Photo by Perry Knotts/Getty Images
The Super Bowl overtime discourse is missing a key point.
We are now three days removed from Super Bowl 58, and the discourse remains focused on what the San Francisco 49ers did and did not do in the overtime period against the Kansas City Chiefs. The debate is heated over Kyle Shanahan’s decision and both whether or not it was flawed and the fact that the 49ers players might have known entirely what was going on.
That debate will rage all offseason and probably early into the 2024-25 season. And yet, that might not actually be the most important thing that comes out of the wild finish to Super Bowl 58. The NFL might have stumbled across a near perfect overtime system and somehow most people are not talking about that.
The NFL has struggled with overtime rules for years. Sudden death meant you absolutely had to receive the ball to start the extra frame. The league switched systems in 2010 to try and balance it out a bit more. If a team kicked a field goal on the first drive of overtime, the other team got the ball. If any other kind of score happened (touchdown or safety), the game was over. The league implemented this rule for the playoffs starting in 2010 and the regular season in 2012.
In 2022, the league made the latest change to the playoffs, but not the regular season. The new rule gives each team at least one possession in overtime, even if the first team scores a touchdown on the first drive. Additionally, if the first team scores and the second team is still on their own first drive when the first overtime period ends, they continue it in the second overtime period like it’s a new “second quarter” of the playoff game.
There is considerable debate about whether or not the 49ers were correct in receiving the ball. Whatever your view, the fact that there are strong reasons for both sides of the debate is evidence this is a good idea. Even better evidence? The analytics back this up that it’s nearly a coin flip which decision is better.
In March 2022, ESPN analyst Brian Burke put together a thread of tweets breaking down the win probability for electing to kick vs. receive the opening kickoff of overtime. You can read the full thread below, but the tl;dr is that the team receiving the first possession wins 50.29% of the time if we assume the second team doesn’t go for 2 if they match a touchdown. If we assume the second team will go for 2 on a tying touchdown, the first team’s win probability is 50.19%.
Super Bowl 58 was the first playoff game to go into overtime since the new rules were implemented in 2022. While one can make an argument for both kicking and receiving, it’s also fair to note that you might choose one strategy for facing a quarterback like Patrick Mahomes and the opposite strategy for facing a quarterback like Zach Wilson. My colleague Kyle Thele used a poker analogy in a Slack conversation about this topic, stating, Shanahan played the cards and not the person. Analytically it made sense, but with Andy Reid and Patrick Mahomes on the opposite sideline, can you really just go with the numbers and hope for the best?
The biggest question in my mind now is not whether Shanahan was right, but how soon will we see this new system implemented in the regular season? If it follows the 2010/2012 changes, we could see it voted on in the league meetings later this spring. There has been no real discussion about it, likely in part because it had never been seen in real game action. Now that we’ve seen what can happen with it in the highest of stakes, maybe we see the NFL get things right and add these rules for the regular season.
The NFL struggles with a lot of things, whether it be using humans on the sideline to manually check the goal-to-go yardage or refusing to implement instant replay across bigger penalties. But they somehow have managed to figure out arguably the best system they will develop for resolving a game in overtime. There is a discussion to be had about the rule extending playing time, but overtime is not a frequent occurrence, and it’s already extended when the first team scores a field goal.
It’s time for the NFL to extend this into the regular season and put the concerns around overtime to bed.
I simulated the NFL’s new rules for postseason OT. Are they fair? Would a team prefer possession first or second? Should the team with possession second kick an XP or go for 2?…
— Brian Burke (@bburkeESPN) March 30, 2022
Yes, these rules are fair. This is the fairest of any of the overtime rules we’ve seen. The team with 1st possession wins 50.29% of the time, assuming the team with 2nd possession doesn’t go for 2 if they match a TD. And yes, that precision is needed because…
— Brian Burke (@bburkeESPN) March 30, 2022
If the team with 2nd possession goes for 2 subsequent to matching a TD, the team with 1st possession wins 50.19% of the time. Very slight difference.
Bottom line, based on these results, you’d want 1st possession and you’d want to go for 2 on a 2nd possession TD. But the…
— Brian Burke (@bburkeESPN) March 30, 2022
…differences are so slight it would take eons of seasons to actually observe the advantage. I ran 120k sims for each strategy.
Some thoughts…
— Brian Burke (@bburkeESPN) March 30, 2022
There are countervailing advantages for each team. The 1st possession team will have 1 extra drive any time there are an odd number of possessions.
The 2nd possession team has a second-mover advantage in that it knows what is needed to survive on its initial drive. And…
— Brian Burke (@bburkeESPN) March 30, 2022
…it can mitigate the 1st possession team’s first-mover advantage by going for 2, but this effect is tiny.
— Brian Burke (@bburkeESPN) March 30, 2022
The sim isn’t perfect, but it does a very good job of matching actual game win probabilities. And it’s especially useful for questions like this where this is no empirical examples on which to train a model.
— Brian Burke (@bburkeESPN) March 30, 2022
There might be enough bias in the sim that these results are wrong, but the error would be so slight and we’re so close to the point of indifference, it might just be smart to go with the wind and let the opponent decide on possession.
— Brian Burke (@bburkeESPN) March 30, 2022
Lastly, I’m not claiming this settles the question. It’s just one preliminary model. FWIW, the same model gives the old sudden death rules a 57% advantage to 1st possession. And it gives a 53% adv under the last OT rule. If anything it might be underselling 1st-poss a tiny bit.
— Brian Burke (@bburkeESPN) March 30, 2022
