We had previously published an article on Home Field Advantage by week examining whether on a league wide basis there was any variance due to the stage of the season. The initial conclusion was that yes, there did seem to be a more pronounced advantage for the home squads in the later weeks of the season.

In particular, over the 1999 to 2002 sample we used, the weeks one through twelve averaged a mere 1.3 points better net for the home sides, while the week thirteen through seventeen games had a whopping 5.3 points per game for the home teams.

For the retest we looked at the 1983 to 1998 time frame, but elected to throw out the 1987 season ("the strike year") and only look at regular season games. Of course, there was some variation in the number of weeks in a season (16 from 1983-1989, 18 in 1993, and 17 in the remaining years). We did not however adjust based on this, treating all week ten games for instance as 'week ten'.

Analysis: While not as striking as the more recent years, the long range data view shows that there is something to the notion of the home field advantage being stronger in the later weeks. This is evident in both the results against the spread, as well as the raw home minus away scores.

If we were to try and put together a simple, yet usable adjustment to add in for most projection purposes, based on the above we would suggest something like this:

For those sticklers for all the details (don't sweat it, we're the same way), here's the chart for the combined samples covering 1983-2002 (excluding 1987):