The End-Game in BAL-PIT

There were two critical decisions at the end of the Ravens-Steelers game Sunday. The first was a 4th down call, and the second...was also a 4th down call, but not the kind I usually harp on.

Down by 4 with 2:44 left in the game, John Harbaugh made the right call to go for the touchdown on 4th and goal from the 2. The fade pass to Anquan Boldin was unsuccessful (are they ever?), but Baltimore went on to win anyway. I won’t bore you with the actual numbers, but this is a perfect example of why going for it is usually the better call. Even when the play fails, it’s far from a death sentence.

The second and more interesting 4th down call came with 1:15 left. The Steelers faced a 4th and 10 from their own 3. An NFL statistician-emeritus asked me if the right call would have been to take an intentional safety. My first reaction was, ‘of course not.’ That cuts the lead to 2 points, making the lead vulnerable to a FG. I thought, what’s harder to do, drive 40 yards for a TD or drive 25 yards for a FG? Little did I know I was falling into the same trap many coaches do.

It turns out that taking the safety might just be the better play. Setting aside the possibility of a blocked punt or long kick return, the numbers are very clear. They were so surprising, I had to go back to the raw percentages just to make sure.

First, let’s look at the punt. Punts from the 3 typically net 38 yds. An offense down by 4 with the ball on the 41 yd-line and 1:10 left has a 0.43 WP. That’s a 0.57 WP for the Steelers.

An intentional safety cuts the lead to 2 points and, following the free kick, would typically give Baltimore the ball at their own 40. This equates to a 0.37 WP for the Ravens, and a 0.63 WP for the Steelers.

I thought there’s no way this can be right. But then I thought how would an NFL coach handle being down by 2 points with a minute to play. He’d play for the “___?___.” As soon as his team entered FG range, his team would (very predictably) run the ball once or twice, spike the ball and hurry on the kicking squad.

Think of it this way, what’s harder, moving 40 yards from your own 40 to the opponent’s 20 for a FG,  or moving 40 yards from the opponent’s 40 into the end zone? They’d be about equally difficult, with the latter being slightly tougher due to the short field in the red zone.

But we’re forgetting something: Field goals are not automatic.

From the 20, FG attempts are successful a little more than 80% of the time. Once that’s taken into account, driving from the 40 into the end zone is the better bet.

Besides, offenses would be very likely to stop well short of the 20. A drive to the 35-yd line would allow a 52-yd FG attempt, which would be successful only about 55% of the time. My hunch is that coaches are suckered into the lure of FG range. If a team gets into range, that’s considered a success, at least for the coaches, because if the kicker misses that’s his own fault.

But FG range is a concept defined by anything but certainty. When the possibility of a miss is considered, a team is better off having to drive the full 40 yds for a TD. The bottom line can be distilled into a simple set of examples: Would you rather move the ball 40 yds for an 82% chance at a win? How about move 25 yards for a 55% chance for the win? How about 30 yards for a 66% chance? What about 40 yards for a 100% chance?

It's worth repeating: Field goals are not automatic.

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

19 Responses to “The End-Game in BAL-PIT”

  1. Kos says:

    "They’d be about equally difficult, with the latter being slightly tougher due to the short field in the red zone."

    Is it really only slightly tougher? I'm only using the eyeball test, but it seems like scoring from 10 yards out is tougher than moving the ball 40 yards late in a game (especially with the prevent-you-from-winning defense coaches use outside of the 20s). I'm not saying the math or your general conclusion is wrong, but I'm guessing that "slightly tougher" isn't accurate based on the ease with which teams move the football outside of the 20s in a two minute drill.

  2. Ian says:

    This is all TV's fault. When you get to the end of a game, they start drawing on that red-line, denoting the kickers career long FG - as though somehow he's certain to score the second you go past the line.

    Defenses do seem to fall foul of the same fallacy though. How many times have you seen teams defending 3rd and 5, say, on about their 35 yard line giving up a long gain because they were trying not to give up even a single yard more? The important thing at that range is to make them kick - not to not allow them to get any closer.

  3. Jeff Clarke says:

    A week after the Belichick game last year, Houston was in this exact situation. They got the ball to the very outer edge of FG range. Then with plenty of time and a timeout left for one more play of 10-15 yards, Kubiak called an extremely conservative run up the middle for no gain. The kicker missed and obviously Kubiak was able to blame the whole thing on the kicker with the "I thought he'd make it and I didn't want to risk a turnover". The bottom line is the potential extra yards were worth far more than the risk of an int.

