I’ve started doing some work on an expansion pack for the game which will be called: Democracy 4: Voting Systems which expands on the way elections are handled in the game. here is a rough bullet point list of new stuff I intend to include:
“Minimum Voting Age” policy
“Maximum Voting Age” policy
“Ban/Restrict Robocalls” policy
“Ban/Restrict Political TV ads” policy
“Government ‘get out the vote’ campaign” policy
“Compulsory Voting” policy
“Voter ID” policy
“State funding of political parties” policy.
“Ban/Limit corporate donors” policy
“Limit party donations” policy
“Mobile phone/online voting” policy
There will also be new features:
A new option on the polling screen will let you switch to see the approval/voting decision lines for likely voters, rather than all of the electorate.
The election results will now show a count of the voters who were disenfranchised by min/max ages.
A choice between FPTP (constituency/state based) or Proportional representation systems will change the way people vote, and boost turnout, and boost turnout for smaller parties.
The player will be forced to pick a single metric to pay attention to in campaigning and their public image, such as ‘Crime’ or ‘Unemployment’. This will boost voters sensitivity to that specific metric for the duration of that term.
A new pre-election campaign screen will show lots of election-specific data, and allow the player to decide on how aggressive and negative to campaign, with a potential boost to effectiveness, but potential impact on trustworthiness if it backfires.
Various policies will now combine with the mission-set level of ‘apathy’ to affect the likelihood of voting.
So far I have a lot of the policies done, but not the other stuff, and obviously no testing done yet Here is the voting results screen with the new disenfranchise count at the bottom:
Pay no attention to the pitiful turnout… thats because none of this is balanced yet!
I am still also working on bug fixes and improvements to the base game, this is running alongside that. It also doesn’t rule out the other DLC ideas initially discussed, its just that this was the most popular choice in the in-game vote by players.
Sounds like good material! One concern I want to bring up is bureaucracy. If the DLC increases the number of policies the player needs to manage then make sure to increase the threshold to trigger bureaucracy when the DLC is active.
I guess most of new policies aren’t going to be uncancellable ones. I was worrying that the voting expansion might add too many of them and end up making the main UI too cluttered. I expect 4 ~ 5 voting policies will be in effect at start.
Worth noting the comment above by Rabid regarding Bureaucracy situation. Maybe there should be bureaucracy modifiers for each DLC enabled.
Polling results of likely voters sound nice. It might be less volatile compared to the current approval ratings (popularity) since those who don’t hold a strong view on the government are less likely to vote. I’ve watched some streams and seen several absurd approval changes in a short period. Hope this will help.
Not sure how a choice between FTPT/Proportional Representation will affect the gameplay. I do acknowledge that it does have an impact on voting decisions and possibly strategic voting behaviors, the biggest difference it brings would be actual gains of seats or electors in my opinion.
‘Target Metric’ is an interesting concept but I think there’s a high chance it just being an easy approval grab as any player should have at least one very favorable & stable simulation (crime for example).
Thanks for the feedback. Actually most of the policies are not enabled by default. Its assumed that there are no restrictions on max vote age, robocalls, ads, or political donations unless the policy is in place, but I can easily nudge up the bureaucracy thresholds if the DLC is present.
I think FPTP/PR has 2 effects:
More turnout for PR systems.
More likelihood of voting for 3rd parties in PR systems. (not sure how to implement this yet… but I will find a way). Thus more likely to get a coalition in a PR system.
The idea behind the target metric is that it has to be picked a long way before the election. I’m thinking of real world examples like Tony Blair saying ‘tough on crime, tough on the causes of crime’ or George Bush with his ‘read my lips. no new taxes’.
More recently the UK also had this ‘triple lock’ promise to keep pensions high.
To make this distinct from manifesto promises, its going to be limited to a single selection from a short list of key metrics, maybe:
GDP
Health
Environment
Education
Technology
Immigration
Crime
And the effects of those policies are then just magnified, where they affect voters. So you may be trying to boost immigration, or cut it, its up to you, but the voters affected will all be very aware of the impacts (The effects will get magnified). There will be no opportunity to change the focus until the next election run-up.
Wow. Literacy tests would be interesting, but we do not currently measure each voters individual IQ or education/literacy. That then opens up a complex system as to how education has been distributed.
I guess we could assume its distribution is linked to equality?
I guess State Schools or School Voucher will have greater impacts on literacy distribution but economic inequality will have a say too. I wouldn’t be surprised to see such tests suppressing poor voters as they are effectively poll tax for voter registration 2.0 imo.
Theoretically, regularly checking the discrepancy between the light blue and the normal dark blue (not shown in screenshot) will give you clues as to if you want to encourage a high or low turnout in the election…
Wrote some more code for this today. By default in Democracy 4, at the start of a game, in a 3 party system there is a lower approval threshold and a higher one, set to 40% and 60%.
Those with approval <40% vote for opposition, 40-60% vote for the smaller 3rd party, and above 60% vote for you.
Now if you switch to PR, those thresholds are even spaced (33% approval each).
This should mean mean more coalition governments in a proportional representation system.
I am currently setting the cost of a voting system change at 20 political capital. In theory this might be higher, in terms of controversy, but I doubt a player would select it as a choice if its made too hard to do. The impact on the game is not earth shattering, unless I guess you are currently FPTP, and are worried about losing, but think an election fought on PR might let you squeeze back in narrowly with a coalition
20 feels too cheap. Actually changing the mechanism of democracy should be quite controversial, and would take a lot of work to get done.
