Historical Attendance Flaws

[Ubiquitous]

Occasionally we have debated the popularity of baseball throughout history, and one of the points that gets brought up is attendance. I just realized that the figures we use are probably wrong. For instance the 1922 Yankees have a total attendance of 1,026,134 and they played 77 home games which gives them an average of 13,326 customers per game. That is the total that is listed on BRef. But here is the problem the Yankees played 5 home double-headers. So in terms of possible paying customer dates they only had 72 home games and their home date average would be 14,252 customers per date. A 7% jump in attendance rates.

Now it is possible that somehow these attendance figures double count the double header. Meaning if 20,000 showed up for the first game the stat trackers also count 20,000 for the second game. If that is the case then one should not subtract the double headers. But if that is what is happening then I think one should go back and subtract the attendance from the second game and only look at total attendance divided by dates.

But I have no idea how the final attendance numbers are figured I don't have access to the original source data for historical attendance figures so in the end I really don't know how they came to the final figure. I have a feeling that they don't count double headers attendance twice but they do count the second game as a home game.

[Brian McKenna]

sort of off the point, but

there were also many morning-afternoon doubleheaders in certain years - for example chi white sox 5/30/1901
game 1 attendance - 1,700
game 2 - 13,000

the current day-night doubleheaders are not new creations