5 ms·

"Sanders won a plurality, but tellingly he did terribly in second alignments." <-- This is a lie, by the way. Second alignments benefited Sanders more than almo

"Sanders won a plurality, but tellingly he did terribly in second alignments." <-- This is a lie, by the way. Second alignments benefited Sanders more than almost any other candidate: https://www.npr.org/2020/02/08/803953713/graphics-how-iowas-vote-totals-shifted-from-the-1st-alignment-to-the-2nd https://www.npr.org/2020/02/08/803953713/graphics-how-iowas-... The poster I'm replying to often sneaks lies into his wall-of-text posts when he thinks no one will notice or call him on it.

rayiner 7y ago
Your own link shows that people went to Buttigieg more than 2:1 over Sanders in the second alignment. Apart from that, the movement in vote total between the first and second rounds combines two effects, but we’re interested in just one of those. Your analysis of the second round data marks a positive count for a candidate where someone moves to that candidate in the second round from a candidate that didn’t hit the 15% threshold in the first round. (That’s the only way for someone to switch.) But it also marks a negative count where a candidate doesn’t meet the 15% threshold in a precinct and his or her supporters have to move to another candidate. We don’t really care about that second effect, because that just measures whether a candidate is able to consistently get above 15%. What we want to predict is, as you knock out Biden, Warren, etc., where do their supporters go? You can find that out directly from the released caucus data, by looking at where supporters of those candidates go in precincts where those candidates don’t hit the 15% threshold. Most of those supporters don’t go to Bernie, which means Bernie is not their first choice or their second choice. Polling backs that up: https://civiqs.com/documents/Civiqs_ISU_banner_book_2020_01_cg3q62ypb43uz6bb.pdf https://civiqs.com/documents/Civiqs_ISU_banner_book_2020_01_.... Right before the Iowa caucuses, Sanders was the first or second choice of 37% of eligible Iowa primary voters. That means for 2/3 of the party, Sanders was neither the first nor the second choice.

rayiner 7y ago

Your own link shows that people went to Buttigieg more than 2:1 over Sanders in the second alignment. Apart from that, the movement in vote total between the first and second rounds combines two effects, but we’re interested in just one of those. Your analysis of the second round data marks a positive count for a candidate where someone moves to that candidate in the second round from a candidate that didn’t hit the 15% threshold in the first round. (That’s the only way for someone to switch.) But it also marks a negative count where a candidate doesn’t meet the 15% threshold in a precinct and his or her supporters have to move to another candidate. We don’t really care about that second effect, because that just measures whether a candidate is able to consistently get above 15%. What we want to predict is, as you knock out Biden, Warren, etc., where do their supporters go? You can find that out directly from the released caucus data, by looking at where supporters of those candidates go in precincts where those candidates don’t hit the 15% threshold. Most of those supporters don’t go to Bernie, which means Bernie is not their first choice or their second choice. Polling backs that up: https://civiqs.com/documents/Civiqs_ISU_banner_book_2020_01_cg3q62ypb43uz6bb.pdf https://civiqs.com/documents/Civiqs_ISU_banner_book_2020_01_.... Right before the Iowa caucuses, Sanders was the first or second choice of 37% of eligible Iowa primary voters. That means for 2/3 of the party, Sanders was neither the first nor the second choice.