So let's suppose there are n candidates running for office. A bunch of random people begin to selectively vote the candidates, each vote being a number from 1 to n. What is a good way to order the candidates based on their vote number and also the number of votes/frequency of votes?
So for example: If candidate A was voted by a bunch of voters, but was the votes were fairly spaced out, he would probably be considered "not as great" as a candidate who was voted less/same amount/more but the votes were shortly after one another.
I was thinking of something like Reddit's Karma system, but that only takes into accounts +1's and -1's (and I'm not even sure if Reddit every gave information on how their Karma system works!)
So for example:
Candidates = [ Bob, Sally, Joe, Sam, Rob]
Person A ranks Sally as #1 (best choice) and [Bob, Rob] as #5 (worst choice) and leaves (he doesn't vote for others).
Person B ranks Sam as #2, Joe as #4, and leaves.
Person C ranks Sally as #1, Joe as #3, and Bob as #5.
So in this case it may be compelling to say Sally is close to being #1, while Bob is probably closer to being #4 or #5. But Sam was only voted once as #2 and as such there's not a whole lot of grounds to say "yes he's definitely #2". Similarly, Rob was only voted once as being #5 but there isn't sufficient votes to definitively say he's worse than Bob.
As such, the raking may become something like:
Sally > Sam > Joe > Rob > Bob
You can probably twist this example to have even more voters and create a few outliar, but I hope his gives you a general idea of what I'm looking for.
