The Chargers’ offense is rounding into form as we approach mid-season after putting up 37 points on theย Minnesota Vikings’ defense which was considered a vauntedย opponent at the start of the season. Game context matters and the Chargers’ defense helped to slam the door shut by the end of the first quarter however, it was still a wildly impressive display by Justin Herbert and the rest of the offensive players.
These numbers represent the best game of the season for the offense but it’s indicative of the upward trend of the overall performances as the Chargers get healthier and their young players are developing with more time on task. Joe Alt returning has provided more stability in the pass game and the development of the run game through Kimani Vidal’s improved output has provided the kind of consistency that can sustain drives across four quarters.
For me the most pleasing aspect of the Bolts’ recent offensive success has been the rookies finding more defined roles within Roman’s scheme. Oronde Gadsden II has been nothing short of a revelation, his combination of route running skills, athletic ability and football smarts is rare for any tight end but for a 5th round rookie, it’s sensational. To haveย found a way to combine all of those traits already means that he has become a genuine threat that defenses have to account for a mere 6 games into his young career. I was higher than most on OG2 as I had him as my TE4 with a high 3rd round grade but even I didn’t expect him to put it all together this quickly.
However this article is not another glowing review of Oronde’s breakout stretch, even though the young man deserves unending praise for the hard work it has taken to get here. Anyone who can change positions and add 40 lbs in two seasons is obviously dedicated to his craft. Instead, this article is about Greg Roman’s use of both Gadsden and another member of the Bolts’ stellar rookie class, and how their deployment could give defenses a head start in their game preparation.
I was inspired to dig into this by a fellow writer, Max Toscano, who wrote a brilliant article about the Chargers’ use of Oronde Gadsden II which I highly recommend reading. Max brilliantly spotted that Greg Roman is signposting when the Bolts intend on running the ball with his usage of #86.
“In a game that represent(ed) the Chargers preferences and tendencies well with a positive game script, the Chargers ran 27 total snaps of 11 personnel. Gadsden was on the field for 23 of those… On those 23 snaps, the Chargers ran the ball once. ONCE!”
Max Toscano, Remember The Tight Ends
Any analysis that spots niche trends will catch my eye so needless to say this deep dive piqued my interest. I therefore went into a mix of the film and data to better understand whether this was based on how the Vikings’ were playing each package or is this something we need to watch out for in the near future.
The data shows that the Vikings followed their normal operating methods by presenting a hybrid system that features no obvious patterns. None of their personnel choices represented above 28% of their snaps in neutral game script situations and for the most part they followed the Chargers lead by matching light boxes for 11 and odd fronts for heavier looks. So this then tells us that the Chargers adjusted choices of personnel was a premeditated decision and I believe I know what has led to this shift in approach.
Before Week 8 kicked off the Bolts had been in light personnel (10 or 11) on 66.9% of the time which was a huge disparity for a team with two 285+ lbs blocking H-backs on the roster. The Chargers showed a significant shift in their offensive approach in Week 8 compared to the previous weeks, with an emphasis of using much heavier personnel usage and a stronger commitment to the run game. This represents a shift back towards the typical brand of Harbaugh led power football we were expecting at the start of the season.
Gap Concepts from Shotgun:
Weeks 1-4: 21 attempts for 142 yards, 0.209 EPA/Play and a 42.9% success rate Weeks 5-8: 11 attempts for 59 yards, -0.046 EPA/Play and a 36.4% success rate
I believe the failings of the gun run gap scheme concepts that the Chargers started the season with have forced Harbaugh and his staff to change course. Without Omarion Hampton to make good on missed interior blocks and reduced gap spacing, the team has had to make a proactive decision to go under center, beef up the personnel and commit to the run in positive game scripts.
This came to fruition this week with a real commitment to going into under center and running the ball from heavier looks. The splits below illustrates the top three personnel groupings from Week 8 and as you can see there are some stark differences from their tendencies over the first half of the season.
