Injuries and home advantage in the NFL

In the first decade of this century players in the National Football League, the NFL community, fans, even the public at large, became aware that multiple concussions, heretofore considered inconsequential, could have devastating consequences later in life.

Since 1978, each one of the 32 teams in the NFL plays 16 games in the regular season. In the 25 years from 1978 to 2004 home advantage in the regular season tended to increase with Game Number (1-16). Then in the following decade (2005-2014) it changed direction and tended clearly to decrease. The change in direction was highly reliable statistically.

The result reported in this paper is an association in time between two striking events, a new consciousness regarding the long-term consequences of concussions in football, and a change in the course of home advantage in the regular season. The paper then advances a possible explanation for this association. The home advantage may be equally well treated as an away disadvantage, the disadvantage being that away players tend to feel on the defensive, that both the hometown crowd and the officials are against them. Injuries put players on both teams on the defensive. The higher the percentage of players on a team who are injured or playing hurt (injury prevalence) the less likely it is that as-yet-uninjured players will adopt an attacking style of play. Injury prevalence increases linearly with Game Number. It turns out, however, that formal considerations require that injury prevalence be the same or close to it for teams playing at home and teams playing away. Therefore, the away disadvantage in total defensiveness (defensiveness due to playing away plus defensiveness due to injury) starts at 1 in the first game of the season, decreases steeply at first, and then decelerates as it approaches .5. This downward course of the away disadvantage in total defensiveness leads directly to a corresponding downward course of the home advantage in game outcome (by the teamwork theory of home advantage).

Further research on the reported association or its explanation may be complicated by continued change in the association itself.