Closing Time Pt II: Offense
This one was close on the scoreboards, but not really on the field… (Niners Noise)
I’ve said before that close games in the NFL are—in the long run—largely 50/50 affairs. The teams that are consistently strong every year are the teams that handily beat the most opponents because a larger winning margin means there’s less chance that a bad bounce, bad call, or single mental error can cost you the game. That’s why a team that has had an otherworldly record in close games one year, like the 2018 LA Rams who went 8-1 in such games (and hopefully the 2019 Seahawks, who went 11-3 in said games last year), is often a prime candidate for regression the following season.
Many believe that an elite quarterback playing at an MVP-ish level greatly changes close game win probability, and while that kind of play under center may provide a boost in spurts, the long-term stats tell a different story.
That’s six first-ballot Hall-of-Famers on six of the winningest teams of the past decade, but —other than the Belichick-Brady machine—none of them are tilting close games heavily in their favor. So what determines how well a team plays down the stretch? And how do we get on that evil Patriots level?
For these purposes, “closing a game” means the offense retains a fourth quarter lead, independent of how well or poorly the defense plays when they’re on the field, and there’s no great stat to measure that. The old world of thinking prioritized time of possession, but in recent years spread-heads and analytics nuts have said there’s no correlation between TOP and winning. That is, like most things, a half-truth.
Yes, while going out of your way to win time of possession above all else is pointless, running more plays, with better success, and converting more yards into first downs is clearly a recipe for winning games, and—particularly in the NFL, where no huddle attacks can only go so fast—that typically means you have the ball longer. When you have the ball for a long drive that usually means you’re getting first downs and either scoring or creating a field position advantage, and scoring (that thing that determines who wins the game) and field position (the greatest determinant towards opposing offensive success) are pretty important. Finally, regardless of your opinions on what the score needs to be and how much time needs to be left in the game before you start bleeding clock, I think we can all agree that if you’re up late and can ice a game by keeping possession, your goal is to do just that—keep possession and win the game. So if you’re a team that’s successful and finds itself up more often while late in games, you’re more likely to have a time of possession advantage.
Over the past three seasons, the top five teams in time of possession have made the playoffs each year—that’s 15-of-15 teams with 23-of-30 of the top ten in T.O.P. going to the post-season over that period. That would seem to imply time of possession at least has some positive correlational effect towards winning games, and—in the 2019 regular season—the Niners were fifth in the league in time of possession and second in the league during the fourth quarter. Theoretically, that means we don’t have a problem.
But if we were to isolate time of possession into the factors that largely make up the positive aspects of the stat (getting yards, getting first downs, etc.) while ignoring the more pointless elements (running the play clock down when you can’t ice the game, taking way too damn long to call plays, etc.) then we might get a clearer picture.
For instance, when you have the lead in the fourth quarter, there is no instance—except, perhaps when you are in an absolute blowout—that you aren’t trying to get yards and first downs. Yards give your defense better field position, which is the biggest factor towards opposing offensive success; first downs give you a new set of downs which churns clock and gives you extra chances to gain yards.
The Niners ranked 10th in offensive yards/drive (33.58 yards) while 5th in points/drive (2.47 points). The first figure is probably somewhat deflated based on the strong play of our defense, which often gave us better field position and capped the number of yards we could travel, while the second figure is likely inflated due to the exact same thing.
Then perhaps the best stat is Football Outsiders’ Drive Success Rate (DSR)? This stat, in which we ranked ninth, examines the percentage of offensive drives that either scored points or secured at least one first down. But it doesn’t separate by quarter, nor does it reward a team for securing four first downs in a drive versus one then punting the ball away. Thus, it too is flawed for these purposes.
Basically this was an incredibly roundabout way of saying that there is no great stat to determining offensive performance in close games, but we were pretty good at it in 2019. Long-term it’s not something we should necessarily be worried about, but it is something that we can improve. After all, the dream is to have close-game stats that resemble the Belichick-Brady outlier percentages as that’s what can vault a team from “perennial contender” to a legitimate dynasty.
So let’s look at some…
POTENTIAL PROBLEM AREAS
Jimmy Garoppolo
Just to get it out of the way, let’s start with the culprit that every angry ex-defensive player on ESPN trying to carve out a niche as a sound bite savant will point to first.
