You’ve heard it once. You’ve heard it a million times. Some brainiac holding box score statistics goes off about how you can’t abandon the run in the NFL. The lecture quickly turns into the genius proclamation that teams who run the ball more tend to win. This is innocuous enough. But like, no shit. Any talking head from ESPN to The View knows damn well that NFL teams run the ball more when winning to ice the game away.

Where shit goes awry is when these idiots connect rushing amount totals as some sort of prescriptive recipe teams should abide after the fact. It’s the single most obvious example of “correlation doesn’t equal causation” in the history of sports and I’m here to beat this idiotic nonsense like the dead horse it should be. Anyone within a mile radius of a 100-level statistics course immediately sees how tragically flawed such a claim is.

Yet this notion not only pervades, but thrives in public discourse:

Yo @NFL. Here are top 5 rushing teams:
1. @ravens
2. @49ers
3. @seahawks
4. @colts
5. @BuffaloBills

Combined record is 45-15

To the teams who are struggling; please think about this. #BaldysBreakdowns

— Brian Baldinger (@BaldyNFL) December 3, 2019

#Bears are 4-1 when David Montgomery rushes 15 or more times/game.

They are 2-5 when he rushes less than 15 times/game.

— The Bears Report (@TheBearsReport) December 4, 2019

The magic number for Rams: Todd Gurley at 20 touches.

LA’s won 19 of last 20 times Gurley gets 20. They are 27-8 in his career, 3-0 vs #Seahawks, when he gets 20.

But Gurley’s had 20 touches just twice in LA’s last 17 games. Both are in wins the last 3 weeks. @thenewstribune

— Gregg Bell (@gbellseattle) December 7, 2019

This is hardly an exhaustive list. There was even a columnist for the Tampa Bay Times just a few weeks ago going off after the Buccaneers rushed eight times in a blowout loss to the Saints.

“Since 1970, a team has run the ball eight times or fewer just 20 times. Those teams are 0-20.”

He went on to proclaim:

“Points ultimately matter most. But running the ball — even when from behind — allows a team to take control of a game.”

I’m going to put this kindly so as not to drag one particular columnist when many others would be as deserving, but this might be the stupidest idea anyone who is in any way affiliated with football got paid to make since the NFL signed off on the replacement refs. To claim prescriptively that teams must rush more than eight times per game blatantly fails to address the game situation that led teams to abandon the run in the first place. It’s like looking back at your blackjack success rate when hitting on all your past 16’s vs a dealer face card. Your total success when finding yourself in this situation is going to be bad. But hitting (much like passing in the NFL) on that 16 is going to help you lose less than standing.

You can see what I mean by this from the 20-game sample given by Tampa Bay Times guy. These 20 losing teams allowed their opponents to score an average of 31.7 PPG. I suspect an argument can be made that when you abandon the run, teams cede time of possession to opponents allowing them more opportunity to score as an effect. It’s a hypothesis, but quite simply it doesn’t check out. These teams trailed by an average of 7.85 points – a two score deficit – after just the first quarter in these 20 games and fell behind a whopping 16.75 point average at half-time.

These teams weren’t bad because they abandoned the run. They abandoned the run because they were bad! The starting situation of this set of games was cherry-picked just like looking back at the last 20 times you got dealt a 16 vs a ten in blackjack. Not exactly rocket science Tampa Bay Times guy.

But to really know if sticking with the run is better than abandoning it, the only thing that really matters is what teams who WIN do, not teams who lose. That is going to be the crux of this analysis. I compiled all the NFL regular season play by play data publicly available (from 2009 to midway through 2019) in order to isolate the percentage of rushing plays called for all teams who won after being down by any margin up to 20 points. But using the whole game’s worth of plays is useless. The only way to determine what winning teams do to come back in games is to consider plays in between when teams fell behind by x and either caught up to tie or take the lead. Nothing matters before. Nothing matters after.

Example: In Super Bowl LI, the New England Patriots were down 28 – 3 with 8:31 left in the third quarter leaving them a deficit of 25 points. Atlanta Falcons fans may recall this. The Patriots came back to tie the game with 57 seconds left and went on to win in OT. In the data I looked at, not enough teams came back from 25 down to be useful, but if 25 points was the deficit we were looking for figures on, we would only consider plays after 8:31 left in the third quarter, and before 57 seconds in the 4th quarter. If “control the game” theory has any merit, we should see evidence that winning teams don’t abandon the run. This of course, is wildly wrong. And as this anecdotal example turned out, the Patriots rushed just 20.5 percent of the time during this span (8/39 plays).

The following table provides the results in their entirety. As you can see, especially for smaller deficits, this is a massive amount of game and play data involved making the upcoming results pretty hard to deny:

Let’s put this table in graph form for a crystal clear picture of what’s going on:

Remember, only teams who came back and won were included in this analysis. The trend here could not be more obvious. Teams who came back to win after trailing by some margin scale back the run more and more as a function of their deficit. The three vertical gray bars point out a striking drop in rush percentage immediately upon falling behind, then again when their deficit surpasses one score (8 points), and once more after falling behind two scores (15 points) if we consider two-point conversions as an individual score. This gives pretty clear evidence that abandoning the run – to a degree – isn’t just what spreadsheet wielding Twitter nerds say to do; it’s what WINNERS in the NFL do. Forgive them for not “taking control of the game” and quitting on their all sacred game plan. When their backs are against the wall and teams deal themselves a shitty 16, winning teams in the NFL increasingly scale back the run upwards of 25 percent on average in favor of passing.

Quick sidebar: Funny enough, rush percentage begins to increase after an 18 point deficit and the trend continues to regress upwards towards the NFL average indicating that teams basically start to give up when losing by 18 points or more. Even if momentarily before managing to come back to win.

Now that we looked at the teams who come back to win from a deficit, how about the inverse? What percentage of plays do teams rush who blow a lead of x? Same theory applies as before; only plays between the time a team obtained a lead of x and blew the lead were considered:

As you can see, teams that blew a lead averaged rushing well above the overall NFL average in order to ensure snatching defeat from the jaws of victory.

The data from these two graphs leads me to make the following inference:

Not only is scaling back the run beneficial for teams who are losing, but NOT doing so is detrimental for teams who are winning.

All in all, while there is no team who won a game when rushing eight times or fewer, it’s incredibly obvious that abandoning the run like all 20 teams from Tampa Bay Times guy’s example did was because they needed to take a risk after getting their doors blown off from the start. They looked at their lousy 16 and had the balls to take a hit. And hey, they might have played like crap that week, but at least they tried. They took the exact strategy that winning teams take to comeback and win in the NFL. They played the odds and took the hit, busted, and had to deal with Tampa Bay Times guy lecturing them even though he’s probably the guy that would stand on an 11.

I just hope that when the dealer rolls that second face card and Tampa Bay Times guy also loses with all his old-school “stay the course” Luddites, they feel like they were in control.

-El Jefe