Quantifying Aaron Rodgers’ supporting casts

Mike Sando of The Athletic wrote his annual column on quarterbacks earlier this year. This annual column almost always puts Aaron Rodgers in the first and best tier regardless of how he produces on the field. Sando’s tiers are based on surveys of NFL scouts, coaches, and other people who actually work inside the game, and should have some idea of what they’re talking about.

This is in stark contrast to “analytics people” (myself included) who know with a high degree of certainty that Rodgers is no longer as productive as he was in his prime, and have posited a number of reasons why this might be so. These include age, injuries, overly conservative play, overly conservative play-calling, and a lack of supporting cast. This last one is the big one, as there is a belief among the analytics crowd that their numbers should see through a lack of support, and citing certain seasons where other elite quarterbacks (usually Brady and Brees) lacked a good supporting cast and still produced.

I tend to side with the analytics crowd on this question and I’ve watched enough tape to have some confidence in Rodgers’ decline, but kudos to Sando for putting in a ton of work for a companion piece specifically on Rodgers this week. First, Sando watched and charted all of the production lost on Rodgers incompletions last season, and assigned blame for the play. He compared his results to Rodgers’ outstanding 2014 season where he led the league in ANY/A, DVOA, and any other metric worth mentioning. What he found was that Rodgers’ supporting cast in 2019 cost him greatly: over 500 yards of lost production, 22 first downs, and 7 touchdowns versus 2014.

It’s interesting to note that they only cost Rodgers 9 receptions compared to 2014 according to his charting, but they were especially poor in hauling in big plays. This makes intuitive sense to me, as the more dynamic days of Nelson/Jennings/Cobb/Jones/Driver/Finley early in the decade led to more medium-to-big chunk plays more routinely. Per Sando, Rodgers is still throwing and executing on those plays, but his guys just can’t haul them in. He cites several specific examples.

If you’re like me, you may be skeptical of Sando’s self-charted numbers. After all, who among us hasn’t been burned by the random PFF grade in the past?But after reading a few times and listening to the most recent episode of the Off the Charts podcast in which Sando is a guest of Aaron Schatz and Matt Manocherian, I’m fairly convinced. There is something here, at the very least.

Manocherian works for Sports Info Solutions, and they produce a metric called “Cumulative Total Points.” To calculate Cumulative Total Points, SIS takes the “expected points added” on the play and assigns credit to every player on the field using all of the various metrics they track. This goes well beyond the typical box score stats and includes contextual elements on things like throwaways. Last season, despite a down statistical year, Rodgers was actually first in Cumulative Total Points for Quarterbacks, and 4th in Total Points per Snap. Essentially, the numbers we have available from SIS tell the same story as Sando’s analysis. Rodgers is as good as ever, or at least was in 2019. Give him some horses, and it will show up in his production metrics. As Sando notes, the Packers haven’t made a wide receiver selection in the first two rounds since 2014 when they nabbed Davante Adams.

Tape and Chemistry

I’ve been thinking a lot about the tangible effects of team chemistry lately, and specifically this piece by Russell Carleton who has at least put a framework around the tangible impact that team chemistry could have on a baseball team. Chemistry is one of the last unsettled frontiers for a sophisticated front office, but the big takeaway from what we do know, as put forward by Carleton, is elegant in its simplicity. Players are people, and we know from many many studies, as well as basic life experience, that the work environment has an enormous impact on productivity.

This should not be surprising to anyone who ever worked on a group project in school or as a part of their job. Some teams collaborate extremely well as individuals lend their talents to the good of the overall project. Others will be dysfunctional, with the most capable member doing what they need to do for survival and everyone else mooching off of their effort. On a dysfunctional team, the non-contributors are incentivized only to tie themselves to the capable and the capable, without hope of true success, will only do enough to ensure they get a good grade or keep their job. Which leads me to this throw.

This is from the Seattle game in 2018, which led to a three-game losing streak culminating in the loss to Arizona that ended the McCarthy era. I’ve often cited this throw as a perfect example of Rodgers’ decline. MVS is open on the play as Adams screens off his man, and a better throw likely gets the Packers a first down and keeps them in the game, and in the playoff race. Despite the stakes, and MVS being open out there, Rodgers failed to get the throw out accurately, and frankly, looked like he didn’t care. And I think the key here is that he probably didn’t care.

