Improving the long-term challenge of the game

Democracy 4 Alpha Feedback
1

One of my design concerns is that the game is too easy after the first term. This is something that is very hard to get right.
Fundamentally there is a design problem because you have to be popular enough to win an election, yet also be in a state where winning the second election is not guaranteed. I’d like to lay out some of my thoughts on this design dilemma and solicit feedback from players who are interested in this topic.

Existing systems:

  1. Long-term external factors like climate change and terrorism could automatically get worse
  2. Cynicism of voter groups could build up until they demand a change.
  3. Complacency of voter groups could build up making them harder to please
  4. Long term negative feedback effects of early game policies could cause problems- a booming economy causing house price issues, maybe high wages driving away business etc
  5. Opposition parties move closer towards you so that your unique selling point as a party is less valuable

These are the existing systems. I think they need refinement and balancing, possibly expanding, and new suggestions are welcome. Some details on each one:

Long-term external factors
These are effectively scripted, in that they progress as the years go on. Currently this affects the following variables: Terrorism, Industrial Automation, Lifespan, Internet Currency Adoption and OilSupply (decreases). It seems it does NOT affect GlobalCO2. Maybe this should rise steadily? I’d suggest none of these currently seem to be rising enough.

The year variable also affects Rare Earth Shortages, Water Shortages, Anti-biotic resistance, Mineral Wealth and Oil Wealth.

Cynicism
This is a weird one because I wonder if it should be rolled into the trust perception instead… Anyway, Cynicism is boosted when the player carries out a policy u-turn (cancel then implement or vice versa) or cuts taxes just before and election or boosts spending just before an election. Also, cynicism is boosted when an election pledge is broken. This is all in the game but we do not draw attention to it anywhere outside of the voter group charts, which likely needs changing…

Complacency
Happy voter groups get complacent over time,. This is all data controlled. If a voter groups happiness is above COMPLACENCY_INCREASE_LEVEL it raises by COMPLACENCY_ACCUMULATOR (or falls by a different amount, below a different level). Its max value is COMPLACENCY_MAX plus COMPLACENCY_TERM_ACCUMULATOR for each term up until COMPLACENCY_ABSOLUTE_MAX. Currently thats from 20% to 75% with 1% per term, so would take 55 elections! clearly wrong. Plus I think this should all be scaled a bit by difficulty surely?

Long-term feedback
This is where the real complexity should come in. We likely need to add a bunch of super-long effects, although the code does limit the total lag of an effect… The big question is what are the negative impacts of success? Maybe GDP growth leads to longer term gridlock, and low unemployment leads to higher wages and reduced competitiveness… house prices being unaffordable for the young as wealthy foreigners buy up desirable property? Maybe stability and democracy leads to house price boom (such as happened in London). Maybe homelessness should trigger more severely due to high house prices?

Opposition parties (opportunism)
This is newly modelled in the game. Basically the opposition parties will move towards the center-of-gravity of the electorate (opportunism). That means that if you have high poll ratings and win a lot of elections by a high margin, the opposition parties will move towards your position and ‘steal’ your policies, taking votes directly from you.

I welcome all suggestions, feedback and ideas on this issue. I want the game to maintain long-term playability for all players, in all play styles. I think even writing this down helps me identify some clear areas where some auto-balance or responsiveness to difficulty settings is required :smiley:

5 Likes
2

RE: Long-term external factors
My personal issue with these things is that they go up without any player input, or even with the player being totally the opposite. While there does exist a world totally external to the country, it feels frustrating to have all the work be done to cut CO2 levels, to prevent Industrial Automation, only for this mysterious “Year” to eventually creep up on me and tell me what I did was pointless. These factors I feel like require at least some way of reducing global adoptation, perhaps having something like the International Fusion Project, but for each of these things, in the form of a treaty. With greater expense leading to reducing global increase and, with a freakishly large investment, even reduction.

RE: Cynicism
Cynicism probably does need to be boosted, though more so for 3rd and 4th terms. US politics is proof that cynicism doesn’t really take hold in just a 2 term window.

RE: Complacency
I don’t think compacency needs to be boosted. It may make it more difficult, but it feels arbitrary. Why bother staying loyal to voter groups and making them happy if they will then NOT vote for you as a result? People vote to protect their happiness.

