## Bucs Botch Fourth Down Call

With 1:10 remaining in the fourth quarter, Tampa Bay led 14-13 on the Saints' 29-yard line. Facing a 4th-and-3, with an 82% chance to win the game, Greg Schiano elected to attempt the 46-yard field goal - about a 64% proposition. If the Bucs hit the field goal, they go up four points, increasing their win probability to 86%. With a miss, the win probability drops to 74% and Drew Brees gets the ball on the 36, needing only a field goal to win.
Schiano also had two other options: punt the ball and pin the Saints deep or go for it. Below are the numbers:

There are two caveats to these baseline numbers. The Saints had a tough time offensively all game in the sloppy conditions, but keep in mind, their offense is still far above average. This means that in any situation where the Saints get the ball back, Tampa Bay's win probability is, in all likelihood, lower than the baselines above. On the other hand, the Saints had no timeouts left and the win probability calculator is independent of timeouts (it assumes an average number of timeouts since it comes directly from historical data).

Let's look at the punt first. A punt from the 29 is expected to net about 14-15 yards, giving a league-average offense an 11% to drive down and score with one minute left. This is actually a higher win probability for the Bucs than if they had successfully converted the field goal. This seems odd given that being up one and being up four is a huge difference, but there are two reasons. First, Brees and company would be expected to have much better field position coming off a kickoff rather than the punt. Second, being down four forces the Saints to be more aggressive offensively as they know they need to score a touchdown, rather than settle for a field goal. This is a similar paradox to what we wrote about last week in the Saints game. More aggressive in the NFL typically means more efficient.

Last, Schiano could have elected to try to convert on 4th-and-3. League-wide, teams convert such an attempt at 57%. If the Bucs convert, the game is over as the Saints had no timeouts left. If they fail, the Saints are in a slightly worse situation than a missed field goal, just due to field position. Overall, both going for it and punting are superior options to a field goal, by about 6-7%; I would lean toward going for it given Brees' offensive prowess. The Bucs also had a 2nd and 3rd-and-6 leading up to the 4th-and-3, so they would have the option to game-plan around going for it on fourth down as well, for whatever that's worth.

In the end, Rian Lindell missed the field goal, Brees quickly drove down the field for a chip-shot 27-yard game-winning field goal. We can't evaluate the decision based on the negative outcome of the missed field goal. But, the win probabilities and probabilities of converting both point toward the field goal as the least desirable choice.

Keith Goldner is the Chief Analyst at numberFire.com - The leading fantasy sports analytics platform - and creator of Drive-By Football.  Follow him on twitter @drivebyfootball or check out numberFire on Facebook

### 13 Responses to “Bucs Botch Fourth Down Call”

1. Jeff Clarke says:

I was at this game. The Bucs left the offense on the field at first. They were just running down the clock and calling timeout with one second on the playclock. Still, nearly everyone in my section was aghast at the possibility that they just might go for it. A field goal was so "obviously" the better call there.

I started to think about political debates where neither candidate is willing to admit to believing anything that is unpopular even if it is demonstrably true.

For the first time, I had some pity for the coaches as politicians. I honestly think Raymond James might have rioted if the Bucs went for it and lost the game. Belichick did it on the road and Schiano is no Belichick.

2. Steve Freeman says:

Why is the win probability the same 74% for missed first down as for a missed field goal?

In the missed field goal case the opponent gets the ball on the 36; if they pass or run, it's around the 29, depending on how many yards, if any, are gained or lost in the failed first down attempt, but still on average, it's an additional seven yards.

3. SportsGuy says:

Some assumptions are being made here on what kind of field position the Saints could expect on the kickoff after the made field goal. How good is the TB kickoff guy? How good is the Saints' return unit?

I'm a Lions' fan. I know that if my team were in the same spot the FG would look a bit better because there is almost zero chance anyone is getting return yards off a Sam Martin kickoff unless he is facing a good stiff headwind.

4. Anonymous says:

Im shocked that Brian hasnt posted an article yet on Ron Rivera's 4th down decision.

5. Mike says:

The win probability gain due to punting really surprises me here. Pushing the offense back 14-15 yards drops their win probability by 15%?

This has to be the only time the numbers would support punting from your opponent's 29.

6. Anonymous says:

Mike,

I questioned JG's decision to kick a field goal down by 4 on KC's 35 with about 4min left. I believe the calculator (if I read it right) said that punting was the best option, followed by going for it, and last was Dallas's actual choice, FG.

-ChrisB

7. Anonymous says:

Vikings totally botched 4th down at end of game, 15% win probability decrease by kicking FG.

8. Anonymous says:

What is the percentage of times the ball would go in for a touchback punting from inside the 30? How does a touchback on a punt change the the decision on what should have been done?

9. Ian Simcox says:

Punt and go for it and practically interchangeable - this a perfect situation for a quick kick call.

The beauty of that is that if the defenders stay in normal formation you pooch kick it over the top, chances of a touchback are massively reduced and ultimately you give them the ball far nearer their own goal line. If a defender stays back expecting a pooch kick then you run a play, 11-v-10.

I know we always say it, but why are coaches so conventional when they get to these situations? Do they all seriously never run through different approaches to end game situations as part of their weekly gameplanning talks?

10. Jeff Clarke says:

I never even thought of the pooch kick. The problem with that is that you are asking your QB to kick the ball, which he might or might not know how to do. It seems kind of risky for the circumstances.

I think that the gains from punting might be overstated in this situation. Isn't part of the reason why punts appear to be a relatively good call based on the presumed overconservativeness of coaches pinned deep in their own territory. Its a given the Saints would have come out in an aggressive mode in these particular circumstances.

11. Anonymous says:

or go for a hard count try to get penalty or fake punt or fake field goal

12. Joke says:

@Ian Simcox,

I think the reason coaches don't break orthodoxy as often as they should is about where their incentives lie. They're not actually trying to maximize winning, they're trying to maximize fan and owner happiness.

Even if an unorthodox play increases the chances of winning from, say, 70% to 75%, the coach may not see the move as in his best interest. If the 25% negative outcome is the one that hits, then the coach gets crucified for making a move the fans don't understand.

You could be standing next to the coach and lay out the probabilities for him, and even if he fully believed the analysis, his perception of what's in _his_ best interest could lead him to decide "Well, whether the odds are 70% or 75%, we're probably going to win this. If I make the typical move and it fails, fans are mad because of the loss. If I make the unorthodox move and it fails, fans are screaming for my head."

If the disparity were, say, 25% rather than the 5% of my example, I'm pretty sure all but the dumbest coaches would go for the unorthodox play. But if it were only, say, 1% I'd think only the most secure and/or ornery coaches would go unorthodox.

13. Joke says:

But as @Steve Freeman points out, I'm not sure there wasn't a small error in this analysis.

A failed FG put the Saints on the 36, and a failed go-for-it puts them on the 29. The WP used for both those cases is 74%. Compare that to the ("successful") punt scenario, which the author implies is expected to put the Saints at the 15. There the WP jumps all the way to 89%. (In all 3 cases, the Bucs' lead is the same -- 1 point).

It seems odd that there would be 74% probability from the 36, 74% from the 29, but 89% from the 15.