🏈 Where is the space in today's NFL?
My Week in Sport(s): the state of QB play, Ellyse Perry and Connor Bedard
Welcome to My Week in Sport(s) — a regular newsletter from Plot the Ball.
In this edition:
🏈 Understanding how NFL teams progress the ball in 2024
🏏 Ellyse Perry’s best WBBL season yet
🏒 Connor Bedard’s creative offensive-zone passing
🏈 Where is the space in today's NFL?
I’m a bit sceptical of lines of argument that begin with the stock phrase ‘stats don’t tell the whole story’.
Really, as I’ve written before elsewhere, it’s individual moments in games which don’t tell the whole story.
A certain on-field action going from happening three times a game on average to just once or twice can be pretty impossible to pick up by ‘watching the games’; for some sporting trends, you’ve just got to look at the numbers.
This is true even for play at one of the most closely observed positions in all of professional sport: quarterback in the NFL.
QB play is the prism through which I watch — and try and understand — the league, and I’ve been following some of the position’s biggest names more closely since I wrote back in late 2022 about how modern dual-threat QBs were more explosive than their predecessors1.
If I was updating that piece now, though, I might have to reconsider one of its premises.
When it comes to running the ball, the analysis is still directionally correct.
In a piece published by ESPN prior to the 2024 season, Bill Barnwell confirmed two crucial points: quarterbacks are rushing more frequently — and are still more efficient when they do so than running backs2.
NFL teams, however, seem to be getting steadily less effective at generating explosive passing plays.
In an excellent recent article for Sportico, Lev Akabas noted that — through Week 5 of the current season — quarterbacks were completing fewer passes of 40 yards or more per game than they have in almost two decades3.
In this week’s newsletter, I wanted to take on my piece from a couple of years ago — and see if I could build on the work of Barnwell and Akabas too.
Can we be even more precise about identifying where space exists on the field in today’s NFL, and how this has changed in recent seasons?
Via Pro Football Reference, advanced data is easily accessible for the NFL from 2018 onwards.
For the passing game, yards are broken down into those gained before the ball is caught by a receiver — known as ‘air yards’ — and those gained after the catch. Rushing yards are split up too, into those gained before the runner enters into contact with a defender and those gained after contact.
In basic terms, these are the four ways that teams can progress the ball upfield — and this comprehensive data set allows us to evaluate how offensive approaches have evolved league-wide over the last seven seasons.
Put another way, it can provide us with insights that watching the games and recalling memorable moments simply can’t compete with.
On average, each NFL team has run around 60 offensive plays per game over this period — so the statistics that follow are presented as ‘yards per 60 team plays’.
This accounts for changing frequency — that is, teams choosing to pass or run the ball more or less over time — as well as the changing effectiveness of passing and rushing.
By plotting the yards gained in each of these four phases of ball progression, a couple of trends emerge: ‘air yards’ — or pre-catch passing yards — are contributing less to teams’ offensive success, and pre-contact rushing yards are contributing more.
So far in the 2024 NFL season, teams have been gaining around 15 fewer air yards per 60 offensive plays in 2024 than they were in 2018 — and 10 more yards before contact when they run the ball.
In relative terms: about 59% of all rushing yards have been gained before contact so far in 2024, up from 55% in 2018; about 51% of passing yards have been gained through the air so far in 2024, down from 53% in 2018.
By combining this analysis with some insights from the league’s proprietary location- and motion-tracking data, we can tell a pretty neat story about the evolution of space in the NFL.
Offensive schemes and choices don’t exist in a vacuum, and a piece on the NFL’s own infrequent analytics blog makes it clear that defensive teams are making different choices than they were a few seasons ago.
Analysis of Next Gen Stats shows that teams have been stationing defenders near the line of scrimmage much less frequently in 2024 than they were in 2019 — on both the interior line, and towards the outside of the field.
Where are these defenders going instead? Teams are taking deeper passing attempts off the table by using “more two-high safety looks”, and filling intermediate areas “with linebackers or hybrid defensive backs positioned 4-6 yards off the line”.
Linking this together with our data: fewer passing yards through the air are an obvious consequence of those more frequent two-safety schemes, and fewer bodies near the line of scrimmage lead naturally to more yards being available for rushers before they initiate contact.
This generation of mobile quarterbacks, then, have a much more favourable environment to run the ball in than their predecessors did — but constraints on their passing options are the price they have to pay.
And they’re increasingly coming to terms with those constraints: completed passes so far this year are travelling only 5.6 yards in the air, and incomplete ones only 11.3.
