I’ve been watching NFL football all my life, and endgame timeout-strategy (being susceptible to analysis by non-football people) has always attracted my attention, but there is one numerical quirk with big potential consequences which I only noticed this season. Consider this scenario:

Late in an NFL game, Team A has the ball and a small lead. Team B has 1 timeout left. Team A, on 1st down, runs it up the middle for no gain. The play ends (to be precise, the 40-second play clock is reset) with 2:43 remaining. What is Team B’s strategy?

The obvious thing is to call the timeout now; when it appears not to matter, I mostly see teams use their timeouts on early downs rather than later downs. But in this case, that would be an enormous mistake! Consider the very likely continuations:

Use timeout now:

2nd down: Play begins 2:43, ends 2:38. Clock ticks down to 2:00. Two-minute warning.

3rd down: Play begins 2:00, ends 1:55. Clock runs to 1:16.

4th down: Punt, snap at 1:16.

Save timeout:

2nd down: Play begins at 2:04, ends 1:59. Two-minute warning.

3rd down: Play begins 1:59, ends 1:54. Team B calls final timeout.

4th down: Punt, snap at 1:54.

Using the timeout early costs almost 40 seconds! It costs almost the full value of a timeout, because it removes all the power of the two-minute warning. I can’t actually cite a game from memory where someone screwed this up, but I don’t think I would have been alert to it until recently. Given that I’ve never heard this mentioned on TV, and given the general propensity of NFL coaches to screw up timeout strategy in more obvious ways, I would expect a high proportion of mistakes when this situation arises.

The bottom line: If the clock resets with roughly 2:41-2:49 remaining, don’t use your timeout now. Wait until after the two-minute mark, to make sure you get proper value from the two-minute warning. Note that in college football there is no two-minute warning, so this whole discussion is moot.

One tiny caveat: The recommended strategy does give Team A a free license to throw the ball on 2nd down, since an incomplete pass won’t matter. They might do so and get a first down, ending the game. But there is no way this possibility cancels out the huge time savings. Team A always has the option to throw it anyway, if they want to take the risk of leaving 40 extra seconds. That is, the waiting strategy is (almost) dominant.