The semi-final data accounts for the complete double header, right? So it's four teams contributing to the semi-final attendance figure, not two as in the championship figure.
Yup the Box scores for the first semi-finals game list no attendance, and the box score for the second semi-finals listed the attendance for the session / double header as a whole. It was relatively easy to find the attendance on the MAC site back to 2011. I had to start getting creative to find old box scores past that, so that's why I stopped at 2010. Here's the year by year attendance data that was used for the team breakouts. (Not an exact science but the best way I could think of.)
Finals
2019 - Buffalo over BG - 7813
2018 - Buffalo over Toledo - 5633
2017 - Kent over Akron - 10376
2016 - Buffalo over Akron - 6719
2015 - Buffalo over CMU - 5266
2014 - WMU over Toledo - 5278
2013 - Akron over Ohio - 12102
2012 - Ohio over Akron - 10301
2011 - Akron over Kent - 8926
2010 - Ohio over Akron - 9533
Semi-Finals
2019 - NIU / BG / Buffalo / CMU = 6044
2018 - Toledo / EMU / Buffalo / Kent = 4625
2017 - Kent / Ohio / Akron / Ball St. = 6065
2016 - Buffalo / Ohio / Akron / BG = 6427
2015 - Buffalo / Akron / CMU / Toledo = 6027
2014 - Toledo / EMU / WMU / Akron = 6318
2013 - Ohio / WMU / Akron / Kent = 10324
2012 - Ohio / Buffalo / Akron / Kent = 8513
2011 - Kent / Ball St. / Akron / WMU = 6174
2010 - Ohio / Miami / Akron / WMU = 5089
Edit: Also remember that the MAC did the double / triple byes from 2010 thru 2015. Some teams knew for a week they'd be in the semis.
Last Edited: 2/18/2020 1:07:57 PM by Eagle66