Without a Bears game this Sunday to keep my busy, I decided to dig into a topic I've been interested in for a while. The question of how much impact Run Defense, Pass Defense, Rushing Offense and Passing Offense have on scoring and winning.
I think it's an especially relevant question for us as Bears fans right now. We've got a roster that, to put it charitably, needs to some work. It's unlikely that we're going to be able to address every area of the team in need of upgrades during the 2023 offseason.
So when faced with the need to prioritize, what's most important? Right now, I'd say our two biggest weaknesses as a team are our Run Defense and our Passing Offense. It's fair to say we're in the conversation for among the worst in the league in both areas. Meanwhile, we've run the ball well this season, and are maybe middle of the pack when it comes to passing defense.
If all facets of offensive and defensive play are equal in impact, we probably should use most of our resources this off-season on the run defense and passing offense. But what if one dimension is significantly more important than others? That has to impact our decision making too.
I decided to take a look at the entire NFL over the past 5 seasons (2017 - 2021), breaking down each team's seasons by Rush Defense, Pass Defense, Rush Offense and Pass Offense. Taking a look at Points Scored/Conceded, Winning %, and also playoff success.
Raw Yardage vs. Yardage Rate
To begin, I had to choose which stats to use to represent how good a team is at these aspects of offense and defense. The first thought might naturally be to simply look at rush/pass yardage gained and conceded. But I think there's a major problem with that method.
The issue is this: The score of a game has a large impact on the frequency with which a team runs or passes. When a team is winning, they call more running plays. When a team is losing, they call more passing plays.
This means that a team that goes 4-12 is going to end up with more yards conceded on defense than a team that goes 12-4. Even if that 4-12 team has a really stout D-Line, they're going to face so many more late-game rushing attempts that it's pretty likely that their raw yardage total will look bad. Meanwhile, that 12-4 team might have a porous D-Line, but by being in the lead all game, they see fewer rushing attempts against them by their trailing opponents, depressing total yards against.
I think every football fan understands this concept intuitively, but I didn't want to leave it un-demonstrated. So I grabbed a play-by-play database of all NFL games between 2009-2018. I wrote a script that tallied every rushing and passing play as well as the score at the time of the plays. Here's the breakdown.
As you can see, the difference is pretty dramatic, especially during the 4th quarter.
I'll put some simple numbers behind it to illustrate a bit. Let's say the average offense runs 64 plays per game (which is roughly true). Applying the W/L percentages, a losing team is going to see around 10 more rushes over the course of those 64 plays than a winning team.
Just looking at the 4th quarter by itself, which we'll say contains 16 plays (1/4th of 64), a losing team will see around 6 more rushes than a winning team.
Obviously, you can apply all this in reverse when looking at raw passing yardage. Winning teams get run on less. Losing teams get passed on more. Raw yardage totals aren't the best method to use for evaluating how good a defense is against the run/pass, nor for evaluating how good an offense is at running/passing.
Instead, I'm going to use rate stats. For rushing, the stat is Yards per Carry (YPC). The number of yards a team gained/conceded divided by the number of attempts.
For passing, the stat is Net Yardage / Attempt (NY/A). Net yardage means passing yards minus yards lost to sacks. And then divided by passing attempts + sacks.
Basically, these stats say: On a given play, how effective is an offense at rushing/passing? Or, on a given play, how effective is a defense at stopping the run/pass?
Offense and Defense vs. Points and Win Percentage
The first thing I did was gather the NY/A and YPC numbers for both offense and defense for every team/season from 2017 to 2021.
I then plotted those numbers individually against that team's points either conceded or scored. (Note that I'm using true points scored/conceded, which strips out special teams and defensive scores/safeties)
The first thing that jumps out is how much more impact passing offense has on points scored compared with rushing offense, and how much more impact passing defense has on points conceded compared with rushing defense.
In all cases, there's a correlation. Broadly speaking, the better your rushing offense is, the more points you'll score, and same goes for rushing defense. But passing offense/defense are about 5 times more strongly connected with points.
I also set up four similar charts, but this time instead of points scored/conceded, I put in the winning percentage of the team.
