I was curious as to how well the domestic attendance percent changes lined up with the TEA data (also what is on wikipedia). Disney reports fiscal years, while TEA is calendar years, so there's the initial error of not measuring the same thing. However, the percent changes in the TEA data track the annual report data closely. The biggest discrepancy is 2017, but that is likely due to TEA including 3 more months of Pandora driven attendance (Oct-Dec 2017), where Disney's fiscal year ended Sept 30.