Both figures are down from the 2018 season.
It’s maybe not surprising, then, that pass attempts have dropped off as a share of offensive plays this season — despite, as Akabas highlighted in his piece, completion percentages continue to trend upward.
In every sport, we can assume that the balance between attack and defence is constantly changing like it has been in the NFL.
We need statistics, though, to be able to tell the full story.
🏏 Run the Numbers
During the 2024 edition of the WBBL, Australian superstar Ellyse Perry has been scoring runs faster than ever before.
But, as we observed recently, batting in women’s T20 cricket is changing. Is Perry simply evolving in lockstep with her peers? Or does this remain her best-ever year even when you adjust it for the league’s current scoring environment?
By adapting a version of the ‘runs vs. expectation’ metric I’m fond of for the T20 format, it’s clear that last season and this one have been the all-rounder’s most efficient batting years in the WBBL.
Perry added 6.6 runs per innings above expectation — calculated by comparing her strike rate to the strike rate of other batters in the match — in 2023, and has recorded a mark of +8.8 runs per innings so far in 20244.
When I looked at Australia’s best T20 batters in depth a couple of years ago, Perry 1.0’s method was distinct from those of her international teammates.
At that point in her career, she was facing more balls per innings than her peers — but scoring considerably more slowly than the other batters in her games5.
Since then, she’s flipped the switch.
Perry 2.0 is a different player entirely: she faces fewer balls per innings, but contributes more with the bat than she ever has before. Across the 2024 season so far, her innings have provided the Sydney Sixers with 86.4 runs above expectation in aggregate — comfortably outpacing even her historic 2018-19 season.
🏒 Watch the Games
While his team are performing much better with him on the ice in his second year in the NHL6, it’s fair to say that Connor Bedard isn’t yet reaching the high expectations that have been set for him at the offensive end of the rink.
At even strength, his point production is actually down from his rookie season: from 2.2 points per 60 minutes played in 2023-24, to just 1.6 per 60 this year.
That drop is almost entirely due to Bedard converting fewer7 of his own shots on goal; in the meantime, he’s still contributing by setting up good shots8 for his Blackhawks teammates.
Bedard had a pair of beautiful primary assists against the Anaheim Ducks earlier in the week, and the second of them was a great example of his ability to access the most dangerous parts of the rink in creative ways.
After a brief period of contested possession early in the second period, the puck is swept towards Bedard — who is stationed behind the left goal post.
The forward receives it on the turn, and drags the attention of attention of every Duck on the ice towards the far corner of the rink as he weighs up his next move. After a couple of seconds, though, the man who should be defending the slot begins to rotate to face one of the other Blackhawks below the goal line.
That’s the only cue Bedard needs.
A quick glance confirms that the centre of the ice will shortly be vacant. With his momentum continuing to take him towards the boards, Bedard slips the puck towards the slot.
As he releases the pass, Blackhawks wing Jason Dickinson is arriving late to the front of the net — and has just enough space to bury a one-timer that puts his team ahead.
You can watch a clip of this sequence here.
The next edition of My Week in Sport(s) will be published in two weeks, on Saturday December 7th.
She has faced a similar number of balls per innings in each season — 29.0 in 2023, and 30.6 so far in 2024 — and her efficiency on a per-ball basis is slightly higher this year (+0.29 runs) than last (+0.23).
Through the end of the 2022 WBBL, she was facing 33.4 balls per innings on average and her scoring efficiency was -0.05 runs per ball compared to expectation.
From a piece by Mark Lazerus for The Athletic last week: “He was minus-30 at five-on-five last season; he’s even this season. His expected goals percentage is up, his shot-share is up, he’s cut his goals-against per 60 minutes nearly in half. Opponents are getting fewer shot attempts, fewer scoring chances, fewer high-danger chances.”
His unusually low shooting percentage of just 5.5% — which has almost halved since last year — is likely to rectify itself as the season goes on.
Bedard is averaging 1.1 assists per 60 at even strength through Thursday’s game against the Panthers; he’s provided the final pass for about four in five of his assists in 2024-25, compared to a rate of around two in three last season.
The second problem is a lack of courage. Perhaps the most over used fallacy of the past few years is the so called “two deep coverages” as some sort of mystical supernatural elixir to deep passing. You know, the magical “scheme” that got Brandon Staley a head coaching gig. Nevermind that two deep safeties have existed for decades as the passing the game exploded in the NFL. Nope, now all you must do is just put two guys in those spots and your all set.
