Figure 6: Development of the average number of minutes played (per match) for Euro 2008-
ь ,
Europeans and players of other nationalities. All observations except those of goalkeepers, and
Austrian or Swiss nationals are included.
2008 nationalities and an associated negative trend for players from other nations.34 This
, , , ∙ , , , u , , , . ,1 ,.„,о,
development is apparent in both seasons but considerably stronger in the 07/08 season,
when players in the treatment group also overtake players in the control group.
Figjure 7 shows the evolutions of the average numbers of goals per match of players
in the treatment group and the control group. Figure 8 tracks the developments of goals
per minute played for observations associated with more than 71 minutes on the field, the
threshold used for the regression analyses with per minute output moasuret as dependent
variables. As before, each Figure includes linear fitted trends in each season, and vertical
lines indicating the period of final Qualificcition detisions.
Figuae 7 shows that the average number ot goals per match increased for both groups
over the course of the 06/07 season. The two trend lines are approximately parallel
although, as illustrated in Figure 6, players in the treatment group on average gained
minutes relative to players in the control group. In the 07/08 season Jtreceding the Euro
2008, oil the cttlitr hand, players in the treatment group dramatitahy improved relative
34We will discuss substitution effects between different subgroups of player, which suggest that such
opposite trends are natural, in more detail below.
28