## Desperation Graphed

[Note: I can't remember if I already posted this or not. I did this back over the summer and saved it for the season. But I don't think I ever clicked 'publish.' ]

How desperate do NFL coaches have to be before they start going for it on 4th down?

Recently I've shown that, as a rule, teams should be going for it on 4th down far more often than they currently do. But what is 'currently do?' How far do coaches let themselves get backed up against the wall before they start actually doing what's best for their chances of winning?

It turns out they have to be pretty darn desperate. The graph below plots the percentage of 4th down situations that teams attempt a conversion, broken out by win probability (WP). The lower the WP, the more desperate the situation, the more often teams go for it. (This is an area chart, so the values are additive.)

Teams don't show any detectable change in 4th down doctrine until their WP is below 0.10. By the time their WP is down to 0.05, they're still only going for it 20% of the time! It's not until it's 0.01 or below that teams go for it more often than kick. That is way, way too late.

### 9 Responses to “Desperation Graphed”

1. Neily says:

I'm quite surprised by this. I'd expect a peak somewhere higher up than a WP 0.01. Surely in most cases where the WP is that low, it's a blow out, in which case, even converting the 4th down leaves the team with a tiny WP.

I'd have thought that there'd be an optimum point for going for a 4th down conversion - the game has to be close enough that you can still win (and yes, I know that with a WP of 0.01 you can still *technically* win, but be reasonable) but not so close that an attempt is an unnecessary risk.

2. Anonymous says:

A team down 14-0 with 2 minutes left in FG range has a WP somewhere around ~0.93, but I can't imagine that kicking a FG at that point is a bad decision. Especially if they get the ball first in the second half.

Plus is this just frequency of 4th down attempts that were attempted to convert? It seems strange to lump 4th and 10 downs with 4th and 1 downs.

3. Anonymous says:

If coaches are indeed more concerned with job security than aggressive play calling, then this isn't that surprising. No one will fault them when things seem lost for going for it on fourth. If they're consistently in that position, then their job security is probably low already.

4. Anonymous says:

Thinking about this more, I don't think this graph is really all that helpful: why would a coach go for it on 4th down if it's more likely to *reduce* his chance to win by going for it? But that's included here, since there's no control for how likely the conversion attempt is, nor how much the likely gain is.

A team down 14 at the start of the 3rd quarter, facing 4th and 20 from their own 20 has a WP of only 0.05, but no one in their right mind would suggest that going for it is the right option.

5. Ian says:

The anons make a good point. Part of the 4th down decision making process is weighing up the likely position after the play. For instance, in the situation above me you would never go for the 4th and 20, down 14 on your own 20 early in the third, even though your WP is 0.05. You would, however, go for it every time if you were down by 6 facing 4th and 1 on your opponents 1 yard line with 1 second to go, and your WP is somewhere out around 40-50%.

I think for a true analysis of coaches decision making you'd need to compare the expected WP of going for it versus the expected WP of kicking it (be it a punt or FG attempt). From there, I imagine plotting the go for it% against the difference of the two ought to shed more light on things - so when the expected WP of kicking is greater than the WP of 'going for it' I'd expect coaches overwhelmingly to kick, but on the other side I wonder how much 'going for it' needs to lead 'kick' by in order to tempt coaches to go for it.

6. Brian Burke says:

That's only if there is a considerable difference in the frequency of the various to-go distances based on WP. I'm sure there is some tiny bias somewhere. Bad offenses will tend to be in low WP situations and may tend to have longer distances to go, but I'd bet a fair amount of money the difference doesn't amount to a hill of beans when it comes to the decision to go for it or not.

I just wanted to know where the break-point is, when coaches become desperate. It's just below .10 WP.

7. Anonymous says:

To quote you, from the article:

"How far do coaches let themselves get backed up against the wall before they start actually doing what's best for their chances of winning? It turns out they have to be pretty darn desperate."

Going for it in all 4th down situations is *not* what's best for these coaches in most of these situations. The 4th down situation is relatively independent of WP (a bit, granted there's probably a bit of a bias as you noted) so if a coach is choosing the option which *purely* gives his team an improved WP, you'd expect the graph to be flat. If you plotted frequency of situations where going for it is a better option than kicking a FG or punting, you could verify that.

That means the graph does represent when coaches begin to throw reason out the window in favor of desperation - but that's not what you said it shows.

8. Anonymous says:

How much of the uptick in going for it on 4th down is due to the droppoff in fieldgoal kicking accuracy?

http://www.nytimes.com/2009/12/23/sports/football/23kickers.html

9. Brian Burke says:

Good question...