As Yogi Berra said, “it’s getting late early” and if past data on QB development holds, it may be doing just that for Hurts…
I was optimistic on Jalen Hurts’ ability to correct his primary issue (accuracy / timing) because of his Completion Percentage Over Expected (CPOE) stats in college which usually translate well to the pros. And he has improved, from a second-worst -8.3 CPOE in 2020 to a middle-of-the-pack +0.8 CPOE in 2021. But 0.8 is not good, putting him 23rd out of 33 QBs this year. And more concerning is the lack of progression – he has actually been getting worse as the season has progressed:
I wrote recently on his struggles in the game’s highest volume passing depth, the short 1-10 yard routes. Hurts still only has 1 game in his career where his 1-10 yard CPOE was positive (week 4 vs. Kansas City). And the one area where he did well last year, the deep 20+ yard targets with a very good 15.4% 2020 CPOE, has also regressed, sitting at 2.11.
And his ability to create explosive plays, previously a strength, has regressed recently. Just counting total explosive passes and runs, Hurts is 5th in the league with QBs over 100 passing attempts but his garbage / non-garbage time splits have worsened through the year. Ignoring garbage time stats where win percentage is 90% or above, he drops to 22nd in the league in creating explosive plays with 17 of his 39 explosive passes coming in garbage time and 5 garbage time explosive passes in the Raiders game alone.
What historical data tells us about QB development
Credits to @ConorMcQ5 for the inspiration here as he looked at how a QB’s CPOE progressed through their college career and after how many passing attempts a QB’s CPOE centered around their end career CPOE.
I did the same for pro QBs below, looking at all QBs from 2010-2021 and the results are similar although it takes a bit longer for pro QB CPOE stats to stabilize (which is probably explained by the longer careers in the pros).
The first chart shows the difference between a QB’s CPOE at a specific career passing attempt vs. what their ultimate career CPOE ended up being. At the far left of the chart are the first career passing attempts for QBs with data points stretching from -40 to +40 and as the chart moves to the right, it reflect the same difference as they have more and more career passing attempts. As an example, if you took a data point that is +15 at career passing attempt 1, that means that QB, on their first career passing attempt, had a CPOE 15 points higher than what their career CPOE ended up being. Obviously, the spread of CPOEs is much higher early in careers but narrows around the axis as their career progresses.
But when do we know, for the most part, when a QB is what they are? The next chart looks at the average difference from a QB’s career CPOE at each career passing attempt. For example, the leftmost data point is right above 15 which means a QB’s first career passing attempt is averages +/- 15% from what their end career CPOE is.
What is instructive is how quickly a QB’s CPOE nears their career CPOE. At 250 career passing attempts, the average QB is +/- 2.5% off of their career CPOE, at 1,000 career passing attempts they are +/- 1% from their career CPOE.
Averages can be misleading and there are always outliers. The classic example is Josh Allen who had poor college and rookie year accuracy numbers but did what most don’t and became one of the most accurate passers the past two years, second in the league with a CPOE of 6.3. and his CPOE is still increasing after 1,750 career attempts. The last chart I will show provides context on this variability in the data:
The box in boxplots show the middle two quartiles, meaning 50% of all QBs are contained inside the box with 25% of QBs below the bottom of the box and 25% of QBs above the box. The whiskers, or end lines, show the min and max data that sits outside these middle two quartiles. Some takeaways:
- At 500 career passing attempts, half of QBs are between +/- 1.76% and 2.04% of their career CPOE with the biggest outlier being 2.52%.
- At 750 career passing attempts, there is very little variability with 50% of QBs between +/- 1.51% and 1.61% with the biggest outlier at only 1.68% off their career CPOE.
What does this mean for Hurts?
Hurts is currently at 390 career passing attempts with a career CPOE of -3.6%. Can he get better? Sure. But he has not shown development here and historical data says that by 500 attempts, it is very unlikely to materially improve his accuracy. Even assuming the maximum improvement we have seen in past data of 2.52%, adding this to where Hurts is today gets him near a 0% career CPOE. That’s just not good enough.
You have to be careful using averages as a rule as they aren’t. There are QBs that didn’t center around their career CPOE until near 1,000 career pass attempts (Kirk Cousins and Dak Prescott for example). But as time goes on, the probability of improving dramatically continues to drop.
Hurts adds value with his mobility and we knew he would improve over the 4.5 games he played in 2020, but the fear was that he would improve to be a middle-of-the-league QB, leaving the Eagles in quarterback purgatory. That is what looks highly likely right now, especially considering there really isn’t an upward trend.
That is where I am now.
There has been a lot of recent debate on if Hurts can even be evaluated this year given the failures of Sirianni and the playcalling. Can we judge wins (which aren’t a QB metric anyway)? No. Can we look at the Eagles offense’s ability to construct drives and score? Also probably no as the playcalling and penalties are a huge impact. But there is no reason we can’t look at the things Hurts needs to improve on – the timing routes, his accuracy on short routes, his ability to consistently see open receivers. He is failing at each of these and there is no upward trend which is reflected in the data at the beginning of the post.
What does this mean for the Eagles?
The Deshaun rumors are exhausting, but they aren’t created out of thin air. It is clear Lurie and Howie would trade for one of the top QBs, whether Deshaun or Aaron Rodgers or Howie’s white whale Russell Wilson, if they could. The issue is they probably can’t. It is known that Watson does not have Philly as one of the teams he would waive his no-trade clause for. Rodgers and Wilson will control where they go as well and the Eagles seriously have to wonder if they will be a desired location given the teams that will be in the hunt for QBs.
Pittsburgh, Miami, Carolina, Washington, Houston, Detroit, New Orleans, Atlanta, and Philly all could be aggressively going after QBs (and if Seattle and Green Bay lose Wilson and Rodgers, add them to the list). The 2022 QB class is widely viewed as a weak one with probably Malik Willis, Matt Corral, and Kenny Pickett as potential first round candidates with maybe Sam Howell and Carson Strong in the mix. But most view none of them in the same class as last year’s group. Somebody will rise and a couple will be drafted high given there will be very needy teams. But there will be much more demand for QBs in 2022 than there are good options.
What do I think the Eagles will do? They will without doubt aggressively go after Wilson and Watson with their potential three 1st round picks but as I said, I think the Eagles have a tough uphill battle to win either of them.
Will they love one of the QBs in the draft? Maybe, I wouldn’t be surprised if this was their Plan B and they have had scouts at recent Pitt and Ole Miss games. I’m a Corral guy currently but don’t have great conviction on him.
But another option could be to stick with Hurts or Minshew (or a yet-to-be-name QB from the factory) for another year if one of them shows some hope this season, trade down and grab another first in 2023 to give them optionality on the next group of QBs.
Even though I think the Eagles fall into QB purgatory, I would continue to play Hurts this year barring a total collapse. The go-forward options aren’t great and there is still time for Hurts to show improvement. The plan was to see what he was this season and we should stick to that as the Raiders loss effectively puts the Eagles out of the playoffs (Dallas, unless something bad happens, has a good path to at least 11 wins and the Eagles will be scratching to find 9 if they right the playcalling and system now).
My issue is what if Hurts shows up against Detroit and the Jets and the Giants (twice)? What do you know from that? He has to show up against the better teams and with the weaker part of the schedule ahead, that means the Chargers game gets circled in red. I really wanted to see more against the better teams to really feel good about him.