Russell Wilson called it “the craziest game” he’d ever played.
The Seattle Seahawks’ Monday Night visit to San Francisco last November was a tumultuous affair that lasted a full 70 minutes. When the dust settled, the Seahawks had erased a 10-point deficit, survived four giveaways, escaped defeat when the 49ers missed a field-goal attempt in overtime, and won it as time expired in the extra period. And with that result, an epic NFL streak approached five decades in length.
After eight consecutive victories, the 49ers had suffered their first loss of the 2019 campaign, joining the league’s 31 other teams with at least one defeat. Thus it was assured that, for the 47th straight year, no team would finish an NFL season with a perfect record over the regular season and playoffs. The legacy of the 1972 Miami Dolphins, the last team to pull off such a season, remained intact.
When Don Shula, the coach of that legendary Dolphins team, passed away earlier this month, the topic of perfect seasons bubbled to the surface again. Of course, Shula’s career produced a litany of achievements beyond that one season. His 328 regular-season wins are the most in league history; Bill Belichick, with 273 wins, is the only active coach within 100 of that total. Shula coached in six Super Bowls, winning Lombardi Trophies in both the ’72 and ’73 seasons, and ranks third in league annals with 19 postseason victories.
While it’s clear that Belichick needs to coach at least five more seasons to approach Shula’s win total, it’s harder to assess the likelihood of a team duplicating the Dolphins’ perfect campaign. The NFL schedule expanded by two games in 1978 to a 16-game regular season. So while Miami’s feat consisted of 17 total wins, any unblemished campaign in future years – barring further schedule changes – would require 19 victories. This figure might go up as soon as 2021, as signs point to the NFL adding another game to the schedule. For now we’ll consider 19-0 the target, and we wonder how likely is such a feat.
To assess those chances, we looked back at the past 25 years in the NFL, initially considering just regular-season action. In that time, only Belichick’s 2007 Patriots survived a regular season without a loss or tie. In fact, that team and the ’72 Dolphins are the only ones since World War II to enjoy perfect regular seasons. That New England team failed to etch its name alongside Shula’s unblemished Dolphins, suffering a shocking loss to the New York Giants in Super Bowl XLII. More on that later.
By tabulating points per possession on both offense and defense, Stats Perform is able to more accurately assess the relative strength of each team, going deeper than win-loss records. Using those ratings as the basis, we ran 100 models of each of the past 25 regular seasons, based on the slate of games that each team played, and searched for teams that went 16-0 most frequently.
Surprisingly, the ’07 Patriots do not top the list, recording a 16-0 mark in 13 of 100 simulated seasons. The leaders were the 1999 St. Louis Rams. The Greatest Show on Turf was projected to go undefeated in the regular season nearly twice as often as the real-life 16-0 Patriots were.
Most Times Going 16-0 in 100 Simulated Seasons, 1995-2019
|Year||Team||16-0 Seasons||Actual Result|
|1999||St. Louis Rams||24||13-3, Won SB|
|2001||St. Louis Rams||20||14-2 Lost SB|
|1996||Green Bay Packers||16||13-3, Won SB|
|2007||New England Patriots||13||16-0, Lost SB|
|1998||Denver Broncos||9||14-2, Won SB|
|2016||New England Patriots||9||14-2, Won SB|
The ’99 Rams were an improbable story. The team had signed free agent Trent Green to be its starting quarterback, only to lose Green for the year when he suffered a preseason knee injury. Enter Kurt Warner, whose career to that point consisted of 10,164 passing yards in the Arena Football League and all of 39 yards in the NFL. By the time the ’99 campaign came to a close, Warner had thrown 41 regular-season TD passes and earned league and Super Bowl MVP honors as the Rams captured the Lombardi Trophy.
How close did those Rams actually come to a 16-0 season? They lost three one-score games by a combined 14 points, and in their first two defeats, the margins were razor-thin. In Week 7 at Tennessee, St. Louis’ Jeff Wilkins missed a 38-yard field-goal attempt with seven seconds left, saddling the Rams with a three-point defeat. It was the only time all season that Wilkins missed from inside 40 yards; he made his other 12 attempts.
One week later, the Rams dropped a back-and-forth affair at Detroit that saw seven lead changes, the final one coming on a 12-yard TD pass from the Lions’ Gus Frerotte to Johnnie Morton with 28 seconds remaining. Detroit’s winning drive was sustained by an improbable 57-yard completion by Frerotte on a 4th-and-26 play. Note that in the last quarter-century in the NFL, fourth-down plays with 20 or more yards to go have been converted at a rate of 10.3%.
Had those two outcomes been flipped, it’s not far-fetched that St. Louis would have been 15-0 heading into its season finale at Philadelphia, having won their 13 other games by at least 13 points each. Against the Eagles in Week 17, the Rams had already wrapped up the conference’s top seed and had no strong motivation to win. Warner exited in the third quarter with the game tied, and star running back Marshall Faulk touched the ball only nine times in a seven-point defeat.
Situations like this highlight the importance of a variable that is hard to account for in assessing the probability of an undefeated season: Will teams that have secured their conference’s No. 1 playoff seed pull back the reins in Week 16 or 17? Consider the 2009 Indianapolis Colts, who clearly chose not to take aim at the ’72 Dolphins. Peyton Manning and Co. were 14-0 when they faced the New York Jets in Week 16. With a five-point lead in the third quarter, Manning was replaced under center by Curtis Painter. The Jets outscored the Colts by a 19-0 margin over the remaining 21 minutes, assuring there would be no undefeated season in Indy.
This much is clear: Almost any team that gets to 14-0 will have guaranteed itself home-field advantage in its conference playoff games and will have the option to rest its key players late in the season. Different teams will have differing philosophies in this situation – the ’07 Patriots clearly went all-out to reach 16-0 – but the fact is that any team approaching a 16-0 season has the option of not pursuing it wholeheartedly.
Another unavoidable and unpredictable factor in forecasting perfect seasons is health. The impact of injuries on any single game can vary widely. But for a team to last an entire season without a loss would, in theory, require good health throughout the 16-game schedule – unless you’re the 1972 Dolphins. That team’s Hall of Fame quarterback, Bob Griese, was injured in the fifth game and did not start again until the Super Bowl.
Undeterred, Miami won nine regular-season games and two playoff contests with Earl Morrall as its quarterback. Remember, this was an era when the running game dominated the NFL, especially in Miami. Led by a pair of 1,000-yard runners, Larry Csonka and Mercury Morris, that Dolphins team rushed the ball 43.8 times per game, with runs on 68.6% of its snaps. No other team in the last 35 years has averaged as many as 40 rushes per contest, and only one had a rush percentage as high as 60%. Withstanding a long-term injury to a starting QB is far less likely than it was 48 years ago – unless there’s a Kurt Warner waiting in the wings.
So let’s get back to our six teams that had the highest likelihood of perfect regular seasons. Naturally, the next step for each is to consider its chances of capping its 16-0 mark with a Super Bowl victory. Each of those clubs reached the Super Bowl, so we simulated games against the three opponents that each faced in the postseason. We took those results and combined them with the odds of the unblemished regular seasons to get each team’s chances to go 19-0.
Number of Times Going Undefeated in 100 Simulated Seasons, 1995-2019
|Year||Team||16-0 Reg. Season||SB Win||19-0 Overall|
|1999||St. Louis Rams||24||62||14.9|
|2001||St. Louis Rams||20||47||9.4|
|1996||Green Bay Packers||16||53||8.5|
|2007||New England Patriots||13||62||8.1|
|2016||New England Patriots||9||29||2.6|
Again, the 1999 Rams are the best: In roughly one out of seven simulated seasons, St. Louis turned in a perfect 19-0. That team might be overlooked in discussions of the best in history, but it did post a regular-season point differential of plus-284, second best in the Super Bowl era behind – who else – the 2007 Patriots (plus-315). The fact that their odds were as high as they were suggests that a perfect season is not unattainable.
Speaking of that ill-fated New England team, we wondered how agonizingly close it came to going 19-0. So we put the two Super Bowl XLII participants, the Pats and the Giants, into the simulator and returned these results: the Patriots won the game 80 times out of 100, including one victory by 52 points! But when the game was actually played on February 3, 2008, the Giants got one of those 20 winning results—proving how difficult it is to accurately predict the likelihood of a Helmet Catch.
Finally, not wanting to overlook teams at the other end of the scale, we wondered about the two teams to go 0-16, the 2008 Detroit Lions and 2017 Cleveland Browns. In what might be called gallows simulation, we played each of those teams against the 2007 Patriots 100 times, giving each club the home field in 50 of the matchups. If you thought the Pats would go 100-0 in either matchup, you’d be disappointed. New England won only 99 times against Detroit and struggled to a 97-3 mark versus the Browns.
It seems that any way you slice it, perfection remains just out of reach for Belichick.
Data modeling by Matt Scott, and research support provided by Evan Boyd.