Home-field advantage in NFL: Toughest places to play, more

Who has best, worst home-field advantage in NFL? (1:56)

The NFL Live crew debates which teams are at the top of the league in terms of home-field advantage and which are at the bottom. (1:56)

Home-field advantage is a tricky thing to pin down. I wouldn't want to go play the New England Patriots in Foxboro in January if my season was on the line -- they've won their past five home playoff games. But they were 3-3 in those same must-win home games between 2008 and 2011, losing to Mark Sanchez and Joe Flacco (twice). Everyone knows how brutal the Seahawks can be in Seattle, but it took a Calvin Johnson fumble on the goal line and a missed penalty call to keep them from going 4-4 at CenturyLink Field in 2015.

Basing anything on eight games per season or something anecdotal like decibel levels or temperature at kickoff won't give us the full picture of who actually enjoys the best home-field advantage in the NFL. And likewise, just looking at a team's record at home can be misleading. If a team sweeps all of their road games and goes 6-2 at home to finish 14-2, do they really have a significant home-field advantage, or are they just a very good team anywhere? And does a home-field advantage conferred by a rowdy fan base transfer over from an old stadium to a new one?

Let's try to get those questions answered here in advance of the 2017 season. I looked into this very topic several years ago and found that the Seahawks, narrowly, had enjoyed the league's best home-field advantage. Is that still the case? And who gets the least out of their home cooking?

Figuring it out

To estimate a team's home-field advantage, I went on a season-by-season basis since 1990 and calculated its average point differential in home games, and did the same for road contests. If you add those together and divide by two, you get the team's observed point differential. Using point differential is better than win-loss record because point differential does a better job of predicting future win-loss record than winning percentage itself.

As an example, the 2016 Dallas Cowboys outscored opposing teams in their eight home games by a total of 75 points, good for an average of 9.4 points per contest. They were good on the road, too, as they went 6-2 while outscoring their opponents by a total of 40 points, or an even 5.0 points per game. Their observed home-field advantage is ((9.4 - 5.0) / 2), which is 2.2 points.

That's not an especially impressive home-field advantage -- it ranked 21st among the 32 NFL teams last season. The Seahawks led the league with an observed HFA of 7.5 points, while the Indianapolis Colts were remarkably 4.9 points per game worse at home than they were on the road. The average home-field advantage was good for 2.6 points, which is right in line with the suggestions that Las Vegas values home-field advantage as being worth 2.5 to three points.

The problem with using single-season data is that the fluctuations from year to year can be massive. Take 2007, when the Detroit Lions posted a 7-9 record and a massive observed home-field advantage of 11.6 points per game. That's the second-largest single-season HFA on the books between 1990 and 2016. The following year, the 0-16 Lions -- with many of the same players and fans in the same stadium -- generated a home-field "advantage" worth minus-6.4 points per game, which was the worst mark on record between those same years. There's no question the fans turned on the team, but it's hard to believe that the swing was worth 18 points per game. (The second-worst home-field advantage, interestingly enough, belonged to a pair of playoff teams: the 1993 Pittsburgh Steelers and the 2007 New York Giants, who, including the postseason, went 10-1 on the road before beating the Patriots in the Super Bowl.)

To get a better sense of who really enjoys home-field advantage, we need to take a larger sample; so let's do that. Keep in mind that these numbers are only for regular-season games and exclude games played on neutral sites. The games played in London and Mexico City in recent years are excluded, as are the ones Buffalo played north of the border in Toronto. Games moved by weather are also excluded, most notably when the Saints were affected by Hurricane Katrina in 2005.

Decade of data

The simplest way to estimate home-field advantage is to take a look at each organization during the past 10 seasons, going back through the 2007 campaign. These numbers don't include the neutral-site or otherwise-abnormal games I mentioned, but if a team moved to a new stadium during those 10 years, I included their numbers from both stadiums. It's not a surprise to see the team at No. 1:

The Seahawks would have been a popular guess for the top spot, and teams like the Packers and Bills that play their home games in frigid temperatures in front of rabid fans wouldn't be far behind. I mentioned the Ravens as a bit of a surprise team several years ago, and they've mostly kept up -- aside from the 2015 campaign, when Baltimore actually had a negative observed HFA of 1.6 points per contest.

The 49ers, though? They have to be a surprise, right? You would figure that the last days of Candlestick would be pushing San Francisco up toward the top of the charts, and indeed, the 49ers had an observed HFA of 3.8 points per game in their longtime home between 2007 and 2013. Despite the complaints about the move to Santa Clara and the franchise bottoming out during the past few years, though, the 49ers have kept things up, as their home-field advantage at Levi's Stadium has been good for an even 4.0 points per contest. I don't believe it, either.

And while I mentioned the Bills and Packers as cold-weather teams that enjoy a comfortable home-field advantage, I'm not sure it really carries over as a trend. The Lions, Cardinals, Texans, and Saints all play in domes and have home-field advantages much higher than the league average. The three non-dome teams in the NFC East all see plenty of bad weather, and haven't been notably effective at home.

