Vox has this fascinating story about a poll conducted November 15-16 for President, using instant runoff voting, condorcet voting, approval voting, and range voting. This article is valuable not only for the presidential results, but for its clear explanation of each type of voting system. Thanks to Alex Hammer for the link.
An interesting article, although I did notice a bias in the presentation. They only showed bar graphs for the methods that resulted in Clinton’s election (instant runoff, condorcet voting, and approval voting). They did not for the two methods that showed someone else winning: Trump with range voting and Gary Johnson with Borda voting.
At least the Vox folks have some brain cells about Condorcet — HEAD to HEAD — math —
which has been around since the 1780s (repeat 1780s).
C math is akin to calculus compared to simple addition math.
BTW, someone pointed out to me a problem with instant runoff voting. It seems not appropriate that the votes of the least popular candidate get transferred first. He suggested that a better alternative would be for the 2nd votes of the 3rd place candidate be transferred first, and then so forth in order of descending vote totals. I have to admit that as a Johnson voter, I like that idea.
I applaud Vox for bringing up the topic of alternative voting systems but saying this candidate or that candidate would have won under one of these alternative systems relies on one big false assumption – that people would have voted the same way. Just consider the election using instant runoff voting – how many more voters would have cast a first preference vote for candidates other than Clinton or Trump?
Walt,
Let’s say that under an IRV election we have the following initial vote results for each candidate:
A 26
B 24
C 23
D 7
Why should candidate C be eliminated (before D) and have the second choice votes of C’s supporters transferred. It’s clear that D cannot win – but it is very possible in this example that once D’s ballots are transferred, C or B or both(depending on the second preference on D’s first choice ballots) could move ahead of A in votes.
@Eric L: True that. True also that candidates could and probably would have campaigned differently — and ditto for if the Electoral College were eliminated . . . or if all 51 states’ electoral votes were allocated proportionally (as arguably required by Section 2 of the 14th Amendment; look up Asa Gordon and the “Mal-Apportionment Penalty”).
Irrelevant since we live in a Republic, not a Democracy. We vote for presidential electors and THEY vote for president. To be anywhere near accurate, you’d have to do this “study” all over again with all of the names of every presidential elector in the country. A more meaningful study would be to tabulate the electoral college results if we had Maine/Nebraska-style district election of presidential electors nationwide. I predict the results would be quite opposite of collective approval voting.
Number votes ONLY show relative stuff — NOT YES approval or NO non-approval.
YES or NO does not show relative stuff.
Thus – there should be both — YES or NO and Number Votes for candidate choices.
Office
YES NO John Q. Candidate
1 2 3 4 etc.
YES NO 1, 2 etc in a minibox
Obviously there would have been a MAJOR change of tone if ALL of the Prez candidates were in a NONPARTISAN election —
i.e. Clinton – Sanders – Trump – etc. yelling at each other for months.