*1st-3rd QTR stats are averages per quarter
During the fourth quarter, Garoppolo’s stats across the board are actually better or equal to his averages across the first three quarters, including considerably fewer turnovers and a higher QB rating in the fourth. Now, the decreased YPA and slightly fewer first downs could mean there’s some stalling out come the fourth… but that’s a bit of a stretch all things considered. When you account for the fewer attempts, the first downs / attempts is almost exactly the same, while the YPA difference is likely in part due to the fact that the Niners predominantly held the lead in fourth quarters this year; more conservative play-calling with fewer deep shots would mean less big gain outliers, which would prevent Jimmy G from boosting his averages.
If that’s not enough to assuage your concerns, think of it this way. In 2019, Patrick Mahomes—the undisputed best QB in football—saw his QB rating drop 15 points and his completion percentage 8% during fourth quarters; Jimmy G’s YPA dip isn’t ideal, but it’s far from evidence that he can’t ball late in games.
Granted, if Jimmy plays better in the last five minutes of the Super Bowl do we win? Yes. Absolutely. But considering him the focal point of our fourth quarter struggles, or, I dunno, rating him the 20th best quarterback in the league, is—at best—short-sighted short-term memory nonsense and a claim that is nearly impossible to make independent of our other issues on offense (which we’ll talk about later).
In the past three years, the Niners are 4-20 without Jimmy G and 21-6 when he’s their starting quarterback. That includes a 9-4 career record in games decided by one score or less and a league-leading four 4th quarter comebacks last season. Garoppolo has a lot of areas where he can improve and improvement should be the expectation, but while he’s at times inconsistent or frustratingly careless with the ball, neither his stats nor his film should suggest that he’s a liability in close games.
Feast-or-Famine Run Game
By practically any measurement, our 2019 run game was a success. While our offensive rushing DVOA finished only 12th in the league, that’s a figure that was greatly dampened by our early season string of offensive injuries. Once Staley, McGlinchey, Juice, and Kittle returned and Mostert started carrying more of the load, we were unquestionably one of the best rushing attacks in the league. In fact, on an expected points added (EPA) per play basis, we were the best offense in football when both Juice and Kittle were on the field.
So why is our run game even being brought up here? I’ve mentioned before that our heavy reliance on zone and—more importantly—stretch run concepts means that our rushing attack is inherently more feast-or-famine than, say, a Harbaugh inside power scheme. We rely on big plays more than small consistent ones. Our stretch concepts and speedy backfield mean we’ve got a much better chance than most at ripping off gains of 10 or more yards, and the horizontal/lateral pressure that we put on defenses because of that opens up the deadly array of reverses, misdirections, and play action passes that we ravaged teams with all year. But it also means, that on a play-to-play basis, we might not be as consistent. That shows in our advanced OL statistics below.
Adjusted Line Yards is the overall ranking, which tries to gauge how a team—in a world where they all had the same league average running back—sets up their running game for success. The Niners’ 8th-best ranking makes sense.
The other two greens are explosive run rate stats, which gauge how often an offense creates gains of 5-10 yards (second-level) or 11+ yards (open field). As expected, our penchant for explosive runs puts us near the top of the league in both of those categories.
But then we get to Power Success, which measures the success of teams when they’re within two-yards from the goal line and/or require two-or-less yards to convert a third or fourth down. Here we’re barely better than average. And when it comes to Stuff Rate, which measures the percentage of runs where a running back is tackled at or behind the line of scrimmage, we’re a below average performer.
Now is this simply the way it is? A give-and-take that we concede in order to be able to witness the sight of Matt Breida scampering at a league-best 22.05mph while he bolts past the Cleveland Browns to open Monday Night Football? Not necessarily. Most teams are zone-leaning now, and—while few run stretch as religiously as we do—there are ways to increase down-to-down consistency without throwing away our big-play ability. Which we’ll get to later.
We are Not a “Roll the Ball Out” Offense
In true West Coast offense-style, we are a very structured and methodical offense. Shanahan’s notoriously long play-calls include checks, sight adjustments, shifts, motion, and—particularly at the beginning of drives—secondary calls baked in if we see an unfavorable defensive front. It’s this complexity and attention to detail that makes our offense so dangerous at all times and so difficult to prepare for. But that also mean’s it’s not a “roll the ball out” offense. In fact, it’s probably the exact opposite of that.
What is a “roll the ball out” offense? It’s a scheme that does its best work out of the gun in the two minute drill, where the concepts and plays are minimized and route and technique adjustments reign supreme; where improvisation and scramble drills are king and defenses’ inability to make checks or substitutions means they’re playing more backyard ball in simple coverages; where dinking-and-dunking, going deep, and then hitting comebacks becomes a broken record. Basically, it’s the second half of every Seahawks passing play and the fourth quarter of all their games (when they are inherently behind and they take the reins off Russell Wilson). These offenses that thrive on improvisation, the vertical game, and passes outside the hashes are inherently more comfortable operating in no huddle situations—with fewer shifts, motions, and pre-snap adjustments. In some cases, it’s preferred because the QBs have less to process and can get in the flow of the game without all those complexities.