    After such a media firestorm over Belichick, hardly anybody noted Kubiak's mistake.

    Coach aggressively and lose---its your fault.

    Coach conservatively and lose---there's always a convenient scapegoat.

  4. Anonymous says:

    Not to mention that kicking a FG in Heinz Field is tough and the Ravens have a bad FG kicker.

  5. Anonymous says:

    Can your WP stats tell if running up the middle on 3rd and 10 was a better decision than trying to pass for the first down? Some friends of mine have been debating that.

  6. Anonymous says:

    "What about 40 yards for a 100% chance?" You're forgetting the possibility of an early touchdown, giving the steelers the ball back with time to score. Of course, your WP takes that into account and I still agree with everything you say.

  7. Anonymous says:

    Brian,

    You are ignoring the relative strengths of the teams here. The Steelers just stopped the Ravens at the goal line. They are playing excellent red zone defense. The Ravens RB were dinged up and Joe Flacco has a history of not doing well in the red zone. It is much easier to defend against the TD then defend against a FG.

    The steelers called a blitz at the same time the Ravens called for mass protect. Housh made a nice fake. Things happen. You've got to trust a defense of the Steelers caliber in the red zone.

  8. JMM says:

    "You are ignoring the relative strengths of the teams here. The Steelers just stopped the Ravens at the goal line. They are playing excellent red zone defense. The Ravens RB were dinged up and Joe Flacco has a history of not doing well in the red zone. It is much easier to defend against the TD then defend against a FG."

    All of that also applies to the take the safety option as well. And yes, some times things happen. A coach can still play the odds.

  9. Anonymous says:

    "You've got to trust a defense of the Steelers caliber in the red zone."

    To me, you can trust the very same Steelers defense to prevent the Ravens from getting into field goal range.

  10. Anonymous says:

    Brian, can you do an analysis on Jim Caldwell's timeout with 36 seconds left? I want to know what are the odds the Jaguars score from there and the odds the Colts stop them twice, get the ball back and score a FG in a short period of time.

  11. Anonymous says:

    "To me, you can trust the very same Steelers defense to prevent the Ravens from getting into field goal range."

    Red zone D and between the 20s D is very different. The Ravens moved it between the 20s all game. They had a good O-line that day and three top receiver threats. You can't stop em between the 20s. Get them down near the goaline where its easy to cover all those targets.

  12. foodmetaphors says:

    I hope the next time Hartley kicks they put the red line at the 10. Also, somehow notate the need to center the ball when you have first and 10 from the 11 with plenty a'time left.

  13. Joseph H. says:

    foodmetaphors, you'd better be a Saints fan--cause after the pass to Henderson, I had the same thought. At LEAST, they could have run twice before kicking on 3rd down. A tiny point in Coach Payton's favor though, is that Pierre Thomas and Drew Brees had just got nicked up, and he thought "May as well get this over with and get out of here."

  14. Anonymous says:

    "An offense down by 4 with the ball on the 41 yd-line and 1:10 left has a 0.43 WP. That’s a 0.57 WP for the Steelers."

    "Baltimore the ball at their own 40. This equates to a 0.37 WP for the Ravens, and a 0.63 WP for the Steelers. "

    Do these numbers change in an expected low-scoring game like BAL-PIT? I expect the WP for the Steelers to be higher than .57 and .63 in the two respective situations since BAL should be less likely to score against PIT than the average NFL team against another average NFL team. But is it possible the .57 goes up more than the .63 in a situation like this? I do not ask with a preconceived notion.

  15. Andy says:

    So this analysis only works cause coaches will make the right decisions on offense going for the touchdown as the clock expires and the wrong decisions going for the field goal. What will happen to this site if half the coaches start making the right calls all the time? That might be a really important variable, but its all circular and crazy.

  16. Kevin says:

    Brian; I also asked you about the possiblility of Buffalo taking a safety with 4th&10 from their own one yard line down 13-10 to Miami in Week#1 with two minutes left.

  17. Anonymous says:

    For 4th downs: the best strategy would be to invest in a QB with punting skills (or punter with a deepish pass, or perhaps running ability) kind of a Vick with punting skills. Then watch the opposition figure it out on 4th downs.

    I'd enjoy that. :)

  18. roseyf16 says:

    Hi Brian,

    The ballyhood Pittsburgh defense correctly proved what the "prevent" defense does for teams trying to protect a lead at the end of a game: It prevents you from winning!

    Keith Rosenkranz

  19. Brian Burke says:

    Hi Keith!

Leave a Reply

Note: Only a member of this blog may post a comment.