In addition to FPTP and PR, how difficult would it be to model a Single Transferable Vote system? By this I mean a system where instead of X-ing one box the voter ranks the candidates 1 through whatever number of candidates.
While I see what you intend with these thresholds changes, I’m not really sure that they will properly depict a third party that loses votes to other major parties under FPTP system. You should be aware that typical winning players have approval distribution ranging from 40% ~ 100%, meaning almost no vote goes to the supposedly first opposition. Therefore, adjusting the thresholds to 33~66% under PR system will only make the largest opposition (called the third party :P) slightly more popular, squeezing some votes from smaller one aka the ‘first opposition’ party.
Confusing to read, isn’t it? But that’s just how the third party works in D4. Even 40% ~ 60% is way too wide to place it to the third place as its name dictates. Anyone who took some time to check the polls will recognize that the third party starts with lots of party members and has high support (>80%).
I’ve recently tried adjusting thresholds to 48% & 64% but they were still not enough to give some rooms for the (supposedly) first opposition. I’m going to try more radical changes such as 50-60 (or even 54-58 for FPTP). I’m also considering making changes on VOTE_CHANCE_FEELING_SCALAR so that middle ground voters (potentially 3rd party supporters) can become less likely to vote under FPTP. However, widening the range to 33-66 is definitely not on my list for experiments.
I’d also like to say that I agree to Rabid regarding PC cost on voting system reform but I assume them to be ‘not final’ :D.
Let’s see an example regarding threshold changes. I did a short do-nothing challenge. Technically I did some things such as choosing dilemma options, enacting some policy suggestions from donors, using electioneering tools, and filling vacant seats in cabinet when someone resigns. But nothing else was allowed.
The last poll before the election said my average approval was 59%. Here comes the result.
(image) Approval Distribution w/ normal dist., cell based
I drew a grid so that I can count them easier. I was able to check approvals of individual voters but you can’t do the same with the image. Therefore I marked the approval ratings of the median voters on each cells.
Did some summaries in Excel too. Setting the thresholds at 54-58 made the third party quite small. I think this is what a typical third party would face under FPTP system. The main opposition has shown a marginal opportunistic behavior (+1%p on Lower Threshold I guess).
The third party can become as big as the main opposition under specific conditions. But I’d still expect it to remain in third place as the main opposition is likely to be opportunistic. Opportunism of just +3%p would give the main opposition a huge advantage over the third party.
Then, what would have happened under the default thresholds of the base game?
The third party will take an electoral victory. Yes, I would have lost this election without the threshold changes. However, the main opposition is mostly guaranteed to be opportunistic and it would have squeezed some votes from the third party. So the margin would have been really small. Anyway, the third party is definitely not in a position to lose votes under FPTP system.
The last one is ‘What If?’ scenario with 33-66 mentioned by Cliff.
Third-party voters under PR system created a landslide victory. No amount of opportunism (of the main opposition) can save my party.
I definitely see what you mean by 33-66 thresholds under PR system. Third party voters will cast their votes without considering ‘strategic’ choices of supporting a lesser evil. But it just doesn’t fit well with gameplay experiences.
Yup, its worth pointing out that when I talk about the opposition party, and then the 2nd opposition party, this isn’t to assume that the first opposition party will always be the ‘main’ opposition. They may well end up the largest party., They are just ‘ideologically’ midway between you and the first opposition party.
I guess where it gets complex is deciding exactly how the thresholds change. I have arbitrarily set things up in the game with starting thresholds of 40 and 60 in a 3 party FPTP system. This means that the ideological opposition is always going to be more popular than the centrists. It could just as easily be set at 20 and 60, giving much more ‘power’ to the centrists…
of course in the PR system, every vote counts, so we end up with a 33,66 split with each party equally likely to get someone’s support, assuming their views are truly random.
This is all fascinating, because it makes me realize that the base game as it stands right now, without this DLC could actually be artificially reducing the chances of a coalition. My stats show relatively few coalition governments, but thats with the 40,60 split. a 20,60 split could result in lots more, assuming a normal distribution of opinion…
I wonder if I should include the values for
THIRD_PARTY_HIGHER_THRESHOLD
THIRD_PARTY_LOWER_THRESHOLD
as values an interested (but not file-editing savvy) player might want to adjust.
Well, at least in my experiences, I have rarely seen the first opposition (more radical one) winning the first place in an election. Even in a do-nothing challenges, you can maintain average approval above 50% if you did okay in electioneering. So 40-60 region is very likely to include significantly more voters than 0-40 region, making the centrists more popular than the radicals in most cases.
So I’d say that 20-60 idea is just slightly less devastating than 33-66 since <40% approval region includes just a handful of voters. The first opposition can’t simply win in a normal condition. The government should mess a lot of things and manage not to get assassinated. If you are going to implement ‘no more strategic voting’ under PR system, it’s the first opposition who needs benefits. 20-60 will only result less coalitions since the moderates will simply obliterate the government if an opportunity is presented, without messing with the radicals.
+Comparing chances of coalitions between 40-60 & 20-60 is actually simple. The governing party will get the same amount of votes so the question is 'whether players can perform mediocre so that they can’t won a majority while not getting defeated by the moderates. You should be able to see why 20-60 is less likely to lead a coalition government.
Check charts below. All of them were collected from do nothing challenge under 54-58 thresholds.
Case 1, won by a bit of majority, shown in my previous comment
Here is the voting results screen with the new disenfranchise count at the bottom:
