN Network charges Williams with eight turnover-worthy throws, a high figure for a quarterback who has played just six games. Those throws include Williams’ three interceptions.

There were between two and three turnover-worthy throws in the Week 6 win over the Commanders, including a dropped Mike Sainristil interception near the goal line on a series that ended with a Bears field goal. Terrion Arnoled dropped an interception in the loss to the Lions. There have been a few batted passes at the line which could be considered turnover-worthy, including Maxx Crosby’s actual interception against the Raiders.

Williams was charged with just 11 turnover-worthy throws all of last year. It’s hard to throw near-interceptions when you are just hurling sail balls or eating dirt!

It bears monitoring as to whether Williams is trading sacks for potential turnovers. For now, let’s note that Williams sometimes wings short passes into heavy traffic between the numbers: a potential problem in an offense heavy on high-precision short passes between the numbers. At the same time, let’s not penalize Williams for interceptions he did not throw.

Let’s also look at YAC for a moment. Two of Williams’ three longest passes of the year so far traveled less than zero air yards: a 41-yard swing pass to D’Andre Swift against the Cowboys and a 55-yard screen to Swift for a touchdown against the Commanders. Here’s a list of Williams’ longest throws of 2025 for those who like to get their fingers grubby with data.

Williams is not, however, benefitting too greatly from Johnson-manufactured micro-passes. Williams ranks 16th in the NFL with 45 passes at or behind the line of scrimmage and 21st with 167 yards on such passes (per Sports Info Solutions). Even adjusting for playing just six games, he is nowhere near league-leaders Bo Nix and Patrick Mahomes in these categories. (Aaron Rodgers, who has also played just six games, has thrown 58 passes behind the line of scrimmage for 255 yards.)

Williams ranked third in <1 air yard pass attempts with 164 in 2024, behind Mahomes and Nix, as offensive coordinator Shane Waldron and his replacements tried to jumpstart the Bears offense with a steady stream of screens that opponents saw coming. This year, Williams is throwing fewer micro-passes, but enjoying them more. That’s a sign of a healthier offense.

Caleb Under Pressure

The sharp decrease in sacks invites a closer look into Williams’ statistics under pressure. So here they are:

2024: 14.6 dropbacks per game, 46.4% completion rate, 5.2 yards per attempt, 1.2 ANY/A, 4 interceptions, 4 sacks per game.

2025: 12.0 dropbacks per game, 41.8% completion rate, 4.6 yards per attempt, 2.3 ANY/A, 1 interception, 2.2 sacks per game.

The reduction in the number of pressures per game is not huge, but it’s interesting, because Williams is NOT getting the ball out quicker.

Williams’ time to throw has increased from 2.92 seconds in 2024 to an NFL-high 3.17 seconds so far in 2025, per Next Gen Stats. I rarely use time-to-throw, because I think it’s noisier than a middle-school cafeteria at Halloween. There are too many variables in the soup.

For example, Williams had a series of scramble-around-and-around throws in the second quarter against the Saints. Based on the little clock at the bottom of the video, he took 10 seconds to get the pass off in at least two instances. I counted three other scrambles where the ball took between six and seven seconds to leave his hand after the snap. A handful of plays like that can skew a 185-pass sample that is calculated down to the hundredth of a second. Two ten-second passes and four one-second passes equal a quarterback taking four seconds per pass to get rid of the ball!

It’s worth noting, though, that the five extra-long time-to-throw plays I studied for the paragraph above were incompletions, in a still-close game, against a weak opponent.

Time-to-throw issues aside, Williams appears to be a better decision maker under pressure than last year, because he has replaced half of his sacks with something. Even moving sacks to the incomplete column is a step in the right direction.

Yet Williams is not a very good passer under pressure: 2.3 ANY/A ranks 25th among the 37 quarterbacks who have thrown 25 passes in such circumstances. Williams ranked 38th under similar parameters last season.

To repeat the refrain: “much better” does not mean “great,” or even “fixed.”

Caleb in the Pocket

Williams appears to be a more efficient pocket passer than last year. Here are some raw numbers:

2024: 31.4 dropbacks per game, 66.7% completion rate, 6.5 yards per attempt, 10.7% sack rate, 5.3 ANY/A

2025: 25.1 dropbacks per game, 66.1% completion rate, 8.4 yards per attempt, 5.1% sack rate, 7.9 ANY/A

Williams is staying in the pocket for six fewer passes per game this year than last. In part, that’s because he is throwing three fewer passes per game than last year, in part because (again) he is playing with the lead more often. But his designed rollout and scramble rates have not changed significantly. The film suggests that he doesn’t decide where to go with the ball quickly enough and/or still bails from a manageable pocket too soon on too many occasions.

In order to get a better sense of Williams as a downfield passer from the pocket, as opposed to the guy we saw flinging a zillion screens for parts of last year, let’s eliminate the passes of less-than-five air yards from the data:

2024: 17.2 attempts per game, 52.9% completion rate, 7.5 yards per attempt, 7.6 ANY/A

2025: 19.3 attempts per game, 55.2% 8.6 yards per attempt, 8.7 ANY/A

Once you begin to fiddle with air yards, sacks disappear from the data, because sacks don’t contain air yards. Yards-per-attempt and ANY/A also converge significantly, particularly for a quarterback with a high sack rate.

The data above reveals that Williams has indeed improved as a “real” pocket passer. He throws downfield more often than in 2024, at a higher completion rate, for more yards per attempt. As already discussed, his sack rate has decreased sharply. His pressure rate on such passes has increased from 26.6% to 35.3%, which suggests that we cannot attribute the change solely to offensive line upgrades or a visit from the Saints.

