The Tape Room

The Scramble Is Football's Best-Kept Secret

When a quarterback tucks the ball and runs, the play works about six times in ten. Nothing an offense calls on purpose comes close, and no offense can call it on purpose.

Ask a coordinator for the best play in his arsenal and you will hear about a shot play off a hard run fake, or a screen thrown into a rotating defense. The play-by-play record has a different answer, and it is a play that appears on no call sheet: the quarterback giving up on the pocket and running.

In every season since 1999, a scramble has come up as a win for the offense, meaning it produced positive expected points, about 59 percent of the time. Everything else an offense does from scrimmage, every called run and every called pass, wins about 42 percent of the time. The gap has been between 13 and 20 points of success rate every single year.

In the modern game the per-play values are almost comical. Since 2016 a scramble has been worth +0.480 expected points each time it happens. An ordinary dropback earns +0.018. A designed run loses value on average. The most efficient common play in football is the one nobody draws up.

Worth nearly half a point every time it happens, and impossible to order up. A scramble exists because the defense covered everyone and left the grass open.

Why nobody can just call more of them

The number is enormous precisely because of how a scramble is born. A quarterback takes off when coverage has held, the rush has opened a lane, and the field in front of him is empty. The escapes that fail never make it into the scramble column at all; they get logged as sacks. So the right reading is strictly conditional: a scramble is worth this much when it happens. It is not evidence that any team should scramble more, and a coordinator cannot manufacture one any more than a poker player can manufacture being dealt aces.

The runner is who he is

What a team can know in advance is which quarterbacks will generate them. How often a passer scrambles per dropback is one of the stickiest traits we have ever measured: the year-to-year correlation has run between 0.66 and 0.92, positive in all 26 season-to-season comparisons since 1999. That is far more stable than accuracy numbers, which persist at roughly 0.4 to 0.5. Whether a quarterback is a runner is not a mood, and it does not change with a new coordinator. The statues stay statues.

And the league has been drifting toward the runners. Scrambles have more than doubled since the pocket-statue era, from about 2.8 percent of dropbacks in 2009 to about 5.8 percent in 2025. The full history is a little humbling: the rate was around 4 to 5 percent in the early 2000s, fell as the league went pure pocket, and only then began the long climb. The position got faster, and coaches stopped treating a quarterback run as a broken play.

The tax that keeps it honest

The dark branch of leaving the pocket is the sack, and the scariest part of a sack is the ball, not the lost yards. Depending on the season, between 10.6 and 20.5 percent of sacks have shaken the ball loose, more than ten times the fumble rate on all other dropbacks, which sits around 1 percent. The tax has eased over the years, from roughly 16 to 21 percent of sacks in the early 2000s to 11 to 13 percent recently, as quicker releases and pocket-safety coaching spread. Even the modern floor makes every sack a turnover lottery ticket.

So the next time a broken play ends with the quarterback sprinting for eleven yards, resist the urge to call it chaos. Play for play, you just watched the most valuable thing that regularly happens on a football field, and the reason it looked easy is the same reason it cannot be scheduled.