Who Dat Report

Here’s each team’s average points scored and allowed including all years from 1978-2012. Data from repole.com and cleaned up by me.

Administrative notes: I marked the team by their current location, so STL includes the Rams games when they were in LA. CLE includes all of the Browns data, BAL is just the Ravens stuff. Also: I had forgotten about the Tennessee Oilers.

Other notes: Pittsburgh’s defense is hurt by the fact that the data starts well into the Steel Curtain era. Tampa and Cleveland have been historically bad on offense and kind of average on defense. Arizona has just been historically bad.

Here’s another plot using DVOA data from Football Outsiders and salary data from The Guardian. This time, I plotted offensive and defensive spending and wins.

Note that The Guardian’s salary data is only approximate, and other sites have different measures of how much teams spent. Oakland’s spending, for example, had to have been a bit higher than this, right? So take this as roughly right, which is good enough for our purposes.

2 interesting things:

(1) five of the seven teams that fired their coach (represented in the graph with a *) spent above league average on offense.

(2) Of course Bill Belichick can take a team with league average salaries and coach them to the AFC Championship game. Of course.

I updated the 2014 offensive and defensive cost-efficiency graphs to include a third variable: number of wins. To do this, I had to switch from Stata to R, which is good, because the whole purpose of this is to help me improve my skills with these programs. Anyway, number of wins is represented by the color of the marker for each team, with darker blues meaning more wins. I also added an asterisk next to teams that fired their coach. (UPDATE: I forgot to put an asterisk by Tennessee, who did fire their coach. Whoops.)

Here’s the offensive graph. Again, DVOA from Football Outsiders and salary data from The Guardian.

Next, defensive. Remember with defense, the LOWER the DVOA, the better, so teams in the lower left quadrant were the most cost efficient.

Interesting that bad, expensive offense seems more likely to get a coach fired than bad, expensive defense. If I can get historical salary data and more free time, that might be worth exploring.

Here’s a scatterplot comparing 2014 defensive expenditures to defensive DVOA. Remember that a lower defensive DVOA is better, so the lower-left quadrant is the most cost-efficient part of this graph. I (somewhat arbitrarily) highlighted the most efficient and least efficient teams.

Looking at this graph, Carolina’s defense was probably the most efficient in the league in terms of production per dollar, probably thanks to young, cheap talent like Luke Kuechly. As for the Saints, they spent a bit above average and were a bit above average, about like you’d expect.

Also notable: how much better Seattle was than the rest of the league, presumably because they are a bunch of cheating cheaters.

DVOA data from Football Outsiders and salary data from The Guardian.

For grins, I plotted offensive DVOA (Football Outsiders’ opponent-adjusted measure of offensive efficiency) vs salary spent on offense. As expected, the teams that spent more money generally tended to have higher offensive DVOAs, though there were plenty of exceptions. The upper-left corner is the “super efficient” quadrant: teams that got above average offenses despite spending below average money. The lower right corner is the “pack your bags, coach” quadrant, where teams spent above average $$$ and got below average output. Note that 4 of the 6 teams in that quadrant fired their coaches, and the Giants likely would have if Coughlin wasn’t a multiple Super Bowl winner.

The Saints weren’t particularly cost-effective, but they were good on offense. They spent a lot and got a lot, which helped them reach the second round of the playoffs. Not bad.

DVOA data from Football Outsiders, salary data from The Guardian, of all places.

I’m working on some visualizations of the Saints’ season in my spare time. Which I don’t have any of, so this could be the last one. Anyway.

Here’s the Saints’ 2014 offensive and defensive DVOA rankings by week. Since it’s the ranking, the lower the better. All data from Football Outsiders. Since this is DVOA, the ratings are a bit flaky in the beginning of the year before becoming more reliable later. The Saints were fairly stable after about week 12, though the offense got a little worse and the defense a little better.

*assuming, of course, that the defense would have prevented the Packers from scoring

A few quick thoughts on Hartley and kickers*:

*Which I’ve updated about 100 times, far too many times to do the cute Internet strikethrough thingy, sorry. I shouldn’t write in a pique of post-loss fury and beer, but that’s life.

1) Remember, it’s really tough to evaluate how good kickers are. The sample size is just too small to make meaningful judgments. Garret Hartley has attempted 57 field goals in his 31-game career. It takes a baseball player about 15 games to have 57 plate appearances. No one would say that we should judge a baseball player based on 15 games, but it’s common to think that we can judge a kicker on a similar sample. We can’t.