In all cases, the relationship is going to be weaker. Which makes sense, because points scored or conceded are an offense or defense specific stat, as are YPC and NYA for each side of the ball. While win % is impacted by both sides of the ball.
What's interesting to me is that plotting NY/A for offense and defense against wins, you can see a pretty noticeable trend. With offensive NY/A in particular, there's a pretty clear relationship.
But when it comes to YPC on each side of the ball, there's essentially nothing there. Those scatter charts are more similar to a random scattering of points than they are to anything particularly meaningful.
To me, this suggests that rushing defense - and rushing offense for that matter - are relatively unimportant when it comes to building a winning team.
I know that's probably a statement that's going to instantly aggravate some folks, so I want to be clear - I'm not saying that stopping the run is meaningless. Or that it would be smart to say 'OK, we're going to purposely be the worst team in the league against the run and expect to win'.
The first set of charts showed that rushing defense and points conceded were connected. Just like rushing offense and points scored were connected. And points are connected to wins. So obviously, rushing defense and rushing offense do matter.
But this data seems to say that they don't matter nearly as much as passing offense and/or passing defense.
Another way of looking at this data
I think those charts show something interesting, but maybe a little nuance is lost. So I also broke this data down another way.
5 seasons of 32 teams playing is 160 team-seasons. That nicely breaks down into quintiles of 32 team-seasons. I sorted and divided the data set into five 20% quintiles by Offensive YPC, Defensive YPC, Offensive NY/A, and Defensive NY/A. In other words, the top 32 team-seasons by Offensive YPC are in the top quintile. Then the next best 32 team-seasons are in the second-to-top quintile, and so on. And then I gathered stats on a range of measures of team success.
I'll run through each of the Off/Def Pass/Run quintile breakdowns, starting with passing offense.
Viewed this way, it's extremely clear that passing offense success is linked very strongly with team success. And it's pretty consistent right across the quintiles and levels, from winning % to points to all levels of playoff success.
It's particularly worth noting that 75% of Conference Championship game participants, 90% of SB participants, and 80% off SB winners came from the top 20% of passing offenses. (Lone SB winner from the middle tier - The 2017 Eagles).
You haven't seen the other 3 tables yet, but trust me - They're not going to look this conclusive.
This table says pretty loudly that if you're a bottom 40% passing offense, don't even bother showing up to the playoffs. And if you're not a top 20% passing offense, you're at a massive disadvantage come playoff time.
Right away, it should be pretty clear that something different is going on when it comes to the relationship between rushing offense and team success. Sure, when you look at winning % and point scored, there's a broad trend in favor of good rushing teams, but with win % it's kind of wonky.
And then when you get to playoff performance, everything starts to fall apart. No SB champs and only 1 SB participant (2018 Rams) comes from the top quintile. meanwhile, 3 SB winners and 5 SB participants had below average rushing attacks. There's no real connection between conference championships or division winners and rushing prowess.
However, one thing does appear to be true: Do not completely suck at rushing the ball. The bottom quintile is barren of any playoff success, and few playoff participants. But it appears that even a mediocre rushing attack is enough to sustain a champion.
Passing defense matters quite a bit too. It's not quite as overwhelming as with passing offense, but the trends are pretty strong.
Unlike with passing offense, it does seem a little more possible to make or win a SB with an average or slightly above passing defense. But overall, if you want to make the playoffs and go deep, you're better off with a strong passing defense. Worth noting, though, that the 4th quintile does have a decent number of division winners and conference championship participants. Having a mediocre-to-bad passing defense is a major liability, but it's not a 100% killer like having a bad passing attack is.
But the bottom quintile is just as much of a wasteland as the first two were. So far, one takeway is that if you're bottom 20% at any one of passing/rushing offense or passing defense, you're going absolutely nowhere.
OK, this table's just a mess now. There's not even a great relationship between points conceded and rushing defense. In the sense that it seems equally possible for an average rushing defense to put up a great defensive points-against season as a really strong rushing defense.
And the bottom quintile is full of SB winners, SB participants, division winners and playoff teams. (This actually seems weird enough that I have a separate theory about it that I'll go into at the end)
But let's take these four tables at face value, as indicators of the likelihood of success based on rush/pass offense/defense.