What’s in fact set is the passivity of offensive play callers. If all a defense must do is position two players in particular spots on the field and the offense immediately eliminates half of its playbook, well yes, that “works” in a twisted sense. It works not because of the defense, but because offenses are voluntarily not challenging those defenses deep. Again, the opportunities are there, but individuals are just not taking them. They are surrendering to them without a fight. Quarterbacks are now instructed to avoid deep passes whenever they see this formation as some sort of complex “read.”
Is it any wonder that deep passes have fallen?
The complete intellectual bankruptcy of this approach is intuitively obvious. When you fail to challenge these defenses deep you are in essence treating the two safeties on each team as equally talented and stellar – so much so, that you don’t challenge any of them - ever. Yet, you know that the variance in safety talent is enormous throughout the League. Why then not test the ones that you believe aren’t as talented? Why play into what the defense wants you to do without a fight?
I’m not sure I have a definitive answer, but I do have a hunch. Offensive coordinators and head coaches are now obsessed with avoiding interceptions. An INT is now the ultimate black mark on an offense, a QB, an OC, or a HC. It is also the stat that the media myopically focuses on. But should it be viewed this way?
Are fumbles not as bad? If they are, why then do I read where Trevor Lawrence had a “terrific game with no interceptions” only to see in the box score that he also lost two fumbles. If those fumbles were instead INTs, how would the narrative change? You know that answer.
Or what about punts? By any definition, a punt is a turnover – a surrender of possession during downs - and yet, we don’t think of it that way. We somehow think of it is “smart football.” We say a team lost possessions “on downs” or that a deep interception is “as good as a punt” – but say nothing of the negative consequences of offensive play calling that yields more punts. When you combine punts with the turnover category, you get a much clearer look at the efficiency of the offense.
The canary in the coal mine is the wide receiver screen. These plays I believe are consistently negative from an EPA perspective. They never yield yardage and often lose yards. Yet, teams keep calling this play over and over each week. Why, when they surely must know that analytically they are a disaster. I suspect because it gives the offense the illusion of “passing” the ball with an extremely negligible risk of an INT. It’s essentially an extended handoff – safe – that keeps the management away from criticism. It increases the number of passes thrown while keeping INTs low.
Coaches ignorantly ascribing “two deep coverages” as the reason why they never throw deep are cut from the same cloth. They get to “blame” the market for their own incompetent risk adverse investment decisions to not take the opportunities that are right in front of them. You shouldn't buy into that.
Thanks again for the interesting read.
Great article. I will offer two counterpoints to the NFL analysis for consideration and to further this important discussion.
Your point around the error of looking at rushing as a single category instead of two distinct categories - designed QB runs and traditional RB runs - is spot on. There is an erroneous notion that has recently become near orthodoxy from the media that we need more of the traditional running game in the mix (handing the ball off to a RB 5 yards behind the line of scrimmage) because of the trend toward lighter boxes and the supposed increase in "running game" efficiency.
RB runs remain wildly inefficient despite alloyed trend charts with overly broad categorizations labelled "Rushing Yards." What teams in fact need, is a higher percentage of runs through the QB position and away from drive killing RB handoffs that keep happening week after week along with the resulting punts and having to go for it on 4th down.
Where we will intellectually have to part ways, however, is your implication that you don't need to watch the games to understand what is happening from a historical perspective. That everything can be gleaned from more analytics, a better calculator, and the right spreadsheet. I could not disagree more. You need all of it.
Let's start with your statement that is ironically, the crux of the problem:
"...[T]his generation of mobile quarterbacks, then, have a much more favourable environment to run the ball in than their predecessors did — but constraints on their passing options are the price they have to pay."
It is not the price they "have" to pay - but it is the price they are in fact paying. Those are two quite different ideas that are being confused here. There are no "constraints on their passing options" as asserted - those options are still there but are simply not being exploited as in the past. Like an investor who never invests in Tech stocks lamenting how they missed a Tech stock run-up, the investment options were there for the taking...they just didn't take them. The constraints were not “imposed” by the market or the game...they were arbitrarily imposed by you on yourself. That’s a distinction with a difference.
This failure to exploit available opportunities is the largest hazard of the ascending dual threat quarterback phenomena. This problem can be further broken into two distinct parts. First, the rise of dual threat quarterbacks has lowered the quality of pocket passing, and passing has, not surprisingly, suffered as a result. These QBs are now valued far more for their "athleticism" - their running ability - than their passing acumen. This was not the case only a few years ago, when quarterbacks were expected to pass from the pocket and runs more rarely utilized. Increasingly, as these dual threat QBs enter the League you are seeing more outstanding runs, but less skillful pocket passing which equates to less deep downfield throws from the pocket.