RE: Long-term feedback
Long-term impacts would definitely be good, but only if there are then solutions for them. Low unemployment leads to higher wages and reduced competitiveness, but an increased population (child benefits, marriage tax allowance, parenting classes, abortion law) will provide enough workers to make workers in the still country still compete enough to make wages lower. High house prices resulting in homelessness due to foreign ownership being able to be countered with the Restrict Foreign Investment policy, therefore slashing the foreign demand, and rebalancing the housing market to be solely domestic (making prices be forced to lower).

RE: Opposition parties
This is fine, though it would still be cool for parties to choose one or two voter groups to always represent and try to steal the popularity from you on

2 Likes
3

I think some areas of the game are too responsive- in particular GDP. It would help if a player making sweeping changes wasn’t immediately rewarded for their efforts, and the full effects took longer than a year or two to kick in.

It occurs to me that the business confidence variable might do to be affected by volatility in the player’s legislation. Not that that is relevant at all to this topic lol

4

I was thinking of saying something to this effect; a lot of policies in my opinion seem to be implemented rather quickly in comparison to real life. I’m just not sure how this proposed change will help with the long-term challenge. It may make the first election harder to win, sure, but then you’ll have a lot of policies partially implemented and those effects coming into fruition will just boost your popularity. I suppose if they come with their problems, then yeah, making implementation times longer would also delay those negative effects to be in the later terms. But then this will just delay the problem: the first two terms instead of the first one will be challenging to win, and in subsequent ones you’ll have smooth sailing.

1 Like
5

Its always a difficult balancing act with implementation times, because it goes against almost everything people assume to be good in game design, which is that the player needs to see a tangible response to their decisions :smiley: Obviously thats at real odds with the real-world that I am trying to model, especially in politics.

In UK politics, its pretty much a trope that the person who gets the blame for a recession or the credit for a boom is the person who comes after those responsible have lost an election.

I should do a comprehensive analysis of the implementation times of all policies to see if some can be extended, and of course there are also the effects as well as the implementation periods to consider.

Something I have not mentioned is the negative effects of an ‘overshoot’. So because it takes maybe 20 turns to implement a policy, and 10 turns for its effects to reach maximum, there is a tendency for players to set policies too high, then when they are finally firing on all cylinders, they may be too high.

So for example a curb on immigration may reach the desired effect just before an election, and win votes, but eventually goes too far and causes a skills shortage. (TBH skills shortage is likely a very under-triggered and much needed situation in these circumstances…)

3 Likes
6

Funny. I find that the game often doesn’t punish enough for “firing on all cylinders”. The effects of most policies seem to be all linear, thus requiring almost “no thought” for their slider positions. There are negative situations of course but often to prevent them the better idea is to maximize some new curbing policies or get completely rid of existing prime culprits. Yes, yes, I know there are policies where the optimum strategy isn’t to maximize, minimize or cancel them for the player’s goal, but diminishing returns seems too much of a rarity to see in this game.

It may be one of the ways to add to the difficulty, indeed. If it was harder for the player to come up with the optimum slider position, they would be more likely to overshoot or undershoot, thus either causing future problems or not getting as good of approval ratings as possible, thus making the opposition stand more of a chance in elections. Optimizing sliders would cost a lot of extra capital in future terms for small gains, giving time for complacency and opportunism to set in.

Combined with long implementation and effects coming into… effect, times there would be even more uncertainty about the optimal slider positions. I probably should find an explanation for the implementation delay, since I don’t pay much attention to it, and instead focus on the last effect to have its effect fully kick in, and that’s the “implementation time” in my usual parlance.

1 Like
7

Long-term external factors
I think this is perfect. Of course rising sea levels and natural disasters are the easy answers, but global industrial trends forcing automation on you besides increasing global tensions and foreign protest movements are also strongly in the news. The existing year variables are perfect, though most of them have very specific resolutions revolving around one or two policies, thereby not requiring an abundance of thought or a widespread change.
Perhaps eventually water rationing and space mining are needed for rare earth and water crises respectively? And so on.
If you’re serious about late-game functionality, I might suggest some kind of unlockable policy system, whereby you’re not able to implement a policy until you’ve met certain prerequisites of stability, or otherwise keep them unaffordable until you’ve established a strong enough government. Having too many policies crowding the list in the early game ought to be avoided.

Cynicism and Complacency
These are interesting concepts, but they’ve truly never affected my games. My main groups get fanatical and stay that way. Because of the ability to grow and shrink voter groups, the late-game tends to be a place where hostile voting groups are often decimated (Which almost universally includes the religious).
A way to reintroduce difficulty, nuance and dare I say realism could sit with nerfing these effects, so you’ve always got a few dissidents to convince around election time. At present I don’t even look at speeches after my first victory, because I know about 2% of the population is opposed to my shit by then.