Essentially, if you want to win a SB, it's crucial to be able to pass well. It's also important (although not AS important) to defend the pass well.
Rushing well seems optional, just don't be awful at it. And defending the rush seems like another luxury rather than a necessity.
The Bears in 2023 and Beyond
I think we all know that we're at the beginning of a rebuild here, meaning the next couple of offseasons are of the utmost importance in adding impact playmakers to a roster currently lacking them.
If we had double the salary cap space and double the draft capital, we could address every area of the roster where we could use an upgrade. But that's not realistic. We can only do so much in 2023. And then the same in 2024.
Also worth noting that, while this analysis has split run/pass defense and run/pass offense into separate buckets, most NFL players contribute to both on their side of the ball. A nose tackle or an ILB is primarily involved in rush defense. A CB mostly impacts pass defense. A WR is mostly involved in pass offense. But many players are contributing closer to equally to both.
Taking all this into account, I believe the Bears should have a strong priority on adding offensive players, especially those who can boost our passing offense. Our pass offense is arguably by far the worst area of this team. And it's also by far the most important when it comes to wins and playoff success.
Even if we were to come to the conclusion that our rushing defense is equally bad as our passing offense, the relative importance of passing offense absolutely dwarfs rushing defense. If we currently are underwater at both, it's far, far, far more important to work on the passing offense.
Justin Fields remains an unanswered question. But the rest of the offensive talent around him is in dire need of upgrade no matter whether Fields becomes the guy or not. I feel extremely confident in saying that this team isn't going to win anything of significance as long as our OL and WR/TE groups are near the bottom of the NFL.
While I think our DL or LBs being poor is an impediment to winning, it's an issue that multiple successful teams have recently overcome. While you can't really find any instances of a team with awful passing-oriented offensive talent having similar success.
Again, I'll emphasize - This is not me calling for us to want to be bad as a run defense, or saying no priority at all should be allocated to defense. Having a crappy defense will harm our chances of winning. Most SB participants were at least pretty good at one or both of stopping the run/pass.
But wanting our mediocre defense to become better while our passing offense is a dumpster fire is incredibly misguided.
I think we can see now that there's not going to be a magic OC/HC change that's going to turn our marginal collection of offensive talent into a top unit. If we want to get better as an offense, two things need to happen - We need a top QB, whether that's Fields or someone else. And we need a major upgrade in offensive talent across the board.
This offseason, I would be strongly disappointed if our #1 pick were to be a defender. If the clear 100% obvious BPA at our pick is a defender, then trade the pick to someone who wants to get a surefire stud.
Spending premium assets on defense right now only guarantees that we're one year further from getting to where we want to be. We cannot get ourselves out of this hole without transforming this offense.
That's really the end of the piece, feel free to skip this unless you want to engage in some additional nerdery. I'd love to hear any comments or thoughts you guys have. Also, let me know if you want the data I used to produce this. I saw Josh put a google docs link at the bottom of one of his pieces, and I thought that was cool. I'd be happy to do something similar.
I mentioned during the Defensive YPC quintiles section that I was weirded out by the number of SB winners/participants in the YPC bottom quintile. Basically, a suspiciously high number of really good teams had awful YPC against numbers.
I think it's possible that there are some limitations to using rate stats only in an analysis like this. For example, let's say a team is really explosive and good on offense, and as a result they tend to play high scoring games and will frequently be winning by large margins.
It's possible (though I don't say this positively, I'm not sure) that they might tend to run higher YPC against numbers than normal. Just because at a certain point you're going to let teams run, and not really sweat it if they're picking up 5 yards a carry while down 17 points.
I did plot out all 160 team/seasons for points scored on offense vs YPC against. Checking to see if maybe higher scoring teams tend to give up higher YPCs against. The result was pretty random to my eye. R^2 of 0.0168, fwiw. I'm not a statistician so I'm not sure how much stock to put in R^2 values. I mostly just use them as a numerical representation of the slope of a trendline.
I don't really see a pattern saying this is true. But regardless, I do think YPC and NY/A aren't the entire story. With that being said, I think what I've found here is true to some extent regardless. I can't imagine adding new data is going to massively change these types of results about the relative impact of Run/Pass Offense/Defense.