Prior to 2008, the opening coin toss in an NFL game was governed by the following rule:
The winner may choose one of two privileges and the loser gets the other:Immediately prior to the start of the second half, the captains of both teams must inform the officials of their respective choices. The loser of the original coin toss gets first choice.
- (a)Receive or kick,
- (b)Goal his team will defend.
Under this scheme, the winner of the coin toss did not actually have much of a choice. If they chose anything other than "receive", it would give the other team the opportunity to receive the ball at the beginning of the game and at beginning of the 2nd half. This explains why in 99.6% of regular-season coin tosses from 1999 to 2007 the winner of the coin toss chose to receive the opening kickoff.
The following is a list of the 8 games from 1999 to 2007 where the team winning the coin toss did not choose to receive the opening kickoff. In 7 of 8 of those cases, the team that lost the coin toss was able to receive the ball at the start of both halves of the game.[1]
| Game | Wind | Toss decision | Kickoffs |
|---|---|---|---|
| Nov 14, 1999, Vikings @ Bears | N 16 mph | Vikings elect to defend north goal | Bears receive twice |
| Dec 5, 1999, Packers @ Bears | N 18 mph | Bears elect to defend north goal | Packers receive twice |
| Jan 2, 2000, Seahawks @ Jets | SW 4 mph | Jets elect to defend west goal | Seahawks receive twice |
| Dec 12, 2000, Lions @ Jets | SE 20-30 mph | Jets elect to defend west goal | Lions receive twice |
| Dec 15, 2001, Cardinals @ Giants | N 18 mph | Giants elect to defend east goal | Cardinals receive twice |
| Dec 30, 2001, Bills @ Jets | WNW 15-25 mph | Jets elect to defend east goal | Bills receive twice |
| Dec 1, 2002, Titans @ Giants | NW 18 mph | Giants elect to defend east goal | Titans receive 1st half, Giants 2nd |
| Nov 7, 2004, Jets @ Bills | SW 25 mph | Jets elect to defend west goal | Bills receive twice |
In all of these cases, the coin toss winner chose a goal to defend (i.e. no one seems to have picked the "kick" option). The Jets did so 4 times, and the Giants did twice. Five of these were games were in the old Giants Stadium (shared by the Giants and Jets). Two happened in the Bears' Soldier Field, and 1 was at Buffalo's Ralph Wilson Stadium. These stadiums have been known to be windy, and the wind speeds recorded at game time do suggest that wind might have been a factor in these decisions.
A new option: Deferring
The rule change implemented in 2008 gave an additional option to the team winning the coin toss: They could now defer the privilege of making the first choice to the second half. The procedure after deferring would be essentially the same as if the deferring team had lost a coin toss prior to 2008.
Teams selected this option 39% of the time in the 2008 season, and just 26% and 30% in the 2009 and 2010. In the following seasons, deferring grew in popularity, surpassing "receive" in 2012 as the most popular option. In the 2018 season, teams winning the coin toss deferred 92% of the time.
But teams have not been equal in their tendency to defer. New England's Bill Belichick gained attention early on for his tendency to defer, and since the rule change, leads the league in doing so 95% of the time. At the other end, the Cowboys have chosen to defer just 26% of the time since the rule change. Most NFL teams (23 of 32) have tended to defer more than receive.
Although some teams were slow to adopt the new strategy, all teams eventually joined the trend towards deferring. the 2018 season marks the first time since the rule change that every single team in the league chose to defer more often than they chose to receive the opening kickoff.
Why did coin toss strategies change?
Why did it take so long for coaches to realize that the "defer" option was preferable to receiving the opening kickoff? If there were good reasons to defer—confidence in half-time adjustments, kicking with the wind in the 4th quarter—wouldn't coaches have already been aware of them prior to the rule change? In that case we would have expected teams to begin choosing "defer" as soon as the option became available.
One possibility is that the popularity of deferring is related to other changes in coaching strategy that happened to coincide with the new coin toss rules. It could also be that coaches were not inclined to develop preferences about the coin toss while it seemed out of their control, and that the availability of a choice led to more reflection and strategizing, and the realization that deferring was preferable.
But yet another possibility is that coin toss strategy doesn't really matter. It doesn't have a significant impact on scoring or who wins the game. Coaches initially favored receiving the opening kickoff after the rule change because that was called "winning" the coin toss in the past, and "winning" is a good word. Then they noticed that Bill Belichick was doing something different and was still winning games and was also being called "clever", so they wanted to be more like him.
In the 9 seasons prior to the rule change, teams winning the coin toss won 49.0% of their games. After the rule change, they won 51.5%, and the win percentage among teams that chose to defer was 52.0%, compared to 50.8% for teams choosing to receive. While it's tempting to attribute these differences to coin toss strategy—no coach would turn down the option to increase win percentage by 1 or 2 percentage points—it's still not compelling evidence of a real advantage. Estimating linear models predicting game results based on what happens during the coin toss yields standard errors that are much larger than the raw differences in score differentials and win percentages. So while it is possible that the coin toss does matter, proving that it matters is difficult.
