I'm interested in the following situation: a team in American football possesses the ball with less than a minute left in the first half of a game and is facing a long field to score. Oftentimes, teams will have their quarterback take a knee here to effectively end the half. This seems to be pretty common and it's commonly accepted as sound strategy, but I'm struggling to see how it is the most beneficial choice for the team. You can sometimes see teams do this at the end of a tie game--"playing it safe" and taking their chances in overtime.

From my point of view, the only outcomes that matter from running an extra play or two are either A) Your team scores, or B) The other team scores --any other result is meaningless. Do you teams really think that running a play (even a "safe" one like a run up the middle) will result in an opponent scoring more often than their own scoring probability? If so, what makes this different than any other play; i.e., why run any offensive play at all at any point?

Another consideration that I've surmised is that a team may be worried about having a possible injury. That seems like a pretty small risk to me and you're just as likely to have an injury happen to the opposing team. I know that these plays aren't usually a big deal, but what am I missing that makes this strategy so well-accepted?

3 Answers 3


It's hard enough to score in American football when you have half an hour to score in. Trying to score when you have 0:30 left in the game? Nearly impossible, if you're on your side of the field.

The fundamental problem here is that the defense knows what's going on. They know you have 0:30 left, and need to get to the 30 yard line or so in order to kick a FG. If you're, say, on your own 30 - then you have 40 yards to go. So the defense just needs to make sure you don't get to the 30, but is perfectly okay with you getting ten or twenty yards - as long as you don't get those 40.

So you're left with the problem, particularly if you are out of time outs, of quickly getting those yards. You're going to run the ball? Might work, and some teams do. HB Screen pass actually is a bit better - lets you get the WRs downfield to clear out the safeties from the box. Or HB draw. That's better, except when it's mercilessly mocked of course. And it doesn't tend to work - maybe Adrian Peterson might knock in an 70 yard run, but more running plays end in fumbles than in 70 yard runs.

Specifically, there have been 68 fumbles lost in 9221 rushing plays so far in 2015. 3/4 of a percent of plays involve a fumble lost. This doesn't count plays that are aborted, either, i.e. fumbled snaps. On the other hand, there have been 23 rushing plays yielding 50 yards or more so far this year. 40 yards is a bit more, but still well under 68 - 43 plays total. And that includes many situations where the defense is playing more of a press defense - something they would never do with 0:30 left to go in the half. All that, and a rushing play that didn't go out of bounds would have to either score or leave enough time on the clock for a FG, right?

Okay, so we're not going to run - we're going to pass, right? Except, that's far worse. Rushing, odds are nothing happens. No fumble, no points, just a few wasted plays.

Passing, though, you have a big risk. When the defense knows you're going to pass, or at least is okay mostly ignoring the run because they can still play such that the RB gets 10-15 yards but no more, you have a lot more interceptions.

202 of these plays have been run in 2015 with 2:00 in the 4th or less remaining, down by a TD or less (but down), and 10 of them resulted in picks. 5% of plays like this result in picks. While most of those aren't pick sixes, that's often because a game-ending pick is brought out of bounds or just downed to avoid the risk of a fumble by the DB who picked it off.

On the other hand, only one was successful - Jameis Winston vs. the Giants - at scoring a TD, and ... oh, wait, no, that one wasn't successful either - the other team scored on a fumbled lateral.

Zero plays in that situation have been successful this year: because defenses know how to prevent them. Some probably have yielded a FG eventually - but it's not common, and it's much more likely to lose the ball.

Zone defenses are the main reason for this. Defense knows you have to pass the ball 20-30 yards, right? So they sit back far off the line, facing the QB, and everyone reads the QB's eyes. That makes it much easier to pick the ball off (rather than just running alongside the WR and trying to keep up with him). Zone prevent defenses are terrible at keeping offenses from advancing the ball 10-20 yards - and in a longer situation, 2:00 drill, they're actually pretty bad defenses. But with 0:30 left? Quite effective.

So - what's your call?

(Of course, as Bill Barnwell loved to point out, coaches still do run these plays... especially the basically pointless HB draws...)

On a side note, I think that the injury bit is relevant, though I suspect it's only because the odds of scoring are so low already. NFL teams have injury risks on every play, so it doesn't really make sense to try a play when you don't need to AND it won't give you a benefit. I think it would be hard to prove that coaches use this reason for kneel downs without asking them; but Bill Barnwell in the many Thank You For Not Coaching articles since 2013 certainly thinks it's relevant, and if he does, that's good enough for me.


I don’t think injury prevention is a consideration for taking a knee. I think there are many factors involved, some of which are, the current score, the offensive team’s ball position, time left on the clock, and, the unwillingness to risk a defensive score on a last minute turnover. There is no “safe” play per se. Fumbles can, and typically do, happen when you least expect them. And, there is something to say about "momentum". Not giving it to the defensive team, especially with a close scoring game, is a typical goal.

That’s not to say that taking a knee is a standard practice. Some coaches will offset the risk by throwing deep down field understanding that if the ball is intercepted, the chances of a defensive score are small with the distance needed to score. But, in the end, why take the chance when you can take a knee and re-group in the locker room during half time?

  • I find it hard to imagine that coaches don't consider injury risk to some degree. Not only does an injury to a star player (say, the star RB...) risk derailing the season and possibly the coach's job, but coaches and players are people, too.
    – Joe
    Commented Nov 30, 2015 at 22:35

After reading Joe's and rrirower's answers and having some time to think about, I'm now comfortable with an answer to the crux of my question which is: if running a play here is a net negative, why run any offensive play ever?

The difference between this situation and a normal drive with ample time remaining is that in the kneel-down situation where you could run one more (or two or three) play, the defense actually might have a better chance of scoring than your team...with an entire drive, your chance of scoring is higher than the defense's.

It's counter-intuitive at first (at least to me), so here's some completely made up numbers to illustrate the principle. Say you have the ball at your own 20 yard line with a few seconds left and you want to run a play. Your odds of scoring might be 0.5% while the defense has a 1.0% chance--the upper hand goes to the defense. The difference with a drive where you have time for more plays is that in the 98.5% of the time where nobody scores, you are hopefully advancing the ball a little further. Each subsequent play garners little chance of a score for either team, but you increasingly improve your odds of scoring while decreasing the other team's odds of scoring as you advance the ball. By the time you make it to the other team's 20 yard line, your chances of scoring on a given play might be 12% (again, completely made up) while the defense now has a 0.25% chance.

Like I said, those numbers aren't real, but the principle stands that in being able to put together a drive, you continually increase your odds of scoring compared to the other team. There exists an inflection point somewhere on the field where anything before it is a net negative for that play (as illustrated by Joe). So a single play would not be a good idea. But a drive has a sort of cumulative effect wherein the plays taken as a whole are more often a net positive.

  • 1
    This isn't really an answer to your question, and really is more of an interesting question in and of itself - though I'm not sure it's a fit for SE, since it's a bit more complicated. But it might be a good question if narrowly scoped. You're sort of right, but missing some big chunks (in part because of your made-up numbers).
    – Joe
    Commented Dec 1, 2015 at 15:15
  • I see what you're saying--your answer addressed the question correctly, I guess I was just trying to work out the one remaining piece of it that wasn't making sense to me which is: if taking a knee here is the right move (and so easily and commonly accepted as being right), what makes it so completely different from another point in the game where taking a knee would be laughed at as an option? Like you mentioned, there are defensive adjustments which is a huge part. Could you elaborate on any of the parts you think I'm not understanding? Commented Dec 1, 2015 at 15:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.