By HCR Staff | October 22, 2020
Associated Press’ Josh Dubow reports yesterday that 2-point conversion tries are up significantly this NFL season, with 58 attempts through the first 6 weeks. It’s the highest number of attempts since the rule establishing the 2-point attempt was implemented in 1994.
According to AP, the success rate is 50% thus far this season, and according to ProFootballTalk’s math, when weighing the success rate of this season’s 2-point attempts versus the success rate of the extra point kick, the 2-point attempt yields a statistical advantage of .06.
Now whether you agree that difference is statistically significant is up to you. Many, many arguments can be made on either side.
What we want to get into is the quality of the 2-point attempt data set.
For example, if we’re looking at just this season’s 50% conversion rate, are analytics staffers at teams – who are advising head coaches – weighing every attempt equally?
If they are, then that means they are saying that a 2-point attempt to win the game – as WFT’s Ron Rivera did last Sunday when trailing 20-19 – is the same as a 2-point attempt in a 3 TD blowout with less than 2 minutes left in a game, for example.
Does this sound right to you?
We’re of the mind that the circumstances surrounding every 2-point attempt differ, and that can distort the data. Now one can try to normalize the circumstances and reduce that to a number to plug into an algorithm, but in the end, trying to put a number to shifting momentum in a game, for example, still requires a qualitative judgment. Once qualitative judgments are used, and by different analytics personnel, the data set gets more and more compromised. At least that’s our view.
The 2-point attempt is here to stay, of course, and its created jobs for people who love numbers. But the trick is knowing when to apply numbers to a specific football situation, and when not to.
And many, many teams still have not gotten the trick.