22 Personnel (2 RB, 2 TE, 1 WR):
27 total plays (38.6% of all plays)
23 run plays (85.2%) vs 6 pass plays (22.2%)11 Personnel (1 RB, 1 TE, 3 WR):
27 total plays (38.6% of all plays)
9 run plays (33.3%) vs 22 pass plays (81.5%)21 Personnel (2 RB, 1 TE, 2 WR):
8 total plays (11.4% of all plays)
4 run plays (50.0%) vs 4 pass plays (50.0%)
The precise balance between 22 and 11 personnel was not simply a factor of game script either. In fact when the game was in neutral game script, the Bolts ran a higher rate of 22 personnel (31.6%) than 11 (26.3%). Despite a one game sample size usually being a waste of time, I think all of this shows that the Bolts are at least attempting to transition to a more balanced team.
They have shown they can genuinely pass the ball in 11 personnel with their revamped set of weapons providing Herbert with the outlets he needs to go into hero mode, even if his line can’t play their part consistently. Now they are flipping things back around to see if they can run the ball with a numerical commitment via heavier personnel to see if they can win using methods conducive to playoff football.
Creating variation
This all sounds very exciting and if Harbaugh and Roman can actually create the kind of balanced offense they have been seeking then their Superbowl window can start to be pried open. However (you knew there was going to be a however after that set up didn’t you) to be able to split tendencies into two different sets of personnel groupings you need to be able to run and pass from both in order to keep defenses guessing. That is where my cautionary instincts kick in as right now that is certainly not the case at the moment.
As Max realised, when the Chargers are in 11 personnel and Gadsden is on the field, they pass the ball 96% of the time. Which on its own should be a stark enough to draw the attention of opposing analysts. However when doing my own analysis I found is that Gadsden isn’t the only signal of Roman’s intentions. Tre Harris is another bastion of the play design, when he’s on the field you’re very likely to see a run. Now this is for a very good reason as the former Ole Miss Rebel is developing into a very good blocker to the point he’s being used on plenty of insert assignments which is surprising given how little blocking he was doing in an air raid offense.
However, as good as he is, if the defense knows it’s coming it’s not going to provide sustainable success. In Week 8 The second round rookie was on the field for 41 of 68 total plays (60.3%) and the Bolts ran the ball on 32 (78%) of those snaps. When Harris was on the field in 11 personnel, the Chargers ran 60% of the time, compared to just 17.6% when he was off the field and Tre was present for nearly all 22 personnel plays (24 of 26), which were heavily run-focused.
What this essentially shows is happening is that Gadsden, Ladd and QJ are on the field for pass plays and Matlock, Fisk and Tre are on the field on running plays. This isn’t hockey guys, this isn’t a line change and you need to be able to mix things up a bit more than that. This linear deployment of resources is not going to cut it against modern defenses. If you allow defensive quality control coaches to tee off on easily split tendencies like we saw in Week 8, then that narrows the game plan enough to give the opposing team an edge over you which you just don’t need to hand them.
The next evolution
Again, this is a one game sample and the Week 8 game plan could have been made with a “we’ll do whatever we want to” mentality against a team they were confident they had an advantage over. However if this is the case and the team are planning on continuing this allocation of resources then there are some adjustments that will need to be made in order to build an offense that can grow into the second half of the season.
The first move would be to let Oronde Gadsden II run block in 11 personnel. The tape shows that he’s not that far behind the other players who are being asked to block in-line and the only way he’s going to get better is by being on the field for them. If the team has aspirations of playing football in January and beyond then they need to let go of the shackles to enough of a degree to alter the way that defenses handle planning for that package. I am not saying the coach staff needs to reimagine their entire methodology but they should be trying to introduce enough deviation to take it from a predictable behavior down a few percentage points towards a slight bias.
55-60% = Mild lean
65-70% = Consistent bias
75-80% = Clear tendency
85+%+ = Predictable behaviorGeneral margins for tendencies in standard sports analytical theory
This theory would also apply to Tre Harris’s use, the team needs to get him onto the field in 11 personnel looks earlier in the game and they need to pass the ball once he’s there even if he’s not even running a route within the progression. I’m making it sound like he’s not a good receiver in himself when that’s absolutely not the case, it’s just that he has four more established options ahead of him so it’s more taking those players off the field to put him on that is causing the staff to hold back.