We are not that offense, and while that doesn’t mean we necessarily struggle in come-from-behind situations, it does mean that our five minute offense and our two minute offense look starkly different. Without a regular outside deep ball threat, teams can collapse down on us, press outside, and send the house, and—when we don’t have the time to run our full playbook—that means we have to drop the methodical flow of our gameplay, which often means the elimination of many of the shifts, motions, and other pre-snap adjustments that makes our attack so dangerous.
Improvisation will always play a role in football, but less of it isn’t necessarily a bad thing (see: the Patriots Dynasty, Bill Walsh’s entire philosophy). After all, Deshaun Watson has to improvise so much because the Texans can’t be troubled with making a coherent game plan in back-to-back weeks (or quarters). Aaron Rodgers, one of the best “roll the ball out” players of all time, often had to showcase his improvisational skills because Mike McCarthy spent the better part of a decade alternating between calling Da Bomb and Sub Zero. A play that’s called open is gonna succeed more often than a play that needs to go off script to get there, but without our whole bag of tricks, our offense can lack spontaneity at times. That can show in performance, where our YPA and big gains shrink late in games, as well as in play-calling, where Shanahan can lean conservative — potentially to offset the threat of exposing problem areas. Which we’ll get to in a second.
Potential Solutions/Improvements
Receiver Maturation:
This chart shows raw EPA (expected points added) lost due to drops last year. As you can see, Jimmy G was near the top of the league. It’s important to note that this is NOT a rate stat, so it is NOT an average of EPA lost per attempted pass. So guys like Dak Prescott or Carson Wentz, who both finished in the top 6 in attempts last season, were considerably less affected by drops on a per-pass basis than guys like Josh Allen or Garoppolo, who both had about 130 fewer attempts on the season.
This chart also assumes that the receiver who dropped the ball would have made zero yards after the catch, which, considering how our offense is built around creating chunk plays and how we’re arguably the top YAC team in the NFL, likely underrates the impact these drops had on Jimmy’s total EPA. So it’s safe to say, Jimmy G was actually impacted more by drops than any other QB in the league last year.
Obviously, we had a ton of issues in our receiver corps until the mid-season trade for Emmanuel Sanders. Now he’s gone and Deebo Samuel, our defacto No.1, is out until at least the beginning of the season with a Jones fracture in his foot. But there’s still a chance for improvement in this area.
If he returns healthy, I think Deebo is in for a HUGE season. We’ve also added Brandon Aiyuk, who is clearly less polished and accomplished than Sanders but brings a major athletic upside that we lacked last season. Kendrick Bourne and Dante Pettis (who may stikk carve out a role) are a year older and more developed, and we’ve functionally added three new wideouts to the team. Trent Taylor is your standard underneath slot receiver, consistent on third downs and short-to-intermediate gains. While Jalen Hurd and Jauan Jennings provide a size and physicality that we’ve lacked on the outside throughout the Shanahan tenure.
Injuries and rust could mean a slow start to the season, but this receiver corps has the potential to be better and much more versatile than last year’s by season’s end. Adding more guys who can get open late in games and catch the ball in tight coverage is clearly a benefit when it comes to closing out games.
Slot receivers and pass-catching backs
No knock on Emmo, but he wasn’t really a true slot receiver. He played a bunch of snaps there for us, but was more of an outside guy who could lineup anywhere rather than the dude you always wanted inside. Third downs are when traditional slot receivers like Taylor—who can get open quick underneath—and physical “big slots” like Hurd and Jennings—who can fight off and wall off physical man coverage—make their money. But it’s also when a dynamic pass-catching back really comes in handy.
Fool me once, shame on me. Fool me twice... still shame on me. But fool me three times…? Jet McKinnon’s recovery has been much better this year, he’s looked great in higher intensity off-season workouts, and this season—much more than in the past two—it actually looks like he could see the field and contribute for the Niners. If that’s the case, his greatest contribution will be as a pass-catcher, and the time we’ll need him most is on third downs. I love Mostert, but last year our best matchup threat out of the backfield was our fullback. That will no longer be the case if McKinnon’s knee is up to speed.
Give it to Mostert?