Long-term feedback
A Democracy 3 mod theorised the return of the Skills Shortage as a result of an overly advanced country. Stem cell research, automation, carbon capture, space industries and tech advantage all boosted the Skills Shortage, meaning you needed every educational policy firing on all cylinders by the late game.
Another suggestion would be the Kesseler Syndrome, the singularity of space debris caused by too many poorly regulated satellites, currently hotly discussed in space science and global diplomacy.
Becoming a leading nation would also put a target on your back, reigniting Cyber Warfare (Russian bots anyone?), as well as hostile posturing and perhaps even proxy wars (Global warming is also all but certain to cause wars in the next few decades).
The fourth industrial revolution is also arguably more nuanced than ludditism. Beating back the riots doesn’t address the evolving nature of labour as machines consume manual and data entry roles. The re-definition of what constitutes payable work and fair standards must be addressed as it was in every industrial revolution hence.
Finally I would suggest global liberalism could apply increasing protest pressure over any of the above issues, as well as calling for reparative justice for historic sins and the third world. This has already pissed people off enough to destroy property this year, so I doubt it’ll blow over as we become an increasingly educated, egalitarian and liberal society.

Opposition parties (opportunism)
This is an interesting challenge that certainly serves to reduce the pool of available votes, assuming it cannot be defended against. Would hostile actions become available in electioneering to discredit or vilify those parties as a means of retaliation? Even without that, it’s a good way to thin the 100% lead you often accrue by term 3.

1 Like
8

Something I have not mentioned is the negative effects of an ‘overshoot’. So because it takes maybe 20 turns to implement a policy, and 10 turns for its effects to reach maximum, there is a tendency for players to set policies too high, then when they are finally firing on all cylinders, they may be too high .

There is no reason to set MOST of policies to max.
I think everyone does that - you could collect data on average policy, simulation and situation settings on each election.
I bet most of policies would have average setting of >0.95
There are few policies, that have drawbacks when setting them on extremes.

Debt still can be completely ignored, if you have enough policies to counter effects of debt crisis.
I bet you would be winning all elections, if you maxed out policies, that doesn’t have drawbacks :smiley:

Each election is easier and easier to win, unless you are trying to minimize expenses.

Technology, education and health are simulations, that are way too easy to max out, and you would try very hard to reduce them to 0.

2 Likes
9

Add Thorium Reactors, Electric Planes, Smart Cities, Smart Highways, Drone Delivery Systems etc. You can add a technology requirement so that players have to reach a certain technology level to implement the policies. This means that the players would have long term goals which the can unlock by keeping us technological progress and trying to win more elections.

1 Like
10

Foreign Military Threats could be an external factor you should look into. Maybe if a player has a very small military spending and a small population then they are prone to Ultimatums or invasions which come in the form of dilemas, emergencies and events and ultimately result in a game over as the play would not win the election due steady or sudden decreases in popularity.

1 Like
11

If you made a super policy out of Income tax, then you could attract players for the long game more as people will have more 3 dimensional policy goals for the tax. I call it a super policy as it would not have the same layout as others in that it would have several parameters: Numb of brackets, Income level for each bracket, and tax level for each bracket. This way you incorporate the flat income tax, negative income tax, and progressive marginal income tax policies into one. Making it a super policy is entirely appropriate as tax is one of the largest factors driving voting behaviour and is hugely politically and economically important. In the short game a player could have the goal of lowering tax by 10%, but in the long game a player can focus on sculpting society by having any tax policy the desire achieved through many gradual reforms.

12

I’ve referred to this in the past as the “pro-policy bias”. Policies in democracy tend to be positive sum so the dominant strategy in the game is just to throw more and more policies at everything.

4 Likes
13

I agree 100% that we need more non-linear effects and more interesting sliders in terms of optimum value. Its purely a matter of time, and multiple balance passes with me adjusting stuff and checking it doesn’t ruin everything :smiley:

1 Like
14

Yes, the game does have a pro-policy bias, and I would like to address this, although its tricky to do so without including a whole bunch of extra starting policies for every country (effectively government initiatives that could be abolished or defunded).

TBH we have under-utilised the games system for sizing of icons. Originally our concept was that by having variable sized icons, we could effectively support an infinite number of simulated variables and policies, but at any one time the real minor ones (for example the street lighting budget) would be small enough to be happily ignored in favor of the more crucial policies like income tax/defence etc.

I should probably revisit that a bit, so there is more scope to reflect a broader range of issues without clutter.