Now looking beyond the stretch of the next few games, the long term vision for Tre should be akin to how the Rams, and now the Seahawks, have used Cooper Kupp to allow 11 personnel groupings to run bigger body personnel concepts. I think the coaching staff are already leaning into Tre’s talents however they need to be careful not to let him become so specialised that teams will be keying off on his appearance.
To enable both of these adjustments it will take a commitment to reducing the amount of looks with both Fisk and Matlock on the field. A week ago I would have said that there was very little chance that this would ever come to fruition as this offensive staff has previously stuck to their guns however the team has shown it isn’t afraid of moves like this as they have seemingly benched Will Dissly in favor of Tyler Conklin as their blocking tight end. Which, by the numbers, seems to be a good call but it’s still a bold call considering what Dissly brought to the table last season.
If Roman is going to commit to the run then he needs to be flexible enough in his personnel choices to pass from heavier sets and ensure those targets aren’t being wasted on four blocking tight ends who have either forgotten how to catch or never should have been near the ball in the first place.
If theyโre in 12p/22p/13p, they canโt really run their dropback pass game and spread the field. You do have to honor the run a bit AND stop PA, but you donโt have to really cover Matlock, Fisk, or Dissly when heโs back, so you can play 2 high and still load up the front
Max Toscano, Remember The Tight Ends
So in reality how do they do this, how do they maintain success rates in both the run and pass whilst increasing the variance of personnel as to not give away their intentions? I think it starts with changing the 11 personnel choices. They need to get Oronde Gadsden II on the field as the lone in-line blocker and run the ball from these looks. Then, at the same time, they need to have Tre Harris on the field in the same looks and pass the ball.
In Week 8 they had them both on the field in 11 personnel on 6 total snaps, 4 of those were passes and 1 was a scramble and 1 was a designed run. This play was scuppered by a slipped block by Gadsden who was asked to base block Dallas Turner, a first round edge rusher, but he wasn’t far away from making this successful. OG2 just needed to adjust his inside foot enough placement a few inches to his left in order to produce the force necessary to twist Turner out of the gap.
I’d like to see the rookies take a more prominent role on other personnel packages too, when they were both on the field in 22 personnel (19 plays) good things happened. These looks earned +0.35 EPA/Play overall with their 16 (84.2%) runs going for +0.10 EPA/Play and their 3 passes going for a whopping +1.70 EPA/Play.
Most of these snaps came with both Fisk and Matlock on the field but I’d be keen to see Tre being slid inside to act in the Cooper Kupp pseudo-TE role and then you can have another passing weapon on the outside. This is especially prevalent as Roman has worked with a lot of tight split bunch concepts which allow you to run heavy personnel concepts from lighter personnel.
In summary, the lean into the heavier personnel worked brilliantly against the Vikings however this split in personnel choices has left Roman with a tell that any half decent defensive coordinator could key off on. I guarantee the Titans linebackers are being told this exact same thing in their meetings this week and without addressing it they risk bringing Tennessee’s defense back into a game in which the Bolts should have a sizable advantage.
This was very interesting, thanks Ryan. My main observation is that you didn’t really comment much on game situation.
Could Harris, Fisk, and Matlock have been in on so many rushing plays because the game situation was favorable for running? I think it is reasonable to believe the Chargers want a balanced offense, and one thing that means is running more often than many other NFL teams in such situations.
I checked Harris and found that he had 16 snaps in the first half. 5 were dropbacks, and 11 were RB runs. Of the 11 RB runs, 6 might be viewed as favorable running situations:
2nd & 4 at MIN 35
1st & Goal at MIN 8
1st & 5 at LAC 46
2nd & 1 at 50
2nd & 3 at MIN 37
1st & Goal at MIN 3
The other 5 times were all on 1st & 10.
None of his 2nd half snaps came with a Chargers lead lower than 14 points.
With this context, perhaps the heavy running with Harris on the field was less of a tendency and more game script driven. In running situations, maybe there is less of a concern about tendencies and tipping the defense and more of a straightforward desire to just get the best players for running play execution on the field.
