4 ms·

100% of Stein voters and 24% of Johnson voters in Florida would've flipped that state. 88.6% of Stein voters in Pennsylvania would've flipped that state. 20.8

100% of Stein voters and 24% of Johnson voters in Florida would've flipped that state. 88.6% of Stein voters in Pennsylvania would've flipped that state. 20.8% of Stein voters in Michigan would've flipped that state. 73% of Stein voters in Wisconsin would've flipped that state.

JoshuaDavid 8y ago
> 100% of Stein voters and 24% of Johnson voters in Florida would've flipped that state. ... assuming that the remaining 76% of Johnson voters abstained. If 50% of libertarians and 100% of green voters voted Democrat, and the other 50% of libertarians voted Republican, Michigan, Pennsylvania, and Wisconsin would have gone to the Democrats, and nothing else would have changed. If 30% of libertarians and 100% of green voters voted Democrat, and the other 70% of libertarians voted Republican, New Hampshire would have gone to the Republicans, and nothing else would have changed. I suspect 30% is closer than 50% in terms of how many libertarians would have voted Republican if there was no libertarian option. Source on calculations: run the following in the console on https://en.wikipedia.org/wiki/2016_United_States_presidential_election https://en.wikipedia.org/wiki/2016_United_States_presidentia... // What fraction of Libertarians would vote Republican if they had // to choose between Democrats and Republicans var fracLToD = 0.3, fracLToR = 1 - fracLToD; // Same for Greens var fracGToD = 1.0, fracGToR = 1 - fracGToD; // Find the table var electionResultsTable = [].slice.call(document.querySelectorAll('table.wikitable')) .filter(t => t.querySelector('thead tr th.headerSort') && t.querySelector('thead tr th.headerSort').textContent.match(/^State/))[0]; // Process [].slice.call(electionResultsTable.querySelectorAll('tr')) // Get rows .slice(2, -1) // Chop off header and footer .map(row => [].slice.call(row.querySelectorAll('td')).map(td => td.textContent)) // Get cell contents .map(cs => ({ "state":cs[0], "d":parseInt(cs[2].replace(/,/g, ''))||0, // democrats "r":parseInt(cs[5].replace(/,/g, ''))||0, // republicans "l":parseInt(cs[8].replace(/,/g, ''))||0, // libertarians "g":parseInt(cs[11].replace(/,/g, ''))||0 // greens })) // grab the cols we care about .map(({state,d,r,l,g}) => ({ state, d, r, l, g, // winnerWith3 is the winner as-is, with third parties winnerWith3:d > r ? 'd' : 'r', // winnerNo3 is the winner as-is, with third parties winnerNo3:d + l * fracLToD + g * fracGToD > r + l * fracLToR + g * fracGToR ? 'd' : 'r' })) // And grab the states where excluding third parties would have changed the results .filter(({winnerWith3, winnerNo3}) => winnerWith3 != winnerNo3)

JoshuaDavid 8y ago

> 100% of Stein voters and 24% of Johnson voters in Florida would've flipped that state. ... assuming that the remaining 76% of Johnson voters abstained. If 50% of libertarians and 100% of green voters voted Democrat, and the other 50% of libertarians voted Republican, Michigan, Pennsylvania, and Wisconsin would have gone to the Democrats, and nothing else would have changed. If 30% of libertarians and 100% of green voters voted Democrat, and the other 70% of libertarians voted Republican, New Hampshire would have gone to the Republicans, and nothing else would have changed. I suspect 30% is closer than 50% in terms of how many libertarians would have voted Republican if there was no libertarian option. Source on calculations: run the following in the console on https://en.wikipedia.org/wiki/2016_United_States_presidential_election https://en.wikipedia.org/wiki/2016_United_States_presidentia... // What fraction of Libertarians would vote Republican if they had // to choose between Democrats and Republicans var fracLToD = 0.3, fracLToR = 1 - fracLToD; // Same for Greens var fracGToD = 1.0, fracGToR = 1 - fracGToD; // Find the table var electionResultsTable = [].slice.call(document.querySelectorAll('table.wikitable')) .filter(t => t.querySelector('thead tr th.headerSort') && t.querySelector('thead tr th.headerSort').textContent.match(/^State/))[0]; // Process [].slice.call(electionResultsTable.querySelectorAll('tr')) // Get rows .slice(2, -1) // Chop off header and footer .map(row => [].slice.call(row.querySelectorAll('td')).map(td => td.textContent)) // Get cell contents .map(cs => ({ "state":cs[0], "d":parseInt(cs[2].replace(/,/g, ''))||0, // democrats "r":parseInt(cs[5].replace(/,/g, ''))||0, // republicans "l":parseInt(cs[8].replace(/,/g, ''))||0, // libertarians "g":parseInt(cs[11].replace(/,/g, ''))||0 // greens })) // grab the cols we care about .map(({state,d,r,l,g}) => ({ state, d, r, l, g, // winnerWith3 is the winner as-is, with third parties winnerWith3:d > r ? 'd' : 'r', // winnerNo3 is the winner as-is, with third parties winnerNo3:d + l * fracLToD + g * fracGToD > r + l * fracLToR + g * fracGToR ? 'd' : 'r' })) // And grab the states where excluding third parties would have changed the results .filter(({winnerWith3, winnerNo3}) => winnerWith3 != winnerNo3)