Homefield Advantage in the NFL

It’s a nebulous topic.  Whether it’s Our House, The 12th Man, or any other name homefield advantage is an intangible that’s often invoked in football.

This season I attempted to use the Power Rank’s neutral site predictive capability and account for home field advantage…  and failed.  I added an arbitrary three points for the home team, a practice that is relatively common when setting a betting line, but a practice that is hardly scientific.

Rather than develop a complex mathematical matrix to divine the true value of the homefield advantage (I’m just not that smart…  I’ll let Ed work on that!) I decided to poke back into the recent history of the NFL to see is homefield advantage is a truly intangible myth or if it is a real and tangible aspect of the game which might one day be quantifiable.

I gathered data from NFL.com for the years 2002-2010, the only years that represent the current 32 team configuration of the league.

The first thing I wanted to look at was regular season performance.

Since 2002 home teams have won over 57% of the time, while road teams have won only about 43% of the time.  That’s a fairly strong indication that homefield advantage is tangible and meaningful, especially since every year has produced winning records for the home team.  The worst year for home teams since 2002 was 2006, when home teams won barely more than 53% of the time.

There’s a lot that goes into why playing at home is advantageous, from the comfort of your own familiar locker room to the jet lag (or lack thereof) of a coast-to-coast trip to field conditions.  One of the biggest and most obvious advantages is the crowd noise produced by a home team’s fans.  I wondered if this factor would increase the homefield advantage during the playoffs, when fans are going to be louder and more interested in every play of each game.

Here are the results for homefield advantage in the playoffs since 2002, with 2010 incomplete (and not counted):

This time there’s an even stronger correlation between home team and victory in the final win percentage, but the playoffs have special conditions that must be considered before jumping to the conclusion that the correlation is in fact stronger.  First of all, in the regular season the home team is more or less random, whereas in the postseason the home team is determined by the strength of their regular season finish…  usually.

In most cases homefield advantage in the playoffs is awarded to the team with the best record (or by tiebreakers if they had the same record) but this is not always the case.  Division winners are awarded home field advantage over wild card teams regardless of record.  That’s why the 11-5 Saints visited the 7-9 Seahawks, the 10-6 Packers visited the 10-6 Eagles (who they had beat on tiebreakers), the 12-4 Ravens visited the 10-6 Chiefs, and the 11-5 Jets visited the 10-6 Colts last week.

So in the wild card round, you have teams with better records visiting arguably worse teams that won their division.  The results from these four games last week?  A 1-3 record for home teams, with the Seahawks pulling of a somewhat miraculous win over the Saints.

So I also broke down the records of wild card games by homefield:

Not only is this correlation weaker than the homefield advantage in the regular season, its also generally weak.  52.8% versus 47.2% doesn’t create insurmountable odds.  In fact, the odds for an away team to win in the wild card round are about equal to the odds of beating the house on a casino Roulette wheel (with 0 and 00 spaces).

So what happens in the divisional and conference championship rounds when homefield advantage almost always goes to the team with a better record?

Now the correlation is at its strongest, but again the home team is already coming in favored since they usually have a better record, especially in the championship round when the home team always has the better record.

Does this just mean that the better team is winning?  I don’t think so.  The correlations here are significantly larger than the correlation in the regular season, and as all football fans know, the best team doesn’t always show up and win.  I’d attribute the correlation in the later playoff rounds to a combination of the two:  a better team playing at home has a very significantly better chance of winning in the playoffs.

So what does that mean for this week?  All of this week’s match ups feature a weaker team visiting a stronger team.  This is the kind of correlation we see in the conference championship round of the playoffs, about a 69% chance for the home team to win versus 31% for the away team.

Seattle losing to Chicago and New York losing to New England seem like safe bets this week by conventional wisdom, but the other two games are less certain.  Baltimore finished with the same record as Pittsburgh and lost the division on tiebreakers.  The  Power Rank’s #2 team Green Bay visit the Falcons, who needed a few lucky breaks to win at home against the Packers in week 12 this season.

If one of these potential upsets occurs, the record for better teams playing at home this week will be 0.750, close to the expected 0.688 correlation that is shown in the conference championship rounds of the last eight playoffs.

So it seems that homefield advantage is certainly tangible, and is particularly strong when it favors an already superior team (according to record) in the playoffs.