Speaking of Mostert, when it comes to upping our short-yardage running game, could it be this simple? Well, no, but it’s still worth looking at this chart from David Lombardi at The Athletic:
Last year Mostert led all RBs in YPC, his DVOA also topped the NFL by a very wide margin, his elusiveness rating was 3rd best amongst RBs, and his success rate (the closest thing we have for a measure of how often he won’t get stuffed) was ninth in the league. While Tevin Coleman’s bigger size, better durability, and reputation would make it seem like he’s the Niners’ short-yardage back of choice, Mostert actually outperformed him, the rest of the Niners’ backfield, and the rest of the league in nearly all situations.
Now like all stats, these should be taken with a grain of salt. Despite adjusting for defense and trying to eliminate as many outside variables as possible, no stat can put these backs completely in a vacuum, and—to Coleman’s credit—he faced boxes of 8+ defenders on a league-leading 40.2% of his runs—nearly 8% more often than Mostert—so his efficiency was bound to take a hit. That being said, it does seem like Mostert—who just got his contract reworked and has been beefing up in the off-season to shoulder a greater load—should be getting a lion’s share of the carries, whether it’s to bleed the clock, convert short-yardage gains, or otherwise. And if he can repeat last year’s performance, that’ll be good news for us.
Get Better on the OL
To me, addressing the OL is clearly the most important step towards getting better at icing games. I think we may be grossly underestimating how poor our interior pass protection was by the end of the season and how much that affected both Garoppolo’s play and Shanahan’s play-calling.
It’s easy to say that Shanahan simply didn’t trust Garoppolo in the playoffs, and—to some extent—I’m sure there’s some validity to that. But if you’re getting interior pressure, you simply can’t call a lot of drop back passes, and with the run-game humming the way that it was down the stretch, our heavy reliance on the ground-and-pound almost made our downfield limitations even more prominent.
According to ESPN’s pass-block win rate, which measures how often a line holds its blocks for over 2.5 seconds, the Niners OL was 26th in the league last year. And while it would be easy to point to the 13 combined games where the Niners were missing a starting tackle (or two) for negatively influencing that stat, our interior OL was the biggest problem.
While Pro Football Focus is far from perfect, it is still the only outfit that rates every single snap of every single offensive lineman and since I don’t hate myself nearly enough to do that, that means it’s still the best reference we have for determining individual OL effectiveness. According to PFF, out of 65 qualifying guards, LG Laken Tomlinson finished 38th in the league in pass block efficiency while Person finished 53rd. I think Tomlinson is actually a tad underrated by that figure, but Person’s grade… yeah, not great.
Despite a small sample size, I’m super bullish on Daniel Brunskill. He stepped in mid-game at tackle against one of the fiercest pass rushes in the league and didn’t miss a beat, started games at left and right tackle and during that—albeit short—run, ranked 18th in run-blocking and 4th in pass blocking efficiency in the entire NFL... amongst tackles (!). While his numbers at guard were actually way worse—largely due to nightmare matchup against Aaron Donald in his first ever start on the interior—he played really well against Seattle and should be considered the front-runner to replace Person as the starter.
Issues to start the game against the Rams and struggles down the stretch against the Falcons were largely the result of major lapses along the interior line, and I don’t have to go into how often Chris Jones had this kind of impact in the Super Bowl. We’ve already added Trent Williams, who—if he’s the same player he was in Washington—will be a major addition, particularly in the run game, but shoring up the right guard position and keeping Richburg healthy at center could mean massive strides in terms of pass protection and our short-yardage run game. That could mean gains in the consistency of our short-yardage runs, as well as the confidence required to dial up more drop back passes, vertical shots, and throws outside the numbers.
Increased Aggressiveness (especially early in games)
There were times last year where it seemed like we relied too heavily on the idea that our defense would smother the opposition, punting balls, kicking field goals, and playing conservative in the first three quarters when we could have been going for knockout blows. While I’d argue the analytics push in football is a bit like the constant commentator references to RPOs (importance overexaggerated and mentioned way too often), I do tend to agree with the overall concepts of its fourth down decision-making charts: in essence, go for it more often. While I would NEVER want to go full baseball manager and be making fourth down decisions based on one or two percentage points in a binder, I think in general, football coaches are too conservative when it comes to fourth-and-short inside enemy territory. As stated above, the best way to beat teams in the NFL is to put them away before the fourth quarter, and if there was an area where the Niners struggled at times last year, it was in letting teams hang around after vastly outplaying them in the first three quarters. Maybe that’s a fourth quarter problem, or maybe we ought to approach the first three a bit differently.
