User:BetterVotingAdvocacy/Big page of ideas: Difference between revisions
Content added Content deleted
| Line 80: | Line 80: | ||
One way to explain [[normalization]] for scored ballots is that a voter attempts to put the maximal margin between every pair of candidates, while still preserving their relative strength of preference between each pair of candidates. Specifically, this means that you try to give your favorite the maximal margin (1 vote i.e. [max score - min score] points) against your least favorite. |
One way to explain [[normalization]] for scored ballots is that a voter attempts to put the maximal margin between every pair of candidates, while still preserving their relative strength of preference between each pair of candidates. Specifically, this means that you try to give your favorite the maximal margin (1 vote i.e. [max score - min score] points) against your least favorite. |
||
One way to understand why a voter should give maximal support to their lesser evil if their favorite isn't viable: if, whenever a candidate is very unlikely to win, the voter pretends they aren't in the election i.e. got eliminated, then eventually the voter will have a candidate who is their "favorite of the remaining candidates". If they are normalizing, then they ought to give this candidate the max score, and likewise give their least favorite of the remaining candidates (i.e. the greater evils) a 0. |
|||
When designing a digital interface for the rated pairwise ballot, it's possible to use various software features to highlight the constraints on how a voter can vote. For example, if the voter submits an incorrect A>C preference, then their A>B and B>C preferences could be highlighted on the screen, with an indicator to show why the three preferences don't line up. Similarly, if the voter indicates a maximal margin for A>B and prefers B>=C, then the software can "fill out" the voter's ballot to put a maximal margin in favor of A in every matchup where the voter prefers B to the other candidate (For clarity purposes, the software should probably indicate, perhaps with colors or something, which matchups the voter filled out and which ones were auto-filled). |
When designing a digital interface for the rated pairwise ballot, it's possible to use various software features to highlight the constraints on how a voter can vote. For example, if the voter submits an incorrect A>C preference, then their A>B and B>C preferences could be highlighted on the screen, with an indicator to show why the three preferences don't line up. Similarly, if the voter indicates a maximal margin for A>B and prefers B>=C, then the software can "fill out" the voter's ballot to put a maximal margin in favor of A in every matchup where the voter prefers B to the other candidate (For clarity purposes, the software should probably indicate, perhaps with colors or something, which matchups the voter filled out and which ones were auto-filled). |
||