SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.

Also, our existing PFR blog rss feed will be redirected to the new site's feed.

Pro-Football-Reference.com ยป Sports Reference

For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.

(Travel) time and temperature, part two

Posted by Doug on October 3, 2006

This is a continuation of yesterday's guest column by JKL:

Today, I am going to look at the effect of climate. This study is a little less exact than calculating the distance between two cities. After all, most cities are not like Rock Ridge; they do not change locations overnight. However, temperatures can and do fluctuate throughout a season. It would be too time consuming, and probably impossible, to find accurate weather information on every divisional game played from 1986-2005, so I tried to do the next best thing.

To get a general sense of the effect of climate, I used the National Weather Service median daily high temperatures for each city from the months of September through December, and averaged those 4 months. For teams that played in a dome, I approximated 74 degrees as the average in-season temperature. Houston, because of the retractable roof, was a potential problem, but as it turns out, the average in-season daily high in Houston (if you consider Houston an outdoor team) is only 2 degrees higher than my guess at dome temperatures.

The list below is similar to the distance list from yesterday. It lists the home team wins and losses in divisional series between 1986-2005, sorted by the average in season temperature difference (in Fahrenheit) between the two cities.

temp diff Home W Home L Home T Pct
0 to 5.0 348 305 2 0.533
5.1 to 10.0 303 258 0 0.540
10.1 to 15.0 161 103 0 0.610
15.1 to 20.0 170 109 0 0.609
20.1 to 25.0 132 81 0 0.620
25.1 plus 124 75 0 0.623

Now, let's cross-reference both distance and climate differences, and see what we get.

temp diff 0-400 401-800 801-1200 1201-1600 1601+
0 to 5.0 0.528 0.561 0.526 0.450
5.1 to 10.0 0.515 0.455 0.552 0.613 0.605
10.1 to 15.0 0.517 0.646 0.631
15.1 to 20.0 0.646 0.574 0.531 0.652 0.667
20.1 to 25.0 0.660 0.625 0.571
25.1 plus 0.650 0.608 0.600

This seems to show that both distance and climate matter. At distances less than 800 miles, so long as the temperature difference is not extreme, the home field advantage is very small. Conversely, at longer distances, so long as the playing environments are very similar, the home field advantage is very small. Divisional opponents from similar playing climates, but at distances beyond 800 miles, have a home record of only 65-64-2. Consider the farthest single divisional matchup in terms of distance, San Fransisco versus Carolina in the pre-2002 NFC West. Despite the change of three time zones and over 2300 miles, the two cities have almost identical in-season average temperatures, and the road team won 8 of 14 games.

However, at longer travel distances, once the cities/playing environments start to deviate in temperature beyond five degrees difference, the home field advantage does kick in and increase dramatically.

To demonstrate the strong effect of playing climate change, here is a look at the home team records in series between the current NFC North teams, since 1961. In 1961, Minnesota joined the NFL, and the 4 teams have been divisional rivals since that year. Following the 1974 season, Detroit moved from Tigers Stadium to the Silverdome. Following the 1981 season, Minnesota moved from Metropolitan Stadium to the Metrodome. The Bears and Packers have played outside, as the Good Lord and Lombardi intended it, the entire time.

team 1 team 2 years Home W Home L Home T Home Pct
Chicago Detroit 1961-1974 13 14 1 0.482
Chicago Detroit 1975-2005 40 21 0 0.656
Chicago Green Bay 1961-2005 45 43 0 0.511
Chicago Minnesota 1961-1981 22 18 2 0.548
Chicago Minnesota 1982-2005 28 19 0 0.596
Detroit Green Bay 1961-1974 11 12 5 0.482
Detroit Green Bay 1975-2005 41 21 0 0.661
Detroit Minnesota 1961-1974 11 15 2 0.429
Detroit Minnesota 1975-1981 11 3 0 0.786
Detroit Minnesota 1982-2005 26 21 0 0.553
Green Bay Minnesota 1961-1981 17 24 1 0.417
Green Bay Minnesota 1982-2005 29 18 0 0.617
Outdoor Outdoor 1961-2005 119 126 11 0.486
Outdoor Dome 1975-2005 149 82 0 0.645
Dome Dome 1982-2005 26 21 0 0.553

Note that the home team is only 119-126-11 when the two teams both played their home games outdoors. Granted, during some of those years, there has been a significant power difference between the haves (Green Bay of the 60s, Minnesota of the 70s, Chicago of the mid-80s) and the have-nots (Detroit, and with condolences to MDS, pick your decade). However, when upsets have occurred, they have been more likely to occur on the better teams home field. Consider 1966 and 1967. Green Bay went on to win the Super Bowl both years, but lost to a rather poor Minnesota team at home (while winning in Minnesota) both seasons.

If you asked most NFL announcers what the biggest factor in creating home field advantage was, most would probably attribute a large part of it to crowd noise. This research casts some doubt on that view. I am guessing that the crowd is no less loud (and probably is louder) in divisional rivalry games against close geographic neighbors, such Bears-Packers, Steelers-Browns, or Eagles-Giants. Despite "crowd noise" in these games where the teams are in close proximity and play their home games in similar environments, the home field advantage has been, and continues to be so far in 2005 extremely small.

This entry was posted on Tuesday, October 3rd, 2006 at 3:29 am and is filed under Home Field Advantage. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.