For example, let’s look at two pillars of (well-executed) aggressiveness: The 2019 Ravens had a good defense, an incredible run game, and they regularly went for it on fourth down. Because while analytic-types LOVE to knock the run-game, short yardage conversions are still the one area where the run game and full backs and all that stuff that “modern day” offensive minds hate are proven to be super effective. The Ravens’ offense was insanely efficient because their run game kept them ahead of the sticks, put them in a position where fourth downs were short and attainable (thus maximizing DSR), got people wide open in their passing attack (thus inflating passing efficiency), and minimized defensive possessions. And when they didn’t convert those fourth downs, they still gave their defense good field position, which—as stated above—is the greatest indicator of offensive success. Where they expectedly fell flat in the playoffs? They had no short-to-intermediate passing game. Running the ball and throwing deep off play action can only work if you can run the ball; if someone loads up against the run then both strengths are nullified. Offenses that are only built off one thing always run into a defense that stops them, and that’s what happened against the Titans last year.
So for a longer track record that doesn’t require having someone with the athleticism and moves of a punt returner at QB, let’s look at the Patriots (although not last year’s Patriots, lol). New England’s offense is built on a shotgun-heavy short-to-intermediate passing game and a heavy-set gap-running scheme; they are literally built to convert 3rd-and-5’s and 4th-and-1’s. It’s no wonder Belichick has always been at the forefront of fourth down aggressiveness. Because like the Ravens, he understands the importance of maximizing possessions.
The good news? It’s not hard to see how the Niners’ offense can be just as good at converting third and fourth downs as the two teams above. Possibly even more so. We have a strong run game and nasty play action attack. Jimmy G is at his best with short-to-intermediate passes and all our weapons are built for YAC. We are quickly assembling a roster full of dudes who block, run, catch, and present massive mismatches all over the field. What linebackers can cover Juice and Kittle? What cornerbacks and safeties can tackle Deebo (and hopefully Aiyuk) in the open field? What the hell do you do with Hurd/Jennings lining up in the slot on third-and-short? If we shore up our short yardage run game and keep Jimmy G as clean as Brady was in his time in New England, there’s no reason to think we can’t be just as effective closing out games… if these improvements come with a boost in aggressiveness.
Here’s Football Outsiders’ Aggressiveness Index scores for last year, which include “Aggressiveness Index,” a measure that compares each coach’s chance of going for it on fourth down versus the league average, and “go for it rate,” which is simply the % of fourth downs the coach attempted to convert.
For reference, the league average AI figure was 1.46 this year
Unsurprisingly, John Harbaugh set a new record for AI and go for it rate in 2019, while Shanahan—ranked 28th this year and 31st last year—was near the bottom of the league in AI. And yes, while his go for it rate was 6th-highest in the NFL, he also had by far the most fourth-and-short situations where the average coach (over time, not even in today’s more aggressive NFL) would have gone for it.
Is being more aggressive always better? Obviously not, and it’s worth noting that Belichick (17th) actually landed near middle of the pack in 2019 (in part likely due to the fact that his offense sucked and his defense was amazing), Andy Reid was only one spot ahead of Shanahan at 27th, and the mean average rank of every playoff qualifier was 16.08, or exactly the middle of the pack in terms of aggressiveness. Going for it more on fourth down won’t overcome issues on first through third, nor will it assure success in the fourth quarter, but—as stated at length above—putting teams out of their misery early is the best way to sustain success in the NFL, and if a lean more aggressive in the first three quarters could mean fewer close games, that’s clearly a step we should take.
Last year, I can understand the Niners’ conservative lean. We had a quarterback coming off an ACL tear, in his first full year as an NFL starter, who is prone to turnovers at seemingly random times. We had a receiver corps that—other than Emmo—didn’t have a single player operating outside of their rookie contract (this year, Bourne is the only player not on his rookie deal and he’s on an RFA so it’s pretty much the same thing). We had an offensive line that excelled in the run game but had a tendency to give up interior pressure—especially on clear passing downs—that could get in the face of your pocket passing quarterback (which is one of the last things you wanna see coming off a knee injury). And we had a defense that was—for certain parts of the season—absolutely unreal.
But this year? With an improved offensive line? With Jimmy G more seasoned and confident? With Mostert (hopefully) taking a greater share of the team’s carries? This year, a more aggressive lean should be expected, in no small part because it should be much more feasible. That may help us close out games in the fourth quarter, but—perhaps more importantly—could help us finish teams off before it even gets to that point.