"The data for this analysis was gathered from the platforms’ respective websites between 6:00 p.m. and 7:00 p.m. ET. " On what day?! And why not use all the prior days in the dataset for estimating accuracy?
Thank you for pointing that out, Robin. The data was collected on the 5th of November 2024. We didn't use all the prior days dataset because it's hard to extract the data from each platform and we see the comparison as snapshots. Maybe in a future analysis we could compare some selected days before the election to see how the predictions were changing. Still a static comparison though. Would you propose a different approach? We really appreciate your input!
"The data for this analysis was gathered from the platforms’ respective websites between 6:00 p.m. and 7:00 p.m. ET. " On what day?! And why not use all the prior days in the dataset for estimating accuracy?
Thank you for pointing that out, Robin. The data was collected on the 5th of November 2024. We didn't use all the prior days dataset because it's hard to extract the data from each platform and we see the comparison as snapshots. Maybe in a future analysis we could compare some selected days before the election to see how the predictions were changing. Still a static comparison though. Would you propose a different approach? We really appreciate your input!