I mentioned this play during Saturday’s NFL Open Thread:
Should San Francisco 49ers quarterback Alex Smith have scored the go-ahead touchdown against the New Orleans Saints? Maybe not.
Here’s what actually happened: With 2:18 remaining in the fourth quarter, the 49ers were down by 1 point with the ball on the New Orleans 28 yard line. On third down, QB Alex Smith ran a bootleg and scampered down the left side of the field and into the end zone. The 49ers attempted the 2-point conversion but missed it, leaving the 49ers ahead by 5 points. That gave the Drews Brees-led Saints offense the ball back with 2:11 left and one timeout.
Not surprisingly, the Saints roared back down the field and scored a quick touchdown themselves; 2:11 is a lot of time for a team like this with so many downfield passing options. Alex Smith’s touchdown was celebrated by the announcers during the game and not a single mention was made that the 49ers could easily have lost that game due to leaving too much time on the clock for the opposing team. (The 49ers did end up winning, but it easily could have gone the other way.)
So let’s back up a second and re-consider a different scenario:
What if on third down, Alex Smith ran down to the 1 yard line and dove to the ground, leaving the ball a foot or two short of the end zone?
In that scenario, the Saints would most likely would have used their last timeout to stop the clock. Then on first and goal, the 49ers could either kneel down or, even better, run a simple running play designed to take time off the clock without actually scoring. They could do the same again on second down and third down. On fourth down, if there is one, they simply kick an easy 17 yard field goal (NFL kickers are at around 99% from that range) and go up by 2 points…. but leave New Orleans with only 40 seconds on the clock and no timeouts.
At first I thought maybe this was another of my half-baked football ideas (like running an offense with 2 different QBs on the field at the same time), but it turns out I’m not alone. The crew at NFL stats guru site FootballOutsiders.com, (who watch a lot more football than I do and regularly appear on the NFL Network) were discussing the same play:
Aaron Schatz: Alex Smith responds with a beautiful 37-yard throw to Davis, who was streaking down the left side of the field in front of Jenkins. Now they’re in field goal range and the 49ers have to make sure they run as much time off as possible so Brees doesn’t get the ball back after any go-ahead field goal.
Or, Alex Smith could run a quarterback sweep on third-and-8 that goes for a touchdown. 2:11 and one timeout left. Should Alex Smith have gone down at the 1 so the 49ers could run out the clock and kick a field goal instead of scoring a touchdown and then giving the Saints the ball back?
Vince Verhei: I think that play was the highlight of the season so far. The Saints had one timeout plus the two-minute warning. I can’t do math right now, but you may have a point.
Tom Gower: The calculation is roughly up five or seven with 2:11 to play versus a 98 percent chance of being up two with about 40 seconds left. I think they’re better off with him taking the knee at the 1.
Aaron Schatz: And of course, the Saints make an insane 70-yard pass where nobody can tackle Graham and they take the lead.
Vince Verhei: Well that’s all moot. Good thing SF left themselves enough time to get the ball back. Maybe Graham should have gone down at the 1 too.
Brian Burke of AdvancedNFLStats.com also agreed:
It was a high-drama end-game, to be sure, but it never needed to happen had Alex Smith slid before crossing the goal line. With 2:18 to play, SF was down by 1. Smith ran a naked bootleg 28 yards for the go-ahead TD. Instead of scoring the TD, had Smith taken a knee SF would have been able to run the clock down under 40 seconds before a chip shot field goal attempt. NO would have had a little over 30 sec and no timeouts to respond with their own FG. That might not be a 100% lock, but it’s a better bet than leaving 2:18 plus a timeout to get a game-clinching TD.
NO didn’t have that luxury. Being down by more than 3, they had to score the TD. But they scored too fast. NO took only 23 seconds to score the go-ahead TD, leaving 1:37 on the clock for SF. During games, I’m often asked, “Team X just scored a 30-yard touchdown, so why did their [Win Probability] dropp according to your stupid model? It’s obviously broken.” No, it’s not broken. It’s that sometimes teams are better off burning clock on their way to a score than scoring instantly.
After scoring the TD, the 49ers smartly went for 2 to make it a 7 point game. A 7 point lead at that point is worth 0.86 WP, and a 5 point lead, should they fail, is worth 0.78 WP. Given a 45% conversion rate, that’s a total WP of 0.82 WP for the TD. Running the clock down to 30 sec with a 2-point lead is worth a 0.91 WP–a much better bet. But hey, then we wouldn’t have seen one of the most exciting endings to a playoff game in a long time.
Here’s Brian Westbrook taking an intentional 1 yard line dive in 2008 against the Cowboys:
And here’s perhaps my favorite example of this strategy. In 2009, the Jets tried to let Jaguars running back Maurice Jones-Drew into the end zone in order to get the ball back with as much time left as possible. But MJD did the smart thing:
The intentional non-score (or its counterpart, the intentionally-allowed touchdown by the defense) is one of those counter-intuitive football strategies that hasn’t quite yet gained full acceptance. Fans and announcers want the highlight reel-worthy touchdown or big goal-line defensive stand. They don’t necessarily get all worked up about clock management. But in many cases you are playing against the clock, and being down a point but still having the ball is often far better than giving the ball back to Drew Brees with lots of time left — as the 49ers found out and were nearly punished for.
