1 - The game will agree to a set of "Election Admins". This can be a subset of the Mods, all the Mods, or different people altogether. Ideally this would be around 5-10 people, and represent all parties. These admins would be responsible for setting the "Modifiers" and as such, should be trusted community members who can and will act in a mature non-partisan manner.
SHOULD THE GAME DECIDE IT WANTS VOTING WITH MODIFIERS:
2 - Create a Voters List; a registry of everyone who plays the game.
Your name must be on the voters list PRIOR TO THE WRIT DROP to vote. No excuses. This will discourage spamvertisments.
3 - Each voter would live in a particular region. They could move at will, but not during a writ period.
4 - At election time, voters will cast their ballots for their party of choice.
5 - Each election admin will get to cast 1 ballot in each region, however they see fit. They may vote for the same party in each region, or, for different parties based on their personal ideology. They may also vote only for the parties they feel did a good job in the election. How they vote is up to them.
6 - The total vote for each party will be adjusted by the total modifier.
7 - The Modifiers will be as such
A - Activity. Share of registered voters that posted in-game at least once* since the last election. *(this can be changed to 2 times, or 3 times etc, if the admins think that's a much better measure)
B - Commitment. Share of votes each party could have given in Parliament, compared to the share of votes actually cast. (IE who misses votes)
C - Debate. Each party would be graded on each debate based on how many members involved themselves in the debate. This would be compared to the share of elected members that party has. If for example 7 NDPers engage in debate on a bill, and 7 Tories do; but the NDP has 14 seats and the Tories only have 7, the Tories will be considered to have been twice as involved in debate. This itself should be adjusted based on the quality of the debate. Members who only stop in to say "Hear Hear" could be counted at half weight, while those who post longer and thought out statements could be counted as double. How quality is weighted will be up to the admins.
D - Legislation. Not only will the number of bills passed count as a modifier, but the quality will as well. Again, exactly *how* this is all weighted will be up to the admins.
E - Overall Quality. This one is entirely subjective.
7 Cont - Each of the above would be equally weighted. Lets say that 22% of registered NDP voters are active, the NDP would then get "22" for this. If they missed 12% of votes, they'd get an "88". In debate maybe they get a "69" given how many people only post "hear hear" comments, however they end up with a "100" for getting the most legislation passed, and a "100" on quality as the entire admin team feels the NDP did an excellent job last term. The number should never go above 100, however, the admin team may decide to grade on a scale where the top party gets 100 and the bottom party gets 0, but it is highly recommended you do not allow one party to have 100 and another have under 50.
Their total modifier would then be "379". Lets assume the Liberal modifier in the election is "241", the Tories "332" the Socialists "283" and the Libertarians "309". As you can see this system will cluster these rankings to ensure one party does not get an insane amount of support.
You would then multiply the real vote result in each region by the modifier. So 17 NDP votes would become 6,443 while 20 Liberal votes would become 4,820. The final step is to total all the votes in an area, then present it as a realistic sounding share of the vote that this region would actually return in a real election. Lets pretend the total votes cast is 69, and this is Ontario; the NDP might thus be presented as having won 1,242,920 votes, compared to the Liberals at 929,826. Note that both the modified vote total and the 'end' vote total have the Liberals on 74.8% of the number of votes the NDP got, but the 'end' vote totals look more 'realistic'
SHOULD THE GAME DECIDE IT WANTS FULL SIMULATED ELECTIONS:
2 - Each election admin will state to the other election admins which party/person they feel should win each seat and why. They may use any means they wish to determine this, but are heavily encouraged to reward activity. The aggregate vote of all admins is then combined to produce a (hopefully) consensus based list of winners in each region; and from that list the narrative of the election is determined.
Activity counts, MP absences, Quality of legislation shall be published - but this is an accountability member. None of these numbers need to be used in determining the election result whatsoever. The accountability to be fair will be on the shoulders of each admin, and these admins may be removed by the community if they feel that they are not being fair.
There's a bonus to this
you can simulate real elections. Have parties make ads and post them in the game. Here are some examples from my past polsim:
(two ads shown on a single document; posters)
(lawn sign design)
(billboard; the man was a party leader)
Thanks to this, you can then judge "the campaign" as part of how well/poorly each party has done.