The NFL recently announced that after scoring a touchdown, teams will be required to kick the extra point from the 15 yard line as opposed to the 2 yard line. This is a pretty big change. And whether you’re trying to improve the quality of your process, or simply trying to make a sporting event more exciting, it’s always good to know what kind of effects your change will have. So I’m going to use Minitab Statistical Software to determine what kind of effect (if any) moving the extra point will have on NFL games.
Extra Points: New Location vs. Old Location
Using the game play finder at Pro-Football Reference, I collected every field goal attempt where the ball was snapped from the 15 yard line over the last 15 years (1999-2014). I also collected the number of extra points that were attempted over the same time period. The following is a time series plot of the percentage of kicks that were made from the 15-yard line (new location) and the 2-yard line (old location).
The accuracy of extra points from the 2-yard line has been right around 99% for the last 15 years. Kickers made a lower percentage from the 15-yard line, but it does appear that their accuracy is increasing. The play-by-play data only went back to 1999, but we can look at field goals from the 30- to 39-yard range (so they would be snapped from the 13- to 22-yard line) over the last 50 years.
We can see that kickers are definitely becoming more accurate. So if we want to compare field goal percentages from the 2 yard line and the 15 yard line, we should use data from only the most recent seasons. I totaled the data from the last 5 seasons, and used a 2 proportions test on the proportion of extra points made from the old and new location.
Over the last 5 seasons, NFL kickers have made 94.2% of their field goals from the 15 yard line and 99.4% of their extra points from the 2 yard line. The p-value is 0.003, which is less than the common significance level of 0.05. This means that the difference is statistically significant, and we can conclude that kickers will in fact make a lower percentage of extra points from the 15 yard line. But will a difference of only 5% create effects that we will actually be able to see?
Statistical Significance vs. Practical Significance
Statistical significance itself doesn't imply that your results have practical consequence. With enough observations, sometimes even trivial differences are likely to become significant. For example, suppose you are testing whether the population mean (μ) for hours worked at a manufacturing plant is equal to 8. If the actual value of the population mean is 7 hours and 59 minutes, a large enough sample will reject the null hypothesis and conclude that the population mean is less than 8 hours. However, this difference has no practical importance.
We should apply the same line of thinking to our extra points. Sure, kickers will miss a few more extra points, but will it be enough to notice in a single football game? Last year an NFL team scored 2.3 touchdowns per game. Since there are 2 teams playing, we’ll round up and call it 5 touchdowns in a single NFL game. Using the probabilities above, I calculated the probability of the number of extra points made in an NFL game with 5 touchdowns.
Extra Points Made |
New Location |
Old Location |
0 |
<0.1% |
<0.1% |
1 |
<0.1% |
<0.1% |
2 |
0.2% |
<0.1% |
3 |
2.8% |
<0.1% |
4 |
22.9% |
2.9% |
5 |
74.1% |
97% |
That difference of 5% adds up pretty quickly. Previously, there was only a 3% chance you’d see a missed extra point in an NFL game with 5 touchdowns. That equals about one missed extra point every 33 games. But going forward, there is a 26% chance you’ll see at least one missed extra point. That equals one missed extra point every 4 games! That's an effect that you should be able to see! Of course, there is a much bigger impact this rule change could have.
Kicking the Extra Point vs. Going for Two
The big “practical” change this could result in is two-point conversations. If you make extra points from the 15 yard line 94.2% of the time, your expected value for points is 0.942. So what percentage of two-point conversations would you need to make in order to have the same expected value?
x*2 = 0.942
x = 47.1%
If you successfully convert a two-point conversion 47.1% of the time, you'll score the same amount of points as kicking an extra point. And if you can convert more than 47.1%, then you'll score more points by going for two. According to this article by Advanced NFL Stats’ Brian Burke, the success rate for two-point conversions is 47.9%. So in the long run, you’ll score more points going for two every time instead of kicking the extra point. Awesome! I can’t wait to see coaches going for two after every touchdown this season. It’s going to make football that much more exciting!
On second thought, no it won't.
When given the choice between two solutions, coaches almost always pick the one that is more risk-averse, even at the cost of their team scoring fewer points. What do I mean by risk-averse? Well, let’s consider a game where you score three touchdowns. Actually, let's consider 20,000 different games where you score three touchdowns!
I used Minitab to simulate 20,000 games, half of which you always kick the extra point and half where you always go for two.
Overall, the two-point strategy scored more points than the extra point strategy. But that’s not what coaches actually care about. Look at the left side of the graph: Out of our 10,000 games kicking extra points, only one had our hypothetical kicker miss all 3 attempts. Meanwhile, there were 1,384 games where the team missed all 3 two-point conversions. When it comes to making decisions, NFL coaches want to avoid the worst-case scenario and kicking the extra point is definitely the safer option. The small gain in points they get by going for two is not worth the potential downside of failing.
One last thing to remember from the previous time series plots is that kickers are still becoming more accurate. And with this new rule change, kickers are going to be practicing kicks from the 15 yard line more than they ever have. Over the last 5 years, kickers made about 94.2% of their kicks from the 15 yard line. But it wouldn’t surprise me at all if that number increases this season, making the already tiny difference in expected points between kicking and going for two all but disappear.
So this new rule change will have an effect on the game, but don’t expect much more than a missed extra point every couple of games. And who knows, in a few years kickers maybe become just as accurate at extra points from the 15 yard line as they are now from the 2.
Maybe at that point they should consider making it a 3-point conversion!