2) Though we don’t really know if Hartley’s a good kicker or a bad kicker, many fans have impressions of how reliable he is. Notably, these impressions may differ in 2012 from what they were in 2009, when he kicked the Saints into the Super Bowl.

3) The difficult thing is this: it will take Hartley years to have attempted enough kicks to get a decent sense for his skill. Morten Andersen kicked 389 field goals for the Saints—the equivalent of about 100 baseball games’ worth of at bats— in 13 years. 13 years! As Jerry Glanville explained, NFL coaches don’t have that kind of time. Football teams don’t have the luxury of riding out slumps and unlucky streaks over a 162-game season, which is why a kicker’s life is lonesome.

4) Kickers may be getting better and better, but it’s still a tough job. It gets tougher from 40+ yards, and even tougher with the game on the line. What many people think of as a “gimme” isn’t exactly that. Let’s look at the numbers:

In 2010 and 2011, NFL kickers attempted 584 field goals between 40 and 49 yards. They made 430 of them, or just under 74%. But, thanks to the holding penalty, Hartley didn’t have to make one high-pressure field goal, he had to make two. If he’d missed either of them, the Packers would get the ball. The average odds of making two consecutive field goals between 40 and 49 yards are lower: 0.74^2, or about 55%.

But that overstates Hartley’s odds because his attempts weren’t 40 yards, they were 43 and 48 yards. Field goals get harder to make with each additional yard of distance. Unfortunately, the data at Pro Football Reference isn’t fine-grained enough to make these distinctions, so we can have to make some assumptions until I have time to dig around and find better data.

First, the 43-yard field goal. It’s reasonable to assume that the field goal data are weighted toward easier-to-make field goals, that is, that there were more 40-yarders attempted than 41-yarders, more 41-yarders than 42-yarders, etc. Since the average accuracy from 40–49 yards was about 74%, it’s also reasonable to assume that the accuracy from 40 yards was considerably higher than average. However, what about the accuracy from 43 yards? If I’m right that the data are weighted toward easier-to-make field goals, then my guess is that the accuracy from 43 yards is pretty close to the overall average, or 74%. So let’s stick with a 74% chance of making the 43-yard field goal.

What about the 48-yard field goal? 48 is a lot closer to 50 yards than 40 yards, so it might be useful to look at how the NFL did on 50-yarders for comparison’s sake.

In 2010 and 2011, NFL kickers attempted 248 field goals from 50+ yards, making 149 (60.1%) of them. This figure includes all field goals over 50 yards, and I’d be willing to bet that the accuracy drops off quickly with each yard over 50. So let’s assume that the chance of making the 48-yard field goal was a bit less than the 74% overall chance for 40–49 yard field goals and a fair amount greater than the 60.1% chance of making a 50+ yarder. Call it a 70%.

If you buy my assumptions, the chance of making both of the field goals was about 0.74*0.70 = 51.8%. If you just look at the straight averages, the chance of making both was about 55%. Either way, that’s about a coin flip before you account for the pressure of the situation, the outdoor venue, the screwy timing and focus issues that happen when you have your kick interrupted several times, etc.

5) Again, Hartley had to make both field goals for the Saints to get the 3 points, because the Packers could have declined the holding penalty if he’d missed the first attempt. So the coin flip/ 50-50 chance is the most relevant analysis. However, it’s interesting to look at the probabilities of all the possible outcomes:

• There was about an 8% chance that he’d miss both field goals (0.26*0.3)
• There was a ~22% chance that he’d make the first and miss the second (0.74*.3)
• There was a ~18% chance that he’d miss the first and make the second (0.26*0.7
• Finally, as mentioned above, about a 52% chance that he’d make them both

Those percentages assume that the outdoor/pressure/screwed up timing conditions didn’t make the kicks harder, which I believe is a false assumption. Either way, what actually happened (make the first and miss the second) was the second-most likely outcome of the two kicks.

6) It comes down to this: you can blame Hartley all you want for missing the kick, but given the success rate of NFL kickers on similar field goals over the last couple of years and the difficulty of the situation, it would have been more surprising if he’d made both of them. Unfortunately for the Saints, that’s what he had to do.