Now, the familiar bad news: 8.7 ANY/A on passes of five-plus air yards from the pocket ranks 20th among quarterbacks with 25 or more such attempts in 2025.

Caleb’s Worrisome Accuracy

Watching Williams’ tape, I was struck by how often things just don’t look good. Chase Young batted two passes at the line in the Saints game. Maxx Crosby volleyball-spiked an interception to himself, but also had a batted pass earlier in the Raiders game, as did Jonah Laulu. Shallow-cross passes arrive behind the receiver or rocket ahead of him. The Bears passing game rarely looks crisp. And Johnson both demands and has promised us kettle-cooked potato chips.

Williams’ bad throw percentage, per Pro Football Reference, is 20.4%, the fifth-worst in the NFL. Michael Penix, Cam Ward, Russell Wilson and Jaxson Dart have worse figues.

Per FTN Network, Williams’ accuracy percentage of 66.9% ranks second-to-last among eligible starters, behind Penix but ahead of Dillon Gabriel.

Next Gen Stats has a metric called expected completion percentage. It essentially prorates completion rates based on the depth of each throw, punishing quarterbacks who dink-and-dunk their way to gaudy rates and (by extension) passer ratings. Next Gen Stats then uses a plus/minus system to compare their expected completion percentage to the actual percentage. According to these metrics, Williams should have completed 70.8% of his passes, not 61.1%. His -9.7 percentage point differential is the NFL’s worst, well ahead of Trevor Lawrence and Gabriel.

Over in the Sports Info Solutions database, I filtered out passes to the sidelines to focus on what I think of as “Ben Johnson” throws, as opposed to sideline shots (which Williams often attempts before going out of bounds). Williams’ on-target rate of 68.0% ranked 38th out of 38 quarterbacks with 25-plus such passes.

These accuracy metrics are historically less useful than you might think. Patrick Mahomes’ Next Gen Stats plus/minus on completion rate is -2.3, below the league average. Spencer Rattler has a high accuracy rate on non-sideline passes. There are all sorts of variables at play, including scouting peculiarities among various stat providers. But finishing in the bottom five in just about every available accuracy metric is undeniably bad.

Combine the fact that Williams ranks at or near the bottom of the league in accuracy metrics AND in time to throw, and the sustainability sirens start to go off. A quarterback who finishes last in a bunch of categories but near the middle in others is either some sort of unicorn or needs to improve on his weaknesses significantly to keep them from dragging him down.

The Doctor Who Got C-Minuses in Med School Can Still Practice Medicine

I hoped to see a whole different Caleb Williams when I watched his film. I was watching victories, after all. Against some mediocre-to-awful defenses. Drake Maye has looked like Thor for a few weeks under similar circumstances. I didn’t expect that level of dominance, but I was hoping to see something broadly similar to Jared Goff: smooth, methodical ball distribution. I conceived this feature as something a little more upbeat after skipping the Bears in my good teams that stink feature from Tuesday.

Instead, I saw defenders like Maxx Crosby and Chase Young swatting down quick screens into the slot. Those plays told me that Johnson may be relying on some of the same tactics Waldron used to get Williams into a rhythm, though not as frequently, and that defenses have figured it out.

There is also still far too much Yakety-Sax scrambling going on. Williams is neither Daniels nor Maye on the run. He spins away from would-be sacks very well, but then he drifts toward the sideline in search of a downfield hero ball. Too many Bears plays look like that old vibrating Electric Football game, where the little plastic players ended up clustered together on one side of the field. That was a problem in 2024. It’s about 25% less of a problem now.

Also: Swift having to reach down to scoop up routine swing passes, wide-open Cole Kmet twisting his body for fastballs up the seam, receivers on mesh concepts looking at each other and wondering who the intended receiver was. The film shows a level of imprecision that the stats have not picked up.

Luckily, the Bears keep facing the Saints and Raiders, and the Commanders kept fumbling for them. The running game clicks a little. The defense makes some plays. Jake Moody actually kicked a few field goals when called upon. The Bears, like their quarterback, can now at least seize the opportunities which are handed to them.

Williams has improved. But he is not improving game-by-game the way Maye has in a similar situation (i.e. against weak defenses). We’re past early-season questions about whether Williams might be benched for Tyson Bagent, but Johnson cannot be satisfied with what he’s seeing right now.

It’s hard to imagine the Bears going on a run against late-season opponents like the Packers (twice), Eagles, Lions and 49ers with Williams still running in circles and spraying footballs for long stretches of games. We’ll know more, about both Williams and the Bears, after the two-game Ravens/Bengals road trip that’s about to start. Both the Ravens and Bengals are at a crossroads, yet both would beat the Bears if Williams plays the way he has for most of his last three victories.

In case you missed the Williams-Aikman “beef:” Williams apparently missed a production meeting with the Monday Night Football crew, which led to speculation that Aikman was being extra critical during the broadcast.

Vindictive criticism would be out-of-character for the understated Aikman. As for missing production meetings: that’s one of those no-big-deal things that would be easier to overlook if Williams had a reputation for diligence and meticulous attention to detail. Both Williams and Johnson also offered barbed responses to Aikman’s criticism. Champions don’t allow 58-year old broadcasters who talk like Hank Hill to get under their skin.

Thankfully, the Aikman saga was the only extracurricular Williams drama since the season started. We no longer have to speculate about Williams’ punctuality or workplace readiness. We have data and film. Both are, let’s say, provisionally encouraging. I have seen a lot of quarterback prospects who had rough rookie years flatline in their second seasons. Williams is by no means flatlining.

But there is a lot of work to do, not to go from good to great, but to go from adequate to good.