Earlier today while cleaning up my Favorited Tweets on Twitter, I came across one I’ve been meaning to look into for a while. It was a “TwitPic” of a simple chart, conceived at UCLA in the early 1970s. This is it:
Credited to Dick Vermeil, an assistant coach at the time, versions of the chart have been around longer, but for whatever reason, this is “The One” that still hangs around college and NFL sidelines, albeit perhaps with a few minor changes unique to each coach. It’s fairly evident that, while getting on the right track, there are some serious problems with the chart, which Vermeil himself has long acknowledged (Battista).
Generally, the chart says to go for two when you can make the lead (either yours or your opponent’s) a small multiple of three (FG) or a multiple of seven, also valuing the 17 point lead and the 10 point deficit. Interestingly, the chart suggests going for two when down by one and by nine, something I certainly can’t recall a lot of coaches doing these days. Since 2006, only three times have offenses attempted a two point conversion when down by one, and only once when down by nine. NFL coaches don’t treat the chart like an ironclad law. Have they changed it for the maximum strategic advantage? Not quite.
Coaches still seem a little hazy about how the game clock affects two point attempts. The chart has nothing to say about the amount of time left in the game, but certainly it’s a factor. If it’s the first quarter, even if you can tie the game, why risk it? There are many possessions left for each team. But with under a minute in the 4th quarter, every coach would go for two to tie the game. When does it become necessary? Depends who you ask. In a 2002 Miami-FSU game with 11:44 left in the fourth quarter, FSU kicked the extra point to take a 13 point lead. Miami subsequently scored two touchdowns, holding FSU scoreless to win the game. FSU coach Bobby Bowden said afterward:
I don’t go for two early. If we missed it, a touchdown and two field goals would beat you. (Hutton)
Apparently, 48+ minutes into a game can still be “early”, and Bowden, head coach of a major NCAA Division I football program, was concerned about Miami scoring on three possessions, even though it would only take two to beat FSU. Herman Edwards, former head coach of the New York Jets and Kansas City Chiefs, liked to wait until 12 minutes remained in the game before considering a two point try (Battista). In 2011, then NFL.com analyst Michael Lombardi (now general manager of the Cleveland Browns) expressed his own thoughts:
Continue to add points. Do it until the amount of possessions remaining is dwindling. The discussion to go for two should never occur until the fourth quarter, when possessions are limited. (Lombardi)
It turns out that even before the 4th quarter, going for two can still be the strategic play. Harold Sackrowitz, a statistics professor at Rutgers University, is as expert as anyone on two point conversions, having spent years studying them and tinkering with probability models. He accounts for potential future possessions, the time left, the score, and even the probability of converting given the strengths and weaknesses of the offense and defense. And with all of that in mind, he knows teams should start considering to go for it in the middle of the third quarter (Battista).
How have two pointers fared recently? Using Pro Football Reference’s Play Finder, one can see that of the last 500 two point conversions attempted in the regular season (as many as they’ll show at once, going back to November 2004), 49.2% (246) succeeded while 50.8% (254) failed. Of the 31 attempts so far in 2013, 45.2% (14) have succeeded while 54.8% have failed. Actually I was a little discouraged by those numbers. 500 is a pretty fair sample, and with a 49.2% success rate, I suspect it yields the same number of expected points as an extra point (due to its much higher, still not 100% success rate). Are teams just going for two in the wrong situation or the wrong time, or are they going for two in the wrong situation, or the wrong time, and still not going for it enough? Mr. Sackrowitz thinks yes, and that teams should go for it more. Vermeil, now retired after a renowned coaching career, agrees:
I don’t think it’s a situation you can perfect. I do say this: I believe it’s probably you don’t go for two points often enough. Now that I sit and watch on TV and don’t have that pressure, I can say, “You guys are chicken.” (Battista)
As time goes on, it will be interesting to see if they are proven right.