Author
Topic: Optimal First Down Run/Pass Balance
(Read 183 times)

betarhoalphadelta · « **on:** **Today** at 10:53:26 AM »

So, not singling out BAB here, because his post followed a lot of discussion in multiple threads about UM and UW playcalling...

Quote from: bayareabadger on Today at 10:13:02 AM

No more enduring belief that an infective offense is running too much on early downs and the box is too stacked.

Sometimes it holds up to scrutiny, often not.

But I thought it's worth discussing.

There's a school of thought that says that analytics supports passing more on first down.

Such as: https://blogs.cornell.edu/info2040/2018/09/18/game-theory-for-passing-on-1st-down/

In short, I think the idea is that passing on first down has a higher success rate than rushing on first down.

Success rate [on first down] is described as gaining 50% of the needed yards necessary to get another first down. We'll limit this to only 1st-and-10 scenarios, avoiding both 1st-and-other (due to penalty) and 1st-and-goal...

I think perhaps limiting the metric to "success rate" is misguided. Because more yards, on average, are gained on pass plays than run plays, I think it biases too much against the risk of an unsuccessful play and what that means for the offense on subsequent downs... Especially since we cannot look at a single down on an island; a down is merely one play in, hopefully, a sustained scoring drive.

I think you have to look at a few things:

All 32 teams so far in the 2025 NFL are averaging 5.0 or more yards per completion. 31 of 32 are averaging more than 5.0 yards per pass attempt. So if the team completes a pass on first down for simply their average completion yardage, statistically it will almost always be a "successful" play via success rate.
An unsuccessful passing play, OFTEN, is zero yards (incomplete pass). The team is now facing 2nd and 10. So if the play is unsuccessful due to being incomplete, it puts them well behind the down & distance. If it's unsuccessful due to sack, it's even worse.
The median completion percentage in the NFL is about 65%. So the play is unsuccessful due to incompletion more than 1/3 of the time.
Only 8 teams so far in the 2025 NFL are averaging 5.0 or more ypc. Which means that if a team gets their average rush yardage on 1st down--the down the defense is more likely playing to stop the run--24 teams would not count that as a "successful" play per success rate.
An average but "unsuccessful" running play, however, for 23 of those 24 teams would basically leave them at either 2nd and 6 or 2nd and 7. We all know football and that's a lot less daunting than 2nd and 10.

So I don't necessarily agree with the "all teams need to pass more on 1st down" idea based on success rate. However, I also don't believe that "teams should nearly always run on 1st down" either.

I believe that rushing on 1st down is the better option--but the problem is that you simply can't do it so often that the defense is expecting it, or you're going to face stacked boxes nearly guaranteeing your ypc is below average. You have to pass enough on first down that the defense must respect both.

But once you have the defense respecting the pass, IMHO rushing on first down is a more optimal strategy than passing.

bayareabadger · « **Reply #1 on:** **Today** at 01:12:58 PM »

At this point, I think all the metrics point more to more passing in almost every spot.

I don’t know that people actually want that. Because if passing more doesn’t “work,” people are also pissed. The baseline might just be that people are pissed.

I can tell you that college teams through three weeks run around 57.8 percent on non-passing downs and 35 percent on passing downs (those are second and longs or third and 6 or longer).

I can also say that ineffective offense, if there’s some running has an interesting effect. Even when the run-pass balance is different, and even if the other team keeps two safeties back, our brains default to the idea that they’re running too much, there are too many defenders to stop it and the runs are all focused in a specific spot (the middle), when the first two are pretty variable, and the last is a weird bit of semantics, as “the middle” is both wide and mobile.

Mdot21 · « **Reply #2 on:** **Today** at 03:10:07 PM »

my thoughts are it's always better to pass the ball a little bit more than you run it. pretty much without question. you want some balance- but you want to skew little bit more towards passing than running. and it's probably a great idea to pass on early downs more in big games vs the better opponents in order to throw defenses off- especially if you are a run heavy run first team. that whole breaking tendencies thing and all.

pass/run calls should be dictated based upon on what the defense is showing pre-snap. problem is there are very very few QB's in college capable of commanding the offense and reading a defense on the fly.