Could Gadsden have been in on so many pass plays out of 11 personnel because he was in on a lot of obvious passing situations earlier in the game when the game situation was neutral or closer to neutral?
I checked his snaps independent of personnel grouping and found that he had 32 offensive snaps in the first half. 22 were dropbacks and 10 were RB runs. Of the 22 dropbacks, 16 were in passing situations:
2nd & 13 at LAC 14
3rd & 13 at LAC 14
2nd & 11 at LAC 30
2nd & Goal at MIN 10
3rd & Goal at MIN 8
2nd & 11 at MIN 36
2nd & 17 at MIN 31
3rd & 17 at MIN 31
2nd & 7 at LAC 9 with 3:27 remaining in 2Q (4 minute offense)
3rd & 5 at LAC 11 with 2:47 remaining in 2Q (4 minute offense)
1st & 10 at LAC 17 with 2:08 remaining in 2Q (4 minute offense)
1st & 10 at LAC 33 with 1:57 remaining in 2Q (2 minute offense)
๏ปฟ1st & 10 at MIN 49 with 1:29 remaining in 2Q (2 minute offense)
๏ปฟ2nd & 3 at MIN 42 with 1:16 remaining in 2Q (2 minute offense)
1st & 10 at MIN 27 with 0:55 remaining in 2Q (2 minute offense)
2nd & 10 at MIN 27 with 0:50 remaining in 2Q (2 minute offense)
I would guess that the defense was expecting pass on all of these plays whether or not Gadsden was on the field.
5 of the other 6 dropbacks were on 1st & 10. All of Gadsden’s first half snaps came with the Chargers leading by 0-11 points.
Maybe these and similar examples aren’t wholly explanatory about the data you posted, but perhaps explain some portion of the samples.
The number one thing I really liked about Roman’s game plan this week was the under center splice action run game from 22 personnel so as much I’ve criticised the one dimensional nature of it, the avenues from this path are wide open. Let’s revisit this with another week of data and film!
This was as much a criticism of predictability as it was wanting Tre and Gadsden to build a chemistry between them.
Ryan,
Another very detailed and interesting read. I have always been a big fan of using the same formations and groupings to be run with multiple plays off of it. As a player, film study was always more difficult when you could not pick up on any particular tendencies or hints just by personnel or formation. The key is to not get repetitive with the plays being used from such sets. This very well could be a situation where Roman is deliberately being so repetitive until either teams realize it and begin to stop it, and/or he has other plays set up to be run in the future. After all, with all the film study an analysis and analytics in today’s game it would be malpractice to reveal all your tricks before midseason and I am sure he is constantly adjusting and tweaking his plays weekly as part of the gameplan or leaning into what has been successful. We still have plenty of games left to see the evolution of these play calls.
Great read.
Could this be a deliberate tactic to set up opponents in the future? Or is it not feasible to all of a sudden start passing out of certain running formations or vice versa. I guess what I’m trying to say is that I can’t judge the degree of difficulty it is for an offense to switch to a different tendency all at once. I suppose it’s not that easy to suddenly flip the switch and that those things happen more gradually most of the time.
All I know is it felt like a rather complete victory against the Vikings. I sincerely hope they can build on that instead of that this game plan gave future opponents the blue print on how to attack/defeat the Charger’s.
PS: I haven’t come to answering your post on the other article. New York was absolutely incredible. Meeting and talking to Kyle and other Chargers fans was such a fun experience. New York in general is also such an awe inspiring place to be. Really fun trip all around.
Excellent article, Ryan.
I’m hoping that they start to mix things up in all the ways (and for all the reasons) you’ve written about.
Yet there’s also this part of me that gains comfort from knowing that Brian Flores probably knew all of these things too on TNF. When Gadsden is the lone TE in 11 personnel it is more likely to be pass. When Fisk/Matlock/Tre are all on the field it’s more likely to be run. This was all intuitive but they couldn’t stop us anyway.
So, yes – let’s mix it up. But let’s also be confident that even when you know what’s coming as a DC, it isn’t easy to stop 2x 270+ pounds FB/TEs, and it isn’t easy to stop m—–f—–ng Oronde Gadsden marauding through the middle of the field!