Back in November 2006 (I have been writing on the internet for too long!), I wrote that teams that scored 13 points had a better winning percentage than teams that had scored 14 points. That post was in response to an insightful eye from regular commenter Bill M., who first noticed the discrepancy. A few days later, Doug expounded on the topic, and came up with a couple of possible explanations.
As it turns out, this phenomenon still hold. From 1970 to 2017 (postseason included), teams that scored exactly 13 points in a game have a 308-871-10 record, for a winning percentage of .263. Over that same period, teams that scored 14 points went just 221-991-4, which translates to a .183 winning percentage.
Are you wondering if this is a relic of an older era? Well, over the last 10 seasons, teams that scored exactly 13 points went 44-206-2 (.179), while teams that scored 14 points went 26-181-0 (.126). In addition, since 2008, teams that scored 20 points (13 + a touchdown) have a 147-221 record (.399), which is a lot better than teams that scored exactly 21 points (14 + a touchdown), who have an ugly 50-117 mark (.299). The same rule holds even more strongly for 27 points (228-80-2, for a .739 winning percentage) vs. 28 points (88-58, or .603).
So, what gives? Let’s stick to the 13 and 14 point situations and begin with an experiment. There have been 138 games from 1970 to 2017 that ended in a score of 13-10. Knowing that, how many games do you think had a final score of 14-10?
I’ll give you the answer in moment, but let’s start with the obvious note: despite the small number of games that end in a 14-13 score, we know that teams that score exactly 13 points win more games than teams that score exactly 14 points. This means, of course, that teams that score 13 points must hold opponents to under 13 points at a higher rate than teams that score 14 points hold opponents to under 14 points, because #math. It’s true that, on average, teams that score 13 points allow fewer points – for whatever reason – than teams that score 14 points. Does that help inform your guess?
As it turns out, just 52 games have ended in a 14-10 final score. Such a result is so rare that in the last 5 years, it’s happened just once: in a Dolphins/Rams game that was 10-0 with 5 minutes left and 10-7 with one minute remaining. So while it’s a tautology, what’s driving this weird result is that teams score 13 points allow fewer points than teams that score 14 points.
Is there something special about 13 points relative to 14 points that’s driving that result? If so, presumably it’s related to 13 points being most often resulting from a touchdown and two field goals, while 14 points typically coming from two touchdowns. Therefore, I looked at all games where teams scored exactly 13 or 14 points by only one way for each score: two field goals, one touchdown, and an extra point, or two touchdowns with two extra points.
In those cases, from 1970-2016 (I don’t have 2017 data incorporated yet), teams that have kicked two field goals and scored one touchdown (with one extra point) are 291-722-10 in 1,023 games, for a 0.289 winning percentage. Meanwhile, teams that scored two touchdowns and two extra points and nothing else went 213-942-4 in 1,159 games, for a 0.186 winning percentage.
The graph below shows the points allowed by teams in these situations on a percentage basis. For example, when teams score 13 points by one touchdown and two field goals, they have allowed just 10 points about 13% of the time; conversely, teams that score 14 points from 7+7 allow only 10 points about 4% of the time.
Teams that score 13 points are much more likely to allow just 3, 7, or 10 points.
So what’s going on here? Part of it, I think, is that teams down 31-7, they aren’t going to kick a field goal late in the game, but if the score is 10-10, they will attempt a field goal.
What do you think? What other studies would you want to run?