I want to be clear that this evolution might ultimately be the superior offensive strategy where you combine the greater efficiency of more quarterback runs while sacrificing some measure of downfield passing. However, downfield passing as a stand-alone strategy will undoubtedly suffer when examined.
If you watch the games, you see this increasing problem every week. These young quarterbacks, groomed in simplistic Air Raid offenses and out of Gun formations for years - do not possess the skill of "stepping up" or "climbing" the pocket to gain time as the pressure moves around them to deliver deeper strikes downfield on timing routes. Instead, at the first whiff of pressure around them, they spin out of the pocket and begin a disastrous rollout.
Roll-outs are the ultimate deception that are spreading like wildfire throughout the NFL. To the QB they "feel" like they are buying more time, but the play is over, and the results are almost always predetermined and unsuccessful. A rollout immediately does three catastrophic things for the passing game - it cuts the field in half (toward the rollout), eliminates any receiver occupying the other half, and ruins all timing routes. The resulting play is now purely ad hoc chaos.
As the QB continues to rollout out, the available half field for a downfield completion is now a rapidly shrinking rectangle with each step the QB takes toward the sidelines while being flooded by fast defenders who are filling that shrinking rectangle that has fewer receivers to target.
Ultimately, only four outcomes are typically available - a sack, a throw out of bounds, an interception made into the shrinking rectangle or a short QB scramble toward the sidelines - and none of them are great in term of air yard stats. Yet, if you review the play just before the rollout, often there are 2-4 yards in front of the QB to step into - the "pocket" - that they immediately surrender for a rollout. The downfield pass was right there for the taking, but they ignored it to take off.
At a psychological level, the problem is how these quarterbacks have been socialized as to what constitutes positional success in high school and college. Over these years, these quarterbacks have operated out of the Gun and have been lauded each time they "take off" and scramble for large gains. It’s what they think makes them special. The defensive players at these lower levels are not as fast or as smart as in the NFL, making these big plays routine. But the NFL is another can of fish entirely.
You saw a similar realization with "Time-to-Throw" stats and college QBs sitting behind massive offensive lines, such as Justin Fields at OSU. They have been coached to wait until the receivers are high school wide open and then gun the ball to them...all as the OL makes that extra time more than available. Unfortunately, waiting that long in the NFL equates to sacks. This in turn drives more rollouts and scrambles as the passing game suffers.
Time to Throw - once universally viewed as a stat for offensive lines – was soon to suffer an “aha” moment - where modern thinking now appreciates that sacks are often a negative result of a quarterback who holds onto the ball too long waiting for a receiver to get open instead of throwing them open on timing routes. It’s now a quarterback stat. These dual threat quarterbacks will have to develop, or they will fail at the NFL level.
It’s why Justin Fields sits on the bench, and Lamar Jackson does not. For Fields, the media continues to fawn over his scrambles while his downfield passing and sack numbers remain poor. Jackson, on the other hand, has steadily improved as a skilled pocket passer who regularly climbs the pocket to find his receivers downfield. Jackson has learned to embrace pocket excellence that is aided with the real threat of his runs – but he is an outlier among the dual threat cadre.
Ironically, the new example for the rollout obsessed quarterback is once again in Chicago - Caleb Williams. Socialized for years in high school and college to believe he can run to success whenever he chooses, he now finds the sledding much tougher. Watch him on any pass play and you immediately see a guy with "happy feet" needlessly moving around in the pocket nervously and looking for an excuse to roll out and take off like at USC. His passing numbers are poor.
I'm not the first person to notice this alarming diminution in pocket passing skills. Tom Brady noted it earlier this year. However, the media mindlessly attacked him because it didn't fit their narrative that dual threat QBs are just so dope with no downsides. It’s just too easy to trot out the tired and intellectually lazy “old guy in the lawn chair” tropes rather than rolling up your sleeves and analyzing the problem.
The early key is whether the dual threat QB coming into the League is "run first" or "pass first" in their approach. Watching Jayden Daniels in Washington, this difference becomes immediately clear. Daniels is calm in the pocket, with his feet set and his eyes downfield. As the pressure increases, he steps up in the pocket - often delivering a deep strike downfield. If that does not yield results, only then does he rollout and shows his athleticism as a runner.
The difference between Daniels and Williams could not be starker. One has advanced pocket passing skills - the other does not. The path to success for one is much shorter – while much longer for the other.