And yes; one of the worst things that happened to Reddit was the proliferation of Vote-Up-If posts.
Here's I think it's interesting that we can basically run a survey experiment based on the choices offered. For instance, we can see that the age of YC.news users is a slightly skewed distribution with a mean of of maybe 24-25 (even as we could nail it down more with smaller bins). To me, that's much more interesting than "vote up".
Don't get me wrong, I like politics and I personally think Ron Paul has a lot going for him. But I didn't start reading the site for its political bent; I started reading it for the interesting links and the tech stuff. I'd simply like to get my political news elsewhere. With the bastardized reddit, I couldn't get away from politics, even with the introduction of sub-reddits.
I would have preferred ranges of four years, but I got an error with that many choices.
But with multiple selections in the general election we wouldn't need primaries and would get more centrist presidents.
How about 1992? Could Bush have won if angry republicans would have a chance to vent their anger by also voting for Perot? Or might Perot have won since voting for a 3rd party candidate wouldn't feel like a lost vote.
To demonstrate the problem with approval voting (all voting systems have problems, as was proven by Kenneth Arrow in 1951): if Obama, Clinton and McCain would all run, McCain would probably win, even if a majority of voters preferred either Democrat.
So, approval voting should allow fair elections with more than two candidates, but in practice it doesn't, because people vote strategically.
> (let voters (flat (map [_ 2]
(map [(item _) 'votes]
((item 126923) 'parts))))
(list (len voters) (len (dedup voters))))
(50 38)