When you look at the teams the Weather Channel ranked as the five worst weather cities in football and how they've performed over different parts of the season, the differences just aren't stark. Since 1990 (expanding the sample to get more games), Buffalo, Cleveland, Green Bay, Kansas City, and Pittsburgh have generated a home-field advantage of 3.0 points in games played in September and October. We'd expect their home-field advantage to improve dramatically as conditions deteriorate in the later months, but from November on, their home-field advantage has been worth a nearly identical 3.2 points per contest.

It's not clear noise is an issue, either, with the Chiefs and Raiders both ranking in the bottom five. If you're wondering whether the Raiders turned things around as the organization roared back to life last season, their home-field advantage in 2016 was only observed to be 0.5 points per game. For whatever the Seahawks are doing in Seattle, sound isn't the only factor driving their success at home.

Method 2: Current stadium-specific

Maybe we're underestimating the impact of a team's specific digs. Let's consider this question by looking specifically at how teams have performed in their current stadium going back through 1990. So for the Cowboys, this means we'll be looking at their games between 2009 and 2016, while every game from the Packers since 1990 would be up for consideration. For teams that haven't yet made it to five years in their present stadium, I substituted their most recent stadium, which affects the Falcons, Chargers, Rams, 49ers and Vikings.

The numbers are mostly similar, but there are a few leaps up and down the rankings:

Jets fans might best be known for their draft-day reactions, but Gang Green has enjoyed a significant home-field advantage since moving into MetLife Stadium, one notably higher than their roommates in blue. The Giants are in the middle of the pack, while their NFC East counterparts all hold up the rear, which seems odd. There shouldn't be dramatic strength-of-schedule concerns for those teams, each of whom moved into a new stadium between 1997 and 2009.

Could it be the new digs? We often hear about how teams lose some of the energy and attitude we commonly associate with a home-field advantage by moving from a tired, old stadium into a more expensive arena. It's entirely possible the loudest fans are getting marginalized by higher ticket prices and personal seat licenses and are being replaced by silent, uninvested corporate clients. Is there evidence of that in the data?

Not really. In looking at teams that have spent five-plus years in multiple stadiums between 1990 and 2016, while they've declined on the whole, it's only by 0.3 points of home-field advantage per game. Those three teams in the NFC East have all fallen off, but it's counteracted by gains for the Giants, Jets, and Seahawks:

The takeaways

Baltimore and Seattle are the two toughest places to play on the road. Regardless of how we calculate the numbers, the Ravens and Seahawks consistently rank among the organizations that improve most dramatically at home. If we calculate this by sheer winning percentage as opposed to point differential, as an example, the Ravens have the largest home-field advantage in the league since 2007, winning 72.5 percent of their games at home but just 40 percent on the road. The Seahawks are third, with the Vikings splitting the two.

Picking between them is tough, but I would say narrowly that the Seahawks have the best home-field advantage, in part because their home crowd might have kept Seattle afloat at times. There have been eight seasons since 2007 in which a team has enjoyed an observed home-field advantage of 10 points or more, and the Seahawks are the only team with more than one of those seasons to their name, with the 2009 and 2012 teams topping double-digits.

The second tier includes a handful of teams with no obviously similar characteristics. After the Ravens and Seahawks, there are six organizations ranking in the top 10 by each of our attempts to estimate home-field advantage: the 49ers, Bills, Cardinals, Lions, Packers and Vikings. Three of them play in the NFC North, but how would they explain the 49ers and Cardinals? How do the Bills figure in?

Some of the anecdotal arguments don't hold up under much scrutiny. The NFC North teams enjoy playing at home in cold weather, but they're often playing their divisional brethren or other teams who play their own home games in frigid conditions, too. Since the NFL went to their current divisional structure in 2002, teams from the AFC and NFC South have lost road games played with a starting temperature of 32 degrees or lower by an average of 5.1 points. Teams from the AFC and NFC North, who should be used to those freezing temperatures, have lost those same games by an average of 4.8 points.

The NFC East doesn't appear to offer much home-field advantage. It's tempting (especially to NFC East fans) to wonder whether the relative difficulty and success of the division would make it less likely those teams play effectively on the road, but there's no relationship between strength of schedule and home-field advantage. The correlation coefficient between the two by my strength of schedule measure, which takes a team's schedule and calculates each opponent's average point differential in games not involving the team in question, is just 0.04.

There's a lot we don't know. It does seem likely that we're missing something in terms of understanding home-field advantage, given that three of the four teams in the NFC North (and most of the NFC West) rank among the league leaders, while the NFC East toils in the doldrums. This is the unfortunate reality of a sport in which teams get only eight home games per year. The Rams spent 21 years in St. Louis and played only 167 regular-season games in their home dome, which is essentially two years' worth of home data for MLB's Cardinals at Busch Stadium.

It's probably best to regress most teams' home-field advantage toward the mean and assume they get about 2.5 to three points at home as opposed to a neutral field. The Seahawks and Ravens might be the exceptions to that rule.