NFL

Mike Smith Doesn't Understand Math, Which is Why the Falcons Attempted a Field Goal Down 4 Late in the Game

USATSI_7570302_153192880_lowres

I don’t know what is more mind-numbingly awful. Mike Smith opting to kick a field goal down by four points with 2:24 left in a game, or the almost universal praise after the game that this was the right decision.

I can go about explaining it any number of ways, but this was a dreadful decision. Here is Mike Smith:

“There’s not a whole lot of fourth and 14′s. We had I believe it was around 2 minutes, 28 seconds, so we had a timeout before the two minute, use the two minute as a timeout, we were going to have an opportunity to get the ball back with the offense, only having to kick a field goal to win. Matt Bryant has kicked a lot of clutch field goals for us, and giving us an opportunity.”

The play was a 4th and 15. There have been 306 attempts on fourth down of between 13 and 17 yards going back to 1999. It’s not an insignificant number. This isn’t a rare case where teams never go for it in this situation at the end of the game. This would be the exact kind of situation where having someone with a sense of actual results in other games would help in decision making.

The NFL Network analysts, most of whom I believe were once competitors who must have had great belief in themselves, acted like this was an impossible feat. God forbid that a football team ever pick up 15 yards on a play by means of this forward pass.

Teams have converted 20.3% of fourth down attempts of between 13 and 17 yards (62 of 306). Let’s just call it 20%. Does that change your view of the call?

Here’s another number to compare it to: 24.8%

That’s the percentage of time that New Orleans has a first down with the ball, and fails to pick up another one. (via Football Outsiders Drive Stats)

Mike Smith punted last year to Brees and was criticised. He can't win because the media is always right.Those numbers aren’t that different, but one is viewed as an impossibility, while the other is just glossed over and vastly overestimated. The payoffs, of course, are much different. Going for it and converting results in an outright touchdown sometimes (a quarter of all those long fourth down conversions actually scored on that play). Even when it doesn’t you are inside the red zone with a first down, and a favorite to score a touchdown with plenty of time to call any play.

Using the win probability estimator at Advanced NFL Stats, I estimate the chances of Atlanta winning based on choosing the long field goal at 9.8% (which accounts for the chances of a miss) and the odds of winning based on going for it at 15.7%.

That is a massive drop. It’s a fascinating decision because it combined a bunch of coaching logical fallacies into one giant crapola soup. Those include:

Sad Falcons Fan

  • SO YOU’RE SAYING THERE’S A CHANCE: Vastly underestimating the actual chances of converting a fourth down, even a long one;
  • THE POINTS DON’T ALWAYS GET TAKEN: Overconfidence in the concept of field goal range and the likelihood of making a kick;
  • ALL WE NEED IS A STOP: Coaches seem to make these decisions with the assumption that the stop is likely. Just look at Smith’s quote;
  • THE DIFFERENCE BETWEEN ONE AND FOUR IS MASSIVE: Coaches overestimate the relative value of being down 1 versus 4, compared to outright having the lead versus still trailing. In reality, the difference between 1 and 4 is not nearly as great as having the lead.

Add all that up, and you get a coach that gives away a chance to take a lead, and opts to try to stay close and hope and pray that a bunch of other stuff goes right.

The Falcons missed the 52-yard field goal and lost.

Related: Fourth Down and Short Situations in 2013: What Do They Show?
Related: Fade Into the Red Zone: A Look at Hot and Cold NFL Pass Plays Near the Goal Line
Related: Coaching Decisions on Fourth Down in the Super Bowl
Related: Ray Rice’s Amazing Fourth Down Conversion Sent The Ravens to an Overtime Win Over the Chargers

blog comments powered by Disqus

Latest Leads

prev.loading
nextloading