Hey, Brass - thanks for finding this! It's right up my alley, as I've been highlighting the inefficiencies of our two-party system for a long time, while fielding better approaches from industry and business (group decision making algorithms). Can't watch it tonight, but I'll hit it tomorrow morning.
In the meantime, I've studied many variations on the theme, most of which are subject to gaming by voters. For example, if you had five candidates and ranked your choices, a savvy voter would put their choice at the top, the most likely one of the others to win at the bottom. If everyone played fair or gamed it that way, it'd work. Problem is, only some folks game the system, thereby invalidating the system, by voting:
- First choice
- Fourth choice
- Third choice
- Second choice
One variation on this which has defied people gaming the system is this:
1. Each party is allowed to submit one candidate. Restrictions on minimum party size keep sagebrush from rolling into the game (keeps the number of parties to a reasonable few)
2. Each voter is given a choice between two randomly-selected candidates. They are to choose the one of the two they feel would be the better candidate.
This approach works very well for situations involving both a large number of voters and any number of options. One situation where we used it was to pick a new squadron patch. A dozen designs were submitted, and seven were approved by the commander's staff as candidates. I developed a program that would randomly print out two of the patches on each of two hundred sheets of paper. Each sheet was given to a member of the squadron who picked on patch over the other.
Had we done a ranked vote, the results would have been questionable.
The reason we needed to do this in the squadron is that we needed a no-kidding, non-gamed rank order of desirable patches in case the higher-ups disapproved one or more of our squadron's top choices.