So here's what we did: we took the goals data we collected for the six leagues we've been talking about (Bundesliga, EPL, Eredivisie, La Liga, Ligue 1, and Serie A) for the 1995-2010 period and calculated the average number of goals per match for seasons and leagues that saw 34 league matches and compared them to those that involved 38 league matches.
And lo and behold, averaged across the leagues and years, we see that there is a distinct difference between the two season/league formats: leagues where teams play 38 matches see 2.54 goals while leagues with 34 matches see 2.82. And that difference is highly significant, statistically speaking. So could this be behind some of the scoring differences we see across leagues?
One caveat: given the relatively small number of seasons involved, it's best not to make too much of any differences, but it's interesting to see nonetheless. So here are goals per match for each league separately, by the number of matches played in the season.
Overall, we do see that goal totals decrease with more matches played in three of the four leagues that experienced seasons of varying length. Goal averages are higher when fewer matches are played in Serie A, the EPL, and Ligue 1, while goal totals were up in La Liga during their 42 match seasons.
But I'm still skeptical, and here's why: the two leagues that have consistently played 34 matches also happen to be the highest-scoring 34 match leagues (the others are Serie A and Ligue 1 during this period), and there may be reasons having nothing to do with season length that drive these outcomes. So there's a good chance the pattern in the first graph is more marked because of their inclusion. And of course, we are not accounting for league quality and the extent to which some teams' seasons are actually much longer than 34 or 38 matches because of European competitions. Still, we thought the pattern was well worth thinking about ...

