Banner

About

The Project

This project is being conducted as part of my (John Kirwan’s) thesis for the MA in Digital Media in the Huston School of Film & Digital Media in the National University of Ireland NUIG.

The Project’s key aim is to demonstrate a possible method for being able to educate people on the effects of different voting systems. Participation is often a much more effective teacher than theory; my hope is to test that theory.

Fans vs. Votes will also hopefully provide some interesting data on the different results of this election depending on the different voting system. This is often a point of discussion after an election, but with few experimental data it generally is people’s opinions, informed or otherwise.

Lastly, it never hurts to try sparking a little discussion in the most fundamental aspect of democracy: How shall we vote?

The Model

The Republic of Ireland, United Kingdom and New Zealand provided the base models for the Proportional, Majoritarian, and Mixed voting systems respectively. This was done because of the relative ease of access to information due to being English speaking Islands with a shared political history.

The systems were heavily simplified; attempting to replicate geographic and historic nuances was beyond the scope of this experiment. New Zealand's system was in particular impacted by this.

The model started from the minimum constituency in Ireland, which returns three elected officials. Using the general representative-to-population numbers, this led to requiring three equivalent UK constituencies. Finally the New Zealand model was factored in, however having been stripped of much of its unique qualities it is a rougher estimation. The main aspects stripped from the systems were the high-number of independent candidates in the Irish system and the Maori seats in the New-Zealand system.

The number of candidates was chosen by a mix of the ideal candidate numbers, determined by looking at past elections, and seeing how many currently running Tv-shows were available for each Genre. The shows were assigned randomly within their parties/genres to prevent an unrealistic scenario of on party only running in a single constituency.

The Voting System

Majoritarian

In the Majoritarian voting system Candidates are ordered into three categories to simulate the smaller constituencies/districts under First past the Post systems.

These are Junior, Regular, and Senior Representatives. They are ordered based on how many seasons the show has run for, ie. Simpsons is a Senior Representative Candidate with its 29 season and 629 episodes while The Last Kingdom is a Junior Representative Candidate with its 2 seasons and 16 episodes. Seniority is determined within the party, thus in some Constituencies relatively young shows are matched again long running behemoths. This somewhat replicates support in an area for one party being stronger than others to some degree.

Mixed

The Mixed System is a 1:2 split. Each Constituency returns are local representative, and each voter votes for their preferred party. After the four local Representatives are returned, the remaining 8 seats are divided proportionally among the seven parties.

The Representatives were chosen from the strongest Candidate of each party by user rating and number of reviews so all candidates were on roughly equal footing. This is a Majoritarian election, so the candidate with the most votes wins.

The Party lists were assigned solely in order of their user-rating, regardless of how many reviews they had. A party that receives 12.5% of the vote across all constituencies receives a seat with the remaining seats assigned to proportionally.

In this system two shows were added, Bob's Burgers for animation, and The X-Files for horror. Both genre-parties were too small to represent in every election, however for mixed system this was more punishing and was thus changed.

Proportional

The Proportional system is list based, each candidate is on the list and the list is ordered according to the voter’s preference. While in real elections a user need not vote all the way down the list for ease of use for users, and for ease of explanation, all candidates will be assigned a vote.

After each vote a winner will be determined by calculating if they meet the quota, which for this election is on quarter of the districts turnout plus 1. The surplus over this number will be then distributed proportionally according to the second preference of those candidates’ voters. After this the lowest ranked candidate is eliminated, and their votes are distributed proportionally according to their voter’s second preference.

This can potentially go on for some time, which will slow down the calculation of results. This will also require the results of each count to be published, as this can seem very strange without seeing the changes over each count.