The NFL has an overtime problem, or so everyone seems to think. Every year we get long screeds from columnists lamenting the unfairness of sudden death overtime. Why? Some people think it’s unfair that one team can get the ball and score, without the other team’s offense ever taking the field. Others point to the fact that the team that wins the toss wins the game 25% more often than the team that loses the toss. And recently, the NFL has taken some action, changing the rules for overtime in the postseason:

Under new NFL overtime rules, a coin toss still determines which team gets possession of the ball first in the overtime period. However, if a team scores only a field goal on the ensuing possession, the opposing team has the opportunity to possess the ball. Thus “sudden death” applies only to touchdowns, not field goals. If the score is still tied after 15 minutes, play continues until a score is made and a winner is determined.

Okay, a few things here. First: the new overtime rule for the playoffs seems overly complicated and doesn’t completely solve the problem of only one team touching the ball. Second, in the regular season, what’s so bad about ending a game in a tie? Why not eliminate overtime completely in the regular season? Well … that’s not going to happen. We hate ties here in America. So with that in mind, here are my two cents about “fixing” overtime.

Consider this: why do teams take the ball first in overtime? Is it because they think they are going to score on their first possession? Well, not necessarily, though that does happen a healthy chunk of the time (30+%). No, they want the ball first because they feel that will put them in the best position to score at some point during overtime. (Mathematically, we can say that this move has a positive expectation.)

But why do they feel this way? Two words: field position. I can’t find what the average starting field position is for teams after a kickoff (little help?) but let’s say it’s the 27 yard line*. Football teams intuitively feel they are more likely to score if they start on the 27 than their opponent, and this intuitive view matches up well with the numbers. According to the famous Berkeley paper on the topic, starting with the ball on the 27 yard line is worth 0.71 points (on average) to the team with possession. That doesn’t sound like a lot, but it’s a positive number, meaning that team is more likely to score than their opponents.

* It’s probably lower than that now with the new kickoff rule, but again, I don’t have any numbers here.

So how do we fix this? Easy: just eliminate the opening kickoff in overtime and give the ball to one of the teams on a pre-determined yard line. Now what should that be? The ideal place to start the ball is one that is worth 0 points to the team with possession; this means that the team that starts on defense is as likely to score as the team on offense. If we go back to the Berkeley paper, we find that the 15 yard line is the place to start. Intuitively, this again makes sense — a coach would have to think pretty hard before simply taking the ball to start overtime if he knew he would start from his own 15. I suspect it would happen more often than not, but certainly a lot less often than it happens now. And if the math is right, you would see the team that started with the ball in overtime win a lot less often than they do now.