QB's like Brady/Manning/Brees made a living off of looking at what the defense was showing and checking in and out of pass/run plays. if defenses were crowding the box with safeties- those guys are checking out of run and into pass- if defenses are playing cover 2 with both safeties back they are running the ball every play down the defenses throats until they run them out of it.

see below:

https://www.youtube.com/watch?v=4RhkMgimx3g

FearlessF · « **Reply #3 on:** **Today** at 03:12:07 PM »

speakin of QB's like Brady/Manning/Brees

depends on your personnel

Mdot21 · « **Reply #4 on:** **Today** at 03:13:53 PM »

Quote from: FearlessF on Today at 03:12:07 PM

speakin of QB's like Brady/Manning/Brees

depends on your personnel

truth of the matter is QB's today aren't given that freedom to control the game anymore.

And how are they expected to be able to do that if they aren't ever given the freedom to do it?

coaches need to trust the QB to call his own plays and have complete freedom to audbile if you ask me.

FearlessF · « **Reply #5 on:** **Today** at 03:17:25 PM »

agreed

also talkin bout the O-coordinator callin plays, personnel determines run/pass balance

if you've got QB's like Brady/Manning/Brees - gonna call more passes

Mdot21 · « **Reply #6 on:** **Today** at 03:21:56 PM »

Quote from: FearlessF on Today at 03:17:25 PM

agreed

also talkin bout the O-coordinator callin plays, personnel determines run/pass balance

if you've got QB's like Brady/Manning/Brees - gonna call more passes

those guys called whatever the defenses gave them basically- which is why they were so god damn good. i've seen games where Brady/Manning would check into run plays 10-15 times in a row just to get defenses out of two deep safety cover 2 looks.

betarhoalphadelta · « **Reply #7 on:** **Today** at 04:11:45 PM »

Quote from: bayareabadger on Today at 01:12:58 PM

At this point, I think all the metrics point more to more passing in almost every spot.

I don’t know that people actually want that. Because if passing more doesn’t “work,” people are also pissed. The baseline might just be that people are pissed.

And this is where I say the analytics/metrics are wrong. Because it's not about "more passing" or about getting A first down. It's about sustaining drives.

The problem is this. Take two teams, one that has a competent rushing and passing offense, facing another that has a competent rushing and passing defense.

If the offense passes on first and second down, they're more likely to convert a first down in those two plays than if they ran on first and second down.
If the offense passes on first and second down, they're more likely to end up in 3rd and 10 after those two plays than if they ran on first and second down. Especially since they're probably sitting and 2nd and 10 after the first [incomplete] pass, causing the defense to change the way they defend because 2nd and 10 is more of a "passing down" than 2nd and 6.

If you only look at the former, it suggests passing is optimal. Because if you're looking purely at "success rate" being tied to getting a first down, passing looks good. But if you're 80 yards from the opposing end zone and you need multiple first downs to score a TD, you need to avoid those 3rd and 10 situations too.

But you need to force the defense into those two deep shell coverages to be able to run effectively. It's about the threat of the pass making the offense more effective overall by forcing the defense to respect it. And you have to pass often enough--and effectively enough--to do that.

On first and 10, a "neutral" game script situation, I think there's more desire to pass often enough that the defense needs to be on their heels. I'm not sure what that number is as a percentage for any given team--and it may be the answer is "more often than almost every team does". But I don't think we should extrapolate that to believe the metrics point to passing more in every spot, at least if you're looking at things like success rate or average yards per play. Because it's about sustaining drives, and passing has higher variability to go along with those higher averages. Variability is more likely to stall your drive IMHO.

medinabuckeye1 · « **Reply #8 on:** **Today** at 05:08:03 PM »

I'm as much of a stats guy as anyone but I think there are two things that tend to get lost in the stats.

First is the consequences of an unsuccessful play. You already covered this somewhat but looking at Ohio State's offense last year:

Howard was 309/423 for 4,010 yards. That is 9.5 per attempt and 13.0 per completion on 73.1% completions.
Judkins had 194 carries for 1,060 yards. That is 5.5 per carry.
Henderson had 144 carries for 1,016 yards. That is 7.1 per carry.

The average completion (13.0 yards) is a first down but even for the Buckeyes with Howard finishing second nationally in completion percentage (among full-time starters), still more than one in four of their pass plays was not completed.

Meanwhile, the two main RBs each averaged a "successful" 1st and 10 play.

Still, unsuccessful plays happen. If you think of it as four consecutive 1st down plays on one drive and adjust from the average a little bit, I think you'll get something like this for passes (best to worst, obviously):

17 yard completion, 1st and 10.
13 yard completion, 1st and 10.
9 yard completion, 2nd and 9.
Incomplete pass, 2nd and 10.

The first three are great but the incomplete pass now puts you in a tough spot.

