If the NFL moves the Extra Point, this is why smart teams will stop kicking them entirely…
Shout out to the pro-football-reference.com database for the stats in this article. You guys rock.
So we’ve all heard the rumblings from the NFL’s competition committee recently: “The extra point is sort of boring. Let’s screw with it!” seems to be the general gist of the conversation.
The Commish himself has gone on record about his distaste for the extra point. “I believe we had five missed extra points this year out of 1,200 some odd. So it’s a very small fraction of the play, and you want to add excitement with every play.”
The gripe might be legitimate. In the last ten season, NFL kickers have completed the extra point attempt a whopping 98.9% of the time (11058 attempts). In 2013, the success rate was 99.6% (1267 attempts). This is as close to automatic as it is going to get in the NFL. The problem, of course, is that there are still 21 players besides the kicker on the field, so major injuries can still occur. Rob Gronkowski, the Pats’ ginormous, extraordinarily gifted tight end, famously broke his forearm on a simple extra-point attempt. While that was a highly visible incident, there are undoubtedly countless, unreported minor traumas that occur during extra point attempts every season. The logic is that the 1.1% of kicks that aren’t made are outweighed in importance by the danger of the play.
Goodell intimated that there are several proposals to change the rule, including one where the point is automatically given after the touchdown unless the offense wants to attempt a two-point conversion. But football purists are certain to be appalled by the idea of taking another step to remove feet from football, so this complete removal of the kick attempt would be a fairly radical move to make in one fell swoop.
So the NFL competition committee is floating a compromise idea: Move extra point kick attempts to the 25 yard line, making the kick a 42 yard attempt rather than its current 19. This solution makes some sense on its face. Teams are still allowed the option of attempting a slightly-riskier but still very makeable extra point, or they can go for the two-point play from the regular spot two yards out. This presumably would leave the strategy of the game intact but provide for a more interesting post-touchdown play.
But this proposal has a major flaw: Any team that chooses to kick in that scenario is run by stupid people. Here’s why:
If they move the extra point kick back pretty much at all, statistically minded coaches will stop attempting it almost completely. There’s a graph above that says so! The yellow line is how much extra points would be “worth” on average from a given yardage, and the horizontal lines are multiple approximations for the expected points of any regular two-point conversion attempt. Extra Point attempts start being worth less than two-point conversions if they are attempted from 15-16 yards out, so if the extra point were moved to the 25, smart teams would go for two almost every time.
I’ll admit the above graph doesn’t have the best labels in the world, so please allow me to elaborate. Explanation after the jump.
This graph accounts for all NFL regular season games in the last ten seasons (2004-2013), and it compares the expected values of an extra point attempt from varying distances to the expected value of a two-point conversion. Expected Points are logged on the vertical axis, and the location of the line of scrimmage during a field goal attempt is logged on the horizontal.
Expected Points (EP) are a simple metric. An extra point is worth 1, so if they are completed 98.9% of the time over a ten year span, it stands to reason that you can expect to score .989 points on average per attempt. Thus the EP for an extra point is… you guessed it, .989. So the yellow line on the graph represents the EP of an extra point attempt taken from a range of yardages, based on the NFL field goal percentage from each spot over the last decade. Granted, you can argue that the extra point is always taken from the center of the hashmarks and this data includes all field goals irrespective of relative location within the hashes. Still, it seems to provide a reasonable approximation of success rate for kicks, and this is my website, so I’ll allow it.
Calculating EP for two-point conversions is trickier. See, over the past ten seasons NFL teams have converted 267 of 534 two-point attempts, good for a nice round 50%. On this basis, one could argue that the EP for two-point conversions should be 1. But given the relative rarity of this play compared to kicks (only 4.6% of all touchdowns were followed by a two-point try), it seemed to me that this success rate, and thus the EP of a two-point attempt, could drastically change if the play were called more often.
Two-point conversions are most frequently attempted in special circumstances: extremely tight games and blowouts. As such, the success rate might be inflated because the offense is scoring garbage-time points against backups in a feeble attempt to regain some scorecard dignity, or it could be inflated because teams are using only their best plays for a “must have it” situation. If two-point conversion attempts became more common, which is likely to happen if the extra point were made more difficult, the conversions might not continue to succeed at quite a high rate. I thought it better to dig a bit deeper and expand my sample.
So I widened my search to include not just two-point attempts, but all offensive plays run from the two-yard line in goal-to-go situations. As it turns out, of all plays that start at the two yard line, only about 44% of them result with the ball in the end zone (This is represented by the thick, orange horizontal line on the graph entitled “TOTAL”). The conversion rate varied by down: 1st down two-yard attempts scored 41.5% of the time (green line), 2nd down 40% (red line), 3rd down 45% (purple line), and 4th down came in at 50%, the same as the conversion rate for two-pointers under the current rules (blue line). I have no way of verifying this, but the data seems to fit well with my “must have it” play call hypothesis. But regardless of the cause, it at least confirms that there is reason to believe that if two-point conversions were attempted more often, there is a good chance the success rate would diminish. I suspect that if two-point conversions were attempted almost every time, the success rate would probably regress to the success rate for the total of all plays run from the two-yard line in goal go go situations (TOTAL).
So on the graph, the yardages where the EP for extra point attempts (yellow line) dip below the horizontal lines are the places in which an extra point attempt is worth fewer EP than a two-point conversion run from the two. Assuming one accepts my conjecture that the orange line is the best representation of future two-point conversion rate, taking an extra point attempt from the 25 yard line is worth .76 EP, while a two-point conversion is worth .88 EP. If a team scores 40 touchdowns per season (which is roughly the output of most middle-of-the-pack offenses), going for two every time would theoretically add 4.8 points to that team’s total score per season than if it went for only extra points. That may not seem like much, but 14 regular-season games in 2013 were decided by one point, so that figure really cannot be discounted. Likewise, let’s say my conjecture that two-point conversion rate will drop is wrong. Let’s say teams keep making it around 50% of the time (blue line). If that were the case, that same 40-TD offense would stand to gain 9.6 points over the course of the season.
As it stands today, both two-point conversions and extra points have roughly the same EP. Right now, if the current success rates were to hold steady, one would have to go for two 100 times to see a 1.1 point benefit. Frankly, that return is not worth the trouble. But if the league is set on moving the extra point back, the 25 yard line is too far. Many teams would likely engage in a long-term strategy to accrue more points by taking the immediately riskier two-point option. It seems to me that if the kick must move back, it should go to around the 15 yard line. Assuming that two-point conversions are attempted more often and their success rate falls to around 44% (.88 EP), then kicks would be worth about the same and the strategy of the game would remain fundamentally intact. If they move it back much further, maybe Roger will get his wish and we’ll see the extra point eliminated by default, only to be broken out in strategically important moments when one point will do just as well as two.