STLR voting

STLR voting (pronounced 'Stellar Voting') is an electoral system for single-seat elections, though it can be extended to a Multi-member system with a sequential elimination method. The name stands for "score then levelled runoff", and can be thought of as a utilitarian version of STAR voting. It is a type of cardinal voting electoral system and while the name is a clear reference to STAR voting it is actually a caomprimes between STAR voting and Score voting.

Voters cast ballots as in Score voting, rating each candidate from 0-5. The two candidates with the highest total are selected as finalists, and then in the "levelled runoff", the scores are levelled such that the higher of the two one each ballot is the maximum. The ratio between the two finalist stays the on each ballot but the ballots are levelled between one-another to ensure equal influence.

Method

1. STLR voting uses a ratings ballot; each voter scores each candidate from 0 to $MAX$

   • A $MAX$ is typically either 5 or 9
   • Candidates who are left blank receive a 0.

2. The scores for each candidate are then summed
3. The two candidates with the highest sums go to the runoff.
4. On each ballot, the scores these two are levelled by multiplying by the number such that the higher scored candidates score equal $MAX$

   • Explicitly, if the score for the candidates are $a$ and $b$ then they are levelled to $a_{new}=a{\frac {MAX}{max(a,b)}}$ and $b_{new}=b{\frac {MAX}{max(a,b)}}$

5. The levelled scores for each candidate are then summed and the winner is the one with the higher sum

Usage

The concept was invented by Equal Vote Coalition Director Keith Edmonds and was first proposed publicly in July 2020.

The originally proposed runoff normalization was that of Instant Runoff Normalized Ratings but was later modified to its current form. The idea came from an attempt to solve issue of arbitrary scale in Score voting. STAR voting solves this with the majoritarian runoff but as a result makes the system a majoritarian system.

Scale in Score Voting

When analyzing the ballots from the French studies and primary election for the Independent Party of Oregon .

Problem of being Majoritarian

STAR is intended to be a compromise between score voting and instant runoff voting.. It is a majoritarian system in the example

Red 51%: A[5] B[4] C[0] D[0] Blue 49%: A[0] B[4] C[0] D[5]

It elects A when both Score and Approval voting elect B. STAR argues that it is better because it recovers the way that people would have voted given the top two utilitarian winners. STAR is basically [| Cardinal Baldwin], but only doing the last round instead of all rounds. It is ONLY applying Baldwins method in the last step to normalize without losing monotonicity. Maybe the Baldwin normalization is not the best form of normalization since it reintroduces polarization and majoritarianism.

One can still solve the scale problem in score voting without resorting to majoritarianism with a different normalization. STAR put the favoured candidate to the $MAX$ but also puts the other candidate in the runnoff to 0. In orger to preserve utility the amount of utility that this candidate is worth to the voter needs to be preserved. Some assumption needs to be made and the assumption for STLR voting is that that relative utility should be constant.