With running plays you get something more like:

8 yard carry, 2nd and 2.
6 yard carry, 2nd and 4.
4 yard carry, 2nd and 6.
2 yard carry, 2nd and 8.

As a fan, if my team gets 6-8 yards on 1st and 10 I'm feeling really good. If they get 4 I'm nervous but still feeling pretty good. If they get 2, that makes things tougher but you are still a threat to run on 2nd and 8 because an average carry of 5.5 (Judkins) to 7.1 (Henderson) gets you to 3rd and 1-3 which isn't bad.

When you throw incomplete on 1st and 10 it more-or-less makes you one-dimensional because running on 2nd and 10 usually results in 3rd and long.

One thing about stats is that I REALLY wish someone tracked 'median' run. Judkins' and Henderson's averages of 5.5 and 7.1 respectively are great but averages are skewed by outliers. Judkins' long for the year was 86 and eliminating just that ONE run drops his average from 5.5 to 5.0. Eliminating Henderson's 66 yard long drops his average from 7.1 to 6.6. My point is that those guys weren't really running for 5-7 carry-after-carry. They were running for more like 3-4 carry-after-carry and then once in a while breaking a long one for 20+.

The second thing is wearing down a defense. My observation is that running teams get better as the game wears on so long as they stay ahead or at least within one score or so. If they fall too far behind, they just become fish out of water and look like crap but if they stay in the game they tend to be great at closing things out late.

For an example of this, look at last year's tOSU/PSU game:
Late in the game Ohio State led 20-13 (PSU had a Pick-6 and two FG's, Ohio State had two TD passes and two FG's). Penn State got to the tOSU 1 yard line but couldn't punch it in so Ohio State took over on downs up 20-13 with 5:13 to go:

1st and 10 from the 1, Howard ran for 4
2nd and 6 from the 5, Judkins ran for 9
1st and 10 from the 14, Judkins ran for 6
2nd and 4 from the 20, Judkins ran for 15
1st and 10 from the 35, Henderson ran for 6
2nd and 4 from the 39, Judkins ran for 2
3rd and 2 from the 43, Howard ran for 3
1st and 10 from the 46, Henderson ran for 7
2nd and 3 from PSU's 47, Henderson stopped for no gain
3rd and 3 from the 47, Howard ran for 7
1st and 10 from the 41, take a knee and go home with a W.

It was a bit of an unusual game (aren't they all). Ohio State dominated the stats but lost turnovers 2:1 including the aforementioned pick-6.

For the game Judkins had 14 carries for 95 yards and nothing over 22. He averaged 6.8 per carry and even if you exclude the 22 yard long, he averaged 5.6 on his other 13 carries. Henderson had 10 carries for 54 yards and nothing over 16. He averaged 5.4 and even if you exclude the 16 yard long, he averaged 4.2 on his other 9 carries. That wears down a defense. That final drive (above) probably doesn't happen against a fresh and rested PSU defense.

For an example of where even unsuccessfully running the ball can wear down a defense look at last year's iteration of The Game:
Michigan's RB finished with 32 for 116 with a long of 27. That is only 3.6 per carry even with the 27 yard long and if you eliminate that it is only 2.9. Note that the 27 yard long occurred on the game-winning drive, LATE in the game. On that drive he had:

8 yards on 1st and 10
4 yards on 2nd and 8, cumulative 12
-2 on 1st and 10, 10
6 yards on 2nd and 12, 16
27 yards on 3rd and 6, 43
5 yards on 2nd and 7, 48
-1 yard on 1st and goal from the 5, 47
2 yards on 2nd and goal from the 6, 49
1 yard on 3rd and goal from the 4, 50

On that drive he had 9 for 50 with the 27 yard long. That is 5.6 per carry. If you back that out of his overall total, prior to that he had:

23 for 66. That is an average of 2.9. It is clearly not good but it contributed to wearing down the opposing defense to the point where he was able to go 9 for 50 on that last drive. He didn't do that against Ohio State's fresh defense but they kept pounding away.

The key though is that it obviously only works if you can stay in the game. Michigan scored their previous 10 points in that game on five yards of total offense (a 2 yard TD "drive" after an Int and a long FG after gaining just 3 yards following a short punt). If Ohio State does ANYTHING other than try to pound the ball against the ONE elite unit on the entire Michigan team (DL) then Michigan doesn't get a chance to run it on a tired defense late in the game.

medinabuckeye1 · « **Reply #9 on:** **Today** at 05:15:29 PM »

Quote from: betarhoalphadelta on Today at 04:11:45 PM

On first and 10, a "neutral" game script situation, I think there's more desire to pass often enough that the defense needs to be on their heels. I'm not sure what that number is as a percentage for any given team--and it may be the answer is "more often than almost every team does". But I don't think we should extrapolate that to believe the metrics point to passing more in every spot, at least if you're looking at things like success rate or average yards per play. Because it's about sustaining drives, and passing has higher variability to go along with those higher averages. Variability is more likely to stall your drive IMHO.

