≡ Menu

The Patriots won Super Bowl XLIX, and whatever your thoughts on the end of the game, there’s no doubt that New England was one of the top teams in the NFL in 2014. But it’s not quite so easy to identify why, at least when looking at the traditional per-play metrics. New England ranked 17th in Net Yards per Pass Attempt and 16th in Net Yards per Pass Attempt allowed, hardly the stuff of Super Bowl champions. The Patriots didn’t stand out as particularly excellent as a rushing offense or a rushing defense, either.

But those passing statistics belie the fact that the Patriots did, in fact, have a great offense this year. Part of the issue was the slow start and a meaningless week 17 game. Beginning in week 5, and excluding the week 17 game, New England scored 487 points, a 34.8 points per game average. That matches what the team did in 2012, when the Patriots had a historically lethal offense. And it’s not too far off from even the heights reached by the ’07 team.

The Patriots passing attack ranked 5th in TD rate, 3rd in INT rate, and 4th in sack rate; as a result, they jump from 17th to 6th when moving from NY/A to ANY/A. But the Patriots were even better at pure scoring.1 That’s been a trend for the team: during the Tom Brady era, New England has fared better in points scored than it has in ANY/A, and fared better in ANY/A than the team has in NY/A. And New England has generally been improving in all three statistics, too.

There is one area where the 2014 Patriots stand out as special. New England had just 13 turnovers all season: 9 Brady interceptions, three Brady fumbles, and one Brandon LaFell fumble. That is tied for the third best ever, although that sounds better than it is. The record for turnovers per game is 10 turnovers per 16 games, a feat accomplished by the 2010 Patriots and then the 2011 49ers. In 2014, the Packers also committed just 13 turnovers, and the Seahawks had just 14. As you might suspect, yes, this does mean that turnover rates have declined significantly in recent history. Take a look at the following graph, which depicts turnovers per 16 games for the average NFL team since 1970.  The purple line shows all turnovers; the blue and red lines are for interceptions and fumbles lost, respectively.

TO fum INT

We know that interception rates have declined significantly; the graph above shows that even with significantly more pass attempts, raw interceptions are also on the decline. What’s also interesting is that fumble rates are declining as well. Part of this may be because of the NFL’s attempt to take special teams out of the game, along with a general emphasis on ball control for running backs and a decline in total sacks.

But here is your stark reminder in how much things have changed in the NFL. The 1979 Steelers were the defending Super Bowl champions, and finished the season as repeat champions. But while the 2014 Super Bowl champions committed just 13 turnovers, the ’79 champions had FOUR TIMES as many turnovers. That’s correct: in 1979, Pittsburgh led the NFL with a whopping 52 turnovers!2

As a general rule, there’s not a huge correlation between turnovers and winning the Super Bowl. The graph below shows where each Super Bowl champion ranked in turnovers.

rk sb champ to

The last 3 Super Bowl champions ranked in the top 4 in turnovers, but my hunch is that more short-term noise than evidence of a trend in the NFL.

  1. While New England moves at a fast pace, they actually ranked 3rd in points per drive and 4th in overall points, because the Broncos had even more drives than New England. []
  2. For what it’s worth, but the Eagles topped all teams with 36 in 2014. []
  • Andrew Healy

    Part of that non-correlation might be the randomness in who wins the Super Bowl. The last couple of years, two of the best teams in football won the Super Bowl and were low in turnovers. In the early 1990s, the best teams kept winning Super Bowls and they were all low in turnovers (notably the 1990 Giants, who were #1 in DVOA that year).

    • sacramento gold miners

      Not a Patriots fan, but New England proved on the field they were the best team this past season. We can pull out all kinds of data, but the only metric which really counts is the scoreboard at the end of these games. Small sample size, yes, but Seattle had a ton of opportunities last Sunday, and just didn’t get the job done.

      Never believed in the randomness of championship sports events, not when so many plays are involved. Each team has opportunities to potentially change the outcome on a single play. It’s hardly asking too much of a team to earn the honor of being called a world champion by proving it on the field. When the Patriots went 18-1 several years ago, they enjoyed a remarkable season, but failed to earn immortality by losing the most important game of the year, against a team they had already defeated on the road.

      • James

        You don’t believe in randomness, yet you acknowledge that one play can change the outcome of a game?

        • sacramento gold miners

          I don’t believe there are any random plays in an NFL game. Each team prepares during the week, but sometimes another club will surprise the other. That falls on the coaches, or a mistake by the players. In a championship game between two excellent teams, one key play can change everything, good or bad. As much as I agree with others about the mistake Seattle made in not running the ball, we can’t say for sure Lynch would have scored. Looking back, there were still numerous chances for Seattle to win, plays not made offensively and defensively. I’ve watched plenty of games where I felt one team outplayed the other, and that team won the stats battle.

          If we’re not going to use the scoreboard to determine the best team, I can’t see having an on field result overturned by another method. The 2007 Patriots themselves will tell you that had their shots in the Super Bowl that season, but it didn’t turn out for them. With Ed Sabol passing away yesterday, I must use a fitting NFL Films line here from the Super Bowl 14 program, “Great teams aren’t always great, they’re just great when they have to be”.

      • Richie

        “Never believed in the randomness of championship sports events, not when so many plays are involved.”

        The Super Bowl featured about 170 plays. I wouldn’t consider that “so many plays” to the point of being able to eliminate randomness.

  • Tom

    My mind has been blown again…the Steelers committed 52 turnovers in ’79? That’s three per game!! True to form, Bradshaw threw three picks in the Super Bowl that year against the Rams…

    • Steve

      A sign of how good that team was that they overcome that to win it all.

    • Bryan Frye

      If you like weird coincidences, the Rams were the second worst team in the league that year at turning over the football, with 49. I’d be willing to bet that’s the only Super Bowl to feature the bottom two teams in offensive turnovers.

      The Steelers committed a turnover in all but one game, which is remarkable. That includes games with 9 (vs. Bengals) and 8 (vs. Chargers) turnovers.

      • Tom

        Bryan – I’d bet money you’re right about that being the only Super Bowl with the bottom two turnover teams…that’s ridiculous. And 9 giveaways in one game? Probably not the record (I’m too lazy to check PFR), but that’s gotta be up there.