Look at the play before.

MVS catches the ball, and instead of turning upfield and likely (maybe?) converting a first down, he runs horizontally, gets drilled, and brings up a 3rd down in Packer territory. It was probably frustrating to watch. In this game Davante Adams did all of the heavy lifting, catching 10 balls on 12 targets for 166 yards. The rookie duo of MVS and Equanimeous St. Brown combined to catch 2 of 7 targets for 24 yards, and yet in crunch time you have Mike McCarthy calling short gimmick plays for MVS. MVS, who was only good as a deep threat (and still is), and has never caught more than 52.1% of his passes in a season. Rodgers was throwing again to a player who had done nothing in this game so far except screwing up the previous play. It’s probably infuriating.

We expect athletes to perform at an elite level all the time because they’re highly paid professionals, but that’s not how life works. Watching Rodgers over the last few seasons is also often infuriating, but think about what it’s like to be him actually having to throw to these guys. It makes complete sense that when he spots a guy open deep for a big play he’s still firing super accurate missiles, and when his receivers let him down, that he’s not as inclined to use proper footwork or technique on the next throw, because hey, what’s the point in doing so when you know your receiver will just let you down on the important plays.

Viewing the Packers through this lens helps to reconcile some of the differences between the Sandos and the Ben Baldwins of the world. Baldwin is happy to accompany his stats on Rodgers with a gif of Rodgers throwing the ball away or missing an easy pass.

(Note: FWIW, Baldwin is skeptical of the SIS data involved:)

It’s undeniable that Rodgers hasn’t been as productive since 2015, but it’s worth noting that his raw completion percentage is way down over the last two seasons, which happens to be the EQSB/MVS era, and losing 9 additional passes to drops isn’t the difference between 62% and 65%. (For the record, his completion percentage would only have increased to 63.1% if you give him 9 extra completions, still low for “elite Rodgers.”) Sando may very well be right about the production lost on all of the big plays he cites, but Rodgers is still less accurate generally as shown by his completion percentage.

I think my explanation reconciles both views nicely. We’ve seen a ton of Rodgers noticeably disgusted at his receivers. We’ve seen it for years. But we’ve also seen him be, frankly, lazy on easy throws that most QBs have to bolster their completion percentages. When you don’t trust your supporting cast, the little things slip.

G-mo, Graham, and MVS

I’ve mostly cited MVS here, but it’s worth noting that he is hardly alone. Sando lays plenty of blame at the feet of Jimmy Graham, and it’s undeniable that Geronimo Allison was one of, if not the worst receiver in football in 2019. Allison had a ton of drops, and failed on several plays that cost the Packers first downs. In Sando’s piece he quotes personnel as mentioning that Graham is almost always in the right place, which is why he was targeted so much, but that 63.3% catch percentage tells most of the story. MVS’s drops explain why they lost so much yardage per drop, as his plays are very boom or bust, and in 2019, they were mostly bust.

If you want to be optimistic about 2020, consider that Allen Lazard, who was far more efficient than Allison or Graham (in a limited sample size), will absorb many of the targets that previously went to both. Even if Lazard is merely average going forward, the upgrade from Allison to average would be one of the biggest upgrades at any position in football. Graham’s departure could also pay dividends by forcing more targets to more efficient receivers. Graham was a “worst of all worlds” option, drawing targets from Rodgers but frequently dropping the ball or doing nothing with it once he hauled it in.

On the other hand, tight ends play a large role in the LaFleur offense, and there’s a good argument that the remaining tight ends are actually worse than Graham, at least as receivers. MVS has purportedly had a good camp, but he’s still shown very little, and he likely opens as the team's third receiver. Adams should be his old self, but if you believe everything above, the success of the season, and a rebound from Rodgers is largely dependent on one MVS/ESB/Jace Sternberger breaking out or Allen Lazard being a true diamond in the rough, even with an enormous increase in targets.

The good news is that Rodgers may not only be fixable, but be fixed. So, how much do you trust Allen Lazard, and how do you feel about them not taking a receiver in a historically great class now?