Where Did All The Lopsided Super Bowls Go?

On paper, Sunday’s Super Bowl between the New England Patriots and Los Angeles Rams has all the earmarks of an instant classic. The teams are evenly matched in their strengths and weaknesses, and the favored Patriots’ pregame chance of victory — 52.8 percent, according to our Elo model — is among the lowest¹ of any favorite in Super Bowl history, meaning the game is close to a toss-up.

In the past, though, this would have guaranteed nothing — and even could have been viewed as a bad omen. (Or at least, it wouldn’t have been reliably good.) It wasn’t too long ago that we were in an era of total Super Bowl stink bombs being a regular occurrence. Some of the best (i.e., worst) examples of this, in terms of time spent with the score lopsided, included:

• Super Bowl XX (1986) — Chicago 46, New England 10
  • Average score margin:² +/-19.6 points
  • Ties/lead changes:³ 3
  • After the Pats took a quick 3-0 lead, the Bears rattled off 44 consecutive points. Chicago ended up with 167 yards and four touchdowns on the ground while burying New England QBs Tony Eason and Steve Grogan under relentless defensive pressure.
• Super Bowl XXIV (1990) — San Francisco 55, Denver 10
  • Average score margin: +/-23.2 points
  • Ties/lead changes: 1
  • John Elway’s Broncos were no match for Joe Montana and the 49ers, who took a 41-3 lead five-and-a-half minutes into the third quarter and cruised to the most dominating victory in Super Bowl history. Fewer than 15 minutes in this game were spent with Denver even being within one score of the lead.
• Super Bowl XXVI (1992) — Washington 37, Buffalo 24
  • Average score margin: +/-12.9 points
  • Ties/lead changes: 1
  • A year after losing one of the most dramatic Super Bowls ever, the Bills were undone by Washington’s 24-0 run in the second and third quarters. Buffalo eventually found itself in a 37-10 hole before a couple of meaningless TDs made the final score seem far more respectable than it actually was.
• Super Bowl XXIX (1995) — San Francisco 49, San Diego 26
  • Average score margin: +/-19.1 points
  • Ties/lead changes: 1
  • Any hope that this matchup might be competitive vanished when Steve Young hit Jerry Rice for a 44-yard TD a scant 1:24 into the game. It was the first of seven 49er touchdowns on the day and part of the 35-10 lead San Francisco would build minutes into the third quarter.
• Super Bowl XXXV (2001) — Baltimore 34, N.Y. Giants 7
  • Average score margin: +/-11.1 points
  • Ties/lead changes: 1
  • Baltimore’s defense completely owned this game, keeping the New York offense from crossing the 50-yard line until late in the first half (they would only run four plays in Raven territory at all) and harassing Kerry Collins into the second-lowest-rated passing performance in Super Bowl history.
• Super Bowl XXXVII (2003) — Tampa Bay 48, Oakland 21
  • Average score margin: +/-11.1 points
  • Ties/lead changes: 3
  • Led by another of history’s great defenses, Tampa jumped out to a 34-3 third-quarter lead and never looked back. (As it turns out, maybe you should change up your audibles and play calls when facing your former coach in the Super Bowl.)

(The list could have also included a few other notable snoozers, such as Washington’s 42-10 win over Denver in 1988 and Denver’s own easy 34-19 win over Atlanta in 1999.)

Sure, there were a few all-time-great Super Bowls interspersed among those stinkers — such as Joe Montana’s final-drive victory in 1989, the aforementioned Giants-Bills classic from 1991, Elway’s first Super Bowl win in 1998 and the Rams’ epic finish against the Titans in 2000. But for most of the Super Bowl’s first four decades of existence, there were just as many duds, games that could make you wonder why so much hoopla was devoted to what would frequently turn into a boring blowout.

More recently, though, we’ve been treated to far better games on the Super Bowl stage — and practically no true laughers. How can we prove it? Statistically, there are a number of ways we can measure the excitement of any given game, and this post from ESPN’s Brian Burke lays out a few of my favorites:

• Excitement Index is the sum total of movement in win probability across an entire game. This rewards back-and-forth games with a lot of uncertainty about the eventual winner until the very end of the game.

• Comeback Factor is the eventual loser’s highest win probability in the game. This essentially measures how dramatic a comeback the winning team needed to stage, with larger numbers representing more unlikely victories.

In concert, these metrics tend to do a good job of measuring what we want out of a thrilling Super Bowl — late-game heroics, big comebacks and results that come down to the wire. With win probability data provided by ESPN’s Stats & Information Group,⁴ I calculated both metrics for every Super Bowl (since the first game was played in 1967) and plotted them out as a time series by season:

In recent years, you can see a pattern emerge: Very few Super Bowls have been duds, according to the Excitement Index. Since that Bucs-Raiders debacle in 2003, 13 of the past 15 matchups have registered an Excitement Index above the all-time average for Super Bowls.⁵ And of the two that didn’t, one was New England’s unprecedented 28-3 comeback against the Atlanta Falcons in 2017, which only graded as below average because so much of the game was spent with the Falcons in command, before the Pats stormed back. (It made up for that lower Excitement Index with the highest Comeback Factor of any Super Bowl by far.)

Over the past decade-and-a-half, the only genuinely terrible Super Bowl came in 2014, when the Legion of Boom Seahawks thoroughly routed Peyton Manning’s favored Broncos, 43-8. That game nearly set new records for the lowest Excitement Index and most lopsided average score margin in a Super Bowl (it ranks second in both categories). But it was also an aberration in an era when great Super Bowls are the norm and duds are very much the exception.

Some of that may be because of the matchups actually becoming closer over time. According to our Elo ratings, the average Super Bowl’s pregame spread during the 2010s has been 2.1 points, the smallest in any decade since the 1970s. By comparison, that number was 3.2 from the 1980s through the 2000s. So it’s not exactly a coincidence that the games have gotten better in the same span — though that fact alone doesn’t necessarily explain all of the uptick in great Super Bowls.

So does all of this mean we’re “due” for a dull game Sunday? No. But we’ve also been extremely spoiled by seeing so many great ones in recent history. We’ve watched just as many Super Bowls hit an Excitement Index of 5.0 in the past 14 seasons as we had in the 38 seasons beforehand. Something might have to give eventually, even if all signs point to this weekend’s game continuing the exciting run.

ESPN’s Brian Burke contributed research.

Check out our latest NFL predictions.