Mathematically speaking...
My prediction is the vast majority of voters in the two major parties will willingly pick from the approved list of crappy choices, thereby sanctioning the system completely, and then complain about how they had to vote for the "lesser of X evils" for the next 4 years.
Having worked as a statistician, I feel your pain.
Statistically speaking, the only way to accurately elect the candidate most desired by the people is to abolish all parties and for all people to rank order each and every candidate. The problem is that without primaries, you'd wind up with hundreds, if not thousands of candidates!
Fortunately, there's an easy, yet effective and highly accurate statistical approach to whittle down the the number of candidates to a reasonable size. It's called a "
simple random sampling."
You'd still hold a "primary," but that primary would consist of a small, representative sample of the population as a whole. The exact number of people in the sample population can be precisely determined by knowing the size of the entire population and the degree of confidence in the results, say, 99%. The result would determine the sample size needed to select the top ten candidates (or Pres/VP running mate pairs) across the board.
If you need to limit it to a subset of the population, such as "all registered voters," then you'd conduct a "
stratified random sample."
Surprisingly, it would only require somewhere between a few hundred and a couple thousand randomly sampled people to simply list their top ten choices. The results will be nearly identical to what would have resulted if you'd had everyone rank order every candidate. That's important, as it's accurate and reliable while remaining straight-forward and cost-effective. Of the top ten candidates, only a few will be strongly at the top. Those towards the bottom would never have been elected by any means.
KEY: You can NOT share the rank order of the top ten candidates! To do so would grossly bias the next step:
Once those top ten candidates are chosen, then you open the polls up to everyone. To virtually eliminate order bias, you simply print each voting card with the candidate's names ordered randomly. That is, each card would list the candidates in random order. Very seldom would you find two cards listing all ten candidates in the same order.
Voters would simply rank order the candidates. When averaged, the leading candidate is the one that's elected. Fairly, inexpensively, and without the massive bias introduced by our current two-party system. You'd still have biases, including those relating to funding, cognition, emotion, and others, but the only way to eliminate these would be to limit funding and strip all identity from the candidates. That, however, would introduce severe propaganda bias, so it's best to leave that one alone.
Various consulting groups and think tanks have capitalized on this understanding, beginning in the 1970s, to facilitate group decision-making with a far higher degree of accuracy than the outmoded board-meeting.
It's high time we drag our country's grossly antiquated voting system out of the 18th Century and into the 21st Century, where it belongs. Our Country's future depends on it.