Reciprocal Score Voting: Difference between revisions

No edit summary
 
(8 intermediate revisions by the same user not shown)
Line 33:
== Analysis ==
 
In the case of any asymmetry in support, the reciprocity ratio is <math>R(j,\phi) < 01</math> for the faction which did not cooperate, and <math>R(j,\phi) = 1</math> for the faction that did cooperate. Therefore, not cooperating penalizes the side which did not cooperate more. In this way, factions are encouraged to cooperate as much as possible to maximize mutual support, forcing them to strike a balance between supporting their favorite as well as supporting alternatives as much as they can. In the case of opposing factions, the mutual lack of cooperation has no effect.
 
This system is [[monotonicity criterion|non-monotonic]] and suffers from a very unusual "reverse [[spoiler effect]]" (see "chicken dilemma" below), in which a larger faction may lose an election by not supporting smaller supportive factions. Therefore, larger factions are encouraged to promote smaller factions as much as possible in order to win.
Line 41:
=== Strategy ===
 
It is always advantageous to top-rate a favorite candidate (RSV passes [[favorite betrayal criterion|no favorite betrayal]]), but the rating of other candidates requires assessment. Strategy under RSV depends on whether a front-runner is among a voter's top choices.
 
* If a front-runner is a top choice, under RSV a strategical voter has no incentive to reduce the rating of similar candidates, as in most systems. The opposite is true, in fact: they are encouraged to amplify their support towards allies as much as possible to obtain more of their support, while striking a balance so that their favorite remains a front-runner. Since faction ratings and mutual support are all interrelated, further strategy would require collaboration across competing groups.
 
* If any of the top choices is not the front-runner, the optimal strategy is to equally top-rate every "underdog" candidate deemed better than the front-runner (especially the favorite), and rating every other preferred candidate in order of perceived "viability". That is, the preferred front-runner would be rated as highly as possible below the favorite "underdogs", in order to maximize the support they receive from the front-runner's faction. Since the front-runner's faction is also encouraged to support the underdogs as much as possible, as per above, this is beneficial to all cooperating factions.
 
Thus factions which mutually collaborate have an advantage over opposing factions which are internally fragmented and uncooperative.
 
=== Exaggerated approval-like voting ===
Line 116 ⟶ 118:
 
Since the incentive for betrayal is eliminated, but the rewards for collaborating are preserved, Reciprocal Score Voting minimizes concerns with Chicken Dilemma scenarios. Thus in any sufficiently large mutual majority scenario, the likelihood of either A or B winning will be very large (although not guaranteed due to the non-ranked nature of the system).
 
=== Reverse spoiler effect ===
 
RSV suffers from a "reverse spoiler effect" in which a large faction may lose by not supporting a smaller faction. This can be a major problem when the smaller faction is a group with fringe or too radical beliefs compared to the main faction.
 
Since such fringe groups would likely be universally hated by every other faction, it may be the case that their support becomes effectively irrelevant in an election. In the extreme case if every other faction gives the fringe faction zero support, then the fringe faction's support to every other faction will be zeroed out. Thus, their effect on the election effectively disappears.
 
If, however, there is significant support for the fringe faction, one way to partially address the issue is to additionally adjust the reciprocity ratio factors to include the total ratings given to each faction, that is, if a fringe faction gets an overall low mean score, then that low mean score is used to lower the penalty of not reciprocating with that faction. This way, sufficiently fringe factions largely give asymmetric support towards more mainstream factions, making reciprocation with them less necessary. Between similar largely acceptable factions, which would receive similar overall mean scores, the effect cancels out and RSV proceeds as normal.
 
Therefore, by taking into account mutual reciprocation as well as overall ratings RSV can be made to promote reciprocation only between factions that have similar levels of overall support. It is unclear how severe this issue can be, however, and further investigations are warranted.
 
=== Passed/Failed criteria ===
295

edits