This is why defenses key so hard on avoiding big plays. If a defense can avoid the big play then the opposing offense has to work their way down the field and usually that doesn't work. Usually there will be a sack or a holding penalty or back-to-back incomplete passes on 1st and second down. Usually there will be something that ends up leading to a 3rd and long situation and a punt.

Midrange variability doesn't hurt a defense as much as it hurts an offense. What I mean is that if a runner averages 5.5 (example above) and he gets 11 on one particular carry that only results in 1st and 10.

Think about holding penalties on the defense then the offense on back-to-back plays. Holding on the defense gives the offense 10 yards and a first down, now try to do it again. Holding on the offense puts them in a 1st and 20 situation where the chance of converting is minimal.

betarhoalphadelta · « **Reply #10 on:** **Today** at 06:11:49 PM »

Quote from: medinabuckeye1 on Today at 05:15:29 PM

This is why defenses key so hard on avoiding big plays. If a defense can avoid the big play then the opposing offense has to work their way down the field and usually that doesn't work. Usually there will be a sack or a holding penalty or back-to-back incomplete passes on 1st and second down. Usually there will be something that ends up leading to a 3rd and long situation and a punt.

And this is what we see these days in the NFL. Patrick Mahomes and the Chiefs were just torching opponents. Throwing massive numbers of explosive plays per game.

So defenses did what you suggest. "Okay, Patrick, we're gonna sit in a 2 high shell. If you think you can dink and dunk us down the field for a 14-play TD drive, good luck. But we're not gonna let you hit the big one."

And that dramatically affected the Chiefs' offense. It didn't "shut them down" of course; they still have a great coaching staff and a lot of talent. But it slowed them down quite a bit. And that's the same thing we've seen with several other teams with the big passing attacks.

It's yin and yang. If you are successful with the run, you suck the defenders down to the LOS and you make them vulnerable to the big play. If you're too successful with the big play, they play back and force you to run and dink and dunk it down the field.

And I still say that as it relates to 1st down, the preference is to run--for exactly what we're saying about the "cost" of the unsuccessful play being higher for the median "failed" pass than the median "failed" run. But if you're so predictable that the defense is selling out to stop the run... You'd better throw more to make them back the F off.

OrangeAfroMan · « **Reply #11 on:** **Today** at 08:11:09 PM »

Quote from: medinabuckeye1 on Today at 05:08:03 PM

I'm as much of a stats guy as anyone but I think there are two things that tend to get lost in the stats.

First is the consequences of an unsuccessful play. You already covered this somewhat but looking at Ohio State's offense last year:

Howard was 309/423 for 4,010 yards. That is 9.5 per attempt and 13.0 per completion on 73.1% completions.
Judkins had 194 carries for 1,060 yards. That is 5.5 per carry.
Henderson had 144 carries for 1,016 yards. That is 7.1 per carry.

The average completion (13.0 yards) is a first down but even for the Buckeyes with Howard finishing second nationally in completion percentage (among full-time starters), still more than one in four of their pass plays was not completed.

Meanwhile, the two main RBs each averaged a "successful" 1st and 10 play.

Still, unsuccessful plays happen. If you think of it as four consecutive 1st down plays on one drive and adjust from the average a little bit, I think you'll get something like this for passes (best to worst, obviously):
17 yard completion, 1st and 10.
13 yard completion, 1st and 10.
9 yard completion, 2nd and 9.
Incomplete pass, 2nd and 10.
The first three are great but the incomplete pass now puts you in a tough spot.

With running plays you get something more like:
8 yard carry, 2nd and 2.
6 yard carry, 2nd and 4.
4 yard carry, 2nd and 6.
2 yard carry, 2nd and 8.
As a fan, if my team gets 6-8 yards on 1st and 10 I'm feeling really good. If they get 4 I'm nervous but still feeling pretty good. If they get 2, that makes things tougher but you are still a threat to run on 2nd and 8 because an average carry of 5.5 (Judkins) to 7.1 (Henderson) gets you to 3rd and 1-3 which isn't bad.

When you throw incomplete on 1st and 10 it more-or-less makes you one-dimensional because running on 2nd and 10 usually results in 3rd and long.

Great point.

I don't think many people realize that a successful pass of 10-15 yards simply hits the reset button on the 10 yard fight. Averaging over 10 yards per pass is almost too good, as it sort of negates its efficacy over a run outcome. Those incompletions are damning every time.

Quote from: medinabuckeye1 on Today at 05:15:29 PM

This is why defenses key so hard on avoiding big plays. If a defense can avoid the big play then the opposing offense has to work their way down the field and usually that doesn't work. Usually there will be a sack or a holding penalty or back-to-back incomplete passes on 1st and second down. Usually there will be something that ends up leading to a 3rd and long situation and a punt.

Good point again.
It reminds me of baseball. The optimal strategy there is to get guys on base and slug them in. Why? Because it's harder to string together a series of singles and doubles to score as many runs. THAT'S what makes OBP such an important stat.

I think running the ball a lot on 1st down is epicurean: it's more about the avoidance of pain (2nd and 10) than the pursuit of pleasure (2nd and less-than-5). Running plays avoid 2nd and 10 a helluva lot more than passing plays do.

OrangeAfroMan · « **Reply #12 on:** **Today** at 08:16:07 PM »

Oh, and just for people's understanding of run/pass percentages, a balanced offense isn't 50/50. Bad teams tend to pass more than they otherwise would because they're always behind. Good teams tend to run more than they otherwise would, because they're ahead and running out the clock.

There is no answer to what percentages are actually balanced, but I set it at 62/38 (run/pass) to 54/46 as balanced. 37% passing and below is run-heavy. 47% passing and above is pass-heavy.
But that's just me, and it's just for a game....and none of it involves NFL stats, just thousands of college teams.

betarhoalphadelta · « **Reply #13 on:** **Today** at 08:43:14 PM »

Quote from: OrangeAfroMan on Today at 08:16:07 PM

Oh, and just for people's understanding of run/pass percentages, a balanced offense isn't 50/50. Bad teams tend to pass more than they otherwise would because they're always behind. Good teams tend to run more than they otherwise would, because they're ahead and running out the clock.

There is no answer to what percentages are actually balanced, but I set it at 62/38 (run/pass) to 54/46 as balanced. 37% passing and below is run-heavy. 47% passing and above is pass-heavy.
But that's just me, and it's just for a game....and none of it involves NFL stats, just thousands of college teams.

I think that's a little bit reductionist. Some teams are just built around the pass. Mike Leach's TTU teams, Joe Tiller's Purdue teams, June Jones' Hawaii teams, etc. Those teams weren't "always behind". Sometimes they were behind against the best teams (because they weren't power teams), but they were still throwing it around against lesser competition because that was their identity.

Your percentages might be viable for helmet teams in CFB, who can just out-muscle lesser opponents. But in the NFL where there is parity, in 2024 there were only 3 teams out of 32 in 2024 who were below 50% passing plays. (Admittedly the Super Bowl champ was #32 of 32.)

Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

The 'Old' CFN/Scout Crowd- Enjoy Civil discussion, game analytics, in depth player and coaching 'takes' and discussing topics surrounding the game. You can even have your own free board, all you have to do is ask!!!

Anyone is welcomed and encouraged to join our FREE site and to take part in our community- a community with you- the user, the fan, -and the person- will be protected from intrusive actions and with a clean place to interact.

Topic: Optimal First Down Run/Pass Balance

betarhoalphadelta

Optimal First Down Run/Pass Balance

bayareabadger

Re: Optimal First Down Run/Pass Balance

Mdot21

Re: Optimal First Down Run/Pass Balance

FearlessF

Re: Optimal First Down Run/Pass Balance

Mdot21

Re: Optimal First Down Run/Pass Balance

FearlessF

Re: Optimal First Down Run/Pass Balance

Mdot21

Re: Optimal First Down Run/Pass Balance

betarhoalphadelta

Re: Optimal First Down Run/Pass Balance

medinabuckeye1

Re: Optimal First Down Run/Pass Balance

medinabuckeye1

Re: Optimal First Down Run/Pass Balance

betarhoalphadelta

Re: Optimal First Down Run/Pass Balance

OrangeAfroMan

Re: Optimal First Down Run/Pass Balance

OrangeAfroMan

Re: Optimal First Down Run/Pass Balance

betarhoalphadelta

Re: Optimal First Down Run/Pass Balance