website statistics
Yes. With 1:06 left to play, SEA snapped the ball on a 1st and goal from the NE 5, and Marshawn Lynch ran it to the NE 1. At this point, Belichick had a decision to make. With 2 timeouts remaining, he could stop the clock preserving time for a possible comeback FG drive to tie the game if SEA scores. Or, he could let the clock run, save his timeouts, and let SEA deal with the added pressure of clock management. 

Clearly, SEA had a preference. They took the full 40 seconds getting the 2nd down play off, trying to take as much time off the clock as possible. They anticipated scoring quickly and did not want NE to have enough time to respond. We all know the final outcome, but that could not have been anticipated with any confidence at the time of Belichick's decision.

At the time, I thought it was a big mistake by NE not to use a timeout. If SEA scores on the next play, NE has over 50 seconds left and a timeout to respond, which is plenty of time for a FG drive in the modern game. And if SEA takes 2 plays to score, NE could use their final timeout to preserve over 45 seconds to respond, still enough time for a FG. There are a lot of different paths based on when or if SEA could score, so the simulator is probably the right tool for this analysis. I set the WOPR to run two scenarios, one in which NE calls a timeout, and one in which they let SEA run the clock down.

The results say using the timeout is the better option by a confident margin. The simulated win probability for using the timeout is 0.231 sWP, and for not using the timeout is 0.189 sWP. (The two results are well outside 4 standard errors of each other.)

The difference of about 4 percentage points of win probability means that NE would have about an 8 to 10% chance at a comeback FG drive to tie, which is consistent with endgame drive percentages. 

I don't buy the suggestion that Belichick could anticipate that SEA would burn so much time before the second down snap that they'd be forced to throw at least once, which would take Lynch out of the equation for a play, and so on. There's just way too many assumptions and possibilities to consider. I'd buy that he could, generally, prefer to put all the pressure on SEA. However, the cost of doing so forfeit any opportunity to respond. The final outcome aside, I think Belichick should have used his timeout.

Leave your comments

Post comment as a guest


Comments (26)

Load Previous Comments
  • Guest - Mikey

    Thanks, Brian! Can you also do the math on whether Belichick should have told his defense on first down from the 5 yard line: "if they hand off to Lynch and he gets through the line of scrimmage, allow him to score"? That's what I was yelling at my TV, and am wondering if the math supports it.

  • Guest - SPH

    So we're talking about a calculated sWP of only .04? Given that all the mitigating factors are likely to favor not calling a timeout (Belichick clearly liking the defense he had on the field and how they had prepared for Seattle's goalline package, Seattle having a historically good defense at defending deep and sideline passes, making a tying FG that much harder than usual), this looks like a situation where it really didn't matter much.

  • Guest - Scott

    Brian Burke, because time is just one factor. Seattle could have made a trade-off if they liked running there, by not letting the clock run so low and giving themselves time to call three running plays, potentially. They chose to take off time at a trade-off of having to call one pass play on 2nd or 3rd down, which is a good value trade. New England's primary goal is to prevent the touchdown; taking the time out hurts that goal by giving Seattle time to get organized. As Belichick said, he liked the goal package match up he had. Also, a time out he takes now is a time out he can't use later, if they score, so the clock benefit of a time out there isn't clear anyway.

    The Patriots real clock management mistake was on the kick-off with 2:02 remaining. Kicking through the end zone means no time runs off the clock, means the Seahawks get a free timeout at the two minute warning. Forcing them to play the kick has a lot of clock value there.

  • Guest - Jonathan

    How is kicking it short any different from giving the other team a free play before the two minute warning. If the ball gets returned from the 5, Seattle gets that play to advance the ball past the 20 (and they are guaranteed a first down at the end of the play).

    Anyway, this post does a good job of explaining the downside of snapping the ball early enough for three scores. You give Brady 40 seconds and two timeouts, much different story from 22 seconds and two timeouts.

  • Guest - RickD

    It's entirely possible that Belichick looked at the personnel on the field, figured that he had the advantage, and that's why he didn't call the time out.

    Given that his team won the game on that play, it seems a bit arrogant to positively assert that a probabilistic analysis should be treated as superior to the thinking that was actually used.

  • Guest - spike

    Understandably, this analysis doesn't assume Belichick or the Pats had additional information to give them an advantage. Such as knowing the likelihood they would run a play in a given formation on a given part of the field. There's a sense of hubris to my argument, but it's pretty well substantiated by the Patriot's success over the years. Is it not the case this analysis actually indicates what an "average" NFL coach / team should have done in their situation?

    I don't know if this is true, but my guess is that the best play caller (not trainer of players) isn't the one that can spit out these numbers and choose the higher probability. It's the one with more information than is provided in the "simulator", such as which players are "in-the-zone" and who thrives in high-stress moments.

    Having said that I love this site and understand what it is and what it's attempting to do. I just believe that the results spit out of these analyses shouldn't pick the ultimate decisions, but inform them.

    Nobody will ever know if Belichick's decision not to call the timeout was an informed one, relying on some relevant information he gathered from game film or receiving a timid hand-shake from Pete Carroll during pre-game. Gut reactions rely on inputs, regardless of our ability to be conscious of them.

  • Guest - bill

    Jonathan, Kicking the ball short may have been a better decision because the Patriots kicked off with 2:02 on the clock. By Kicking the ball short the receiving team would likely have to spend more than 2 seconds on the return, thus negating the 2 minute warning and using up one of the time-outs Seattle had left.

  • Guest - JC

    Maybe. The stats must show this. But don't you sort of feel that if BB calls a TO there, Seattle scores a TD. They probably change their play, or at least urge Wilson not to risk a turn over. I can't help thinking that the pressure contributed to the pick. It felt to me like a dice roll by BB - but one based on hours and hours of watching film and suspected what they might run.

  • Guest - zopzoop

    NASA official said that there are aliens on the moon! Many things can not let people think, the United States Apollo 11 lunar month after no moon on the plan, the former Soviet Union has no plans to lunar landing. Recently, some experts reported that the US astronauts found aliens on the moon a lot of information, China canceled the moon program! This is the world's space departments to communicate with each other to prevent accidents.Zenith's name and the origin of many Swiss traditional watch brand names are not the names of the founders, nor the names of watchmakers or important mechanical inventions. Legend of the founder of the brand Georges Favre-Jacot had to walk in the night sky, looked up and looked into the sky, even from the universe to get his mysterious inspiration. A large celestial system running around the Polaris made him think of the sophisticated precision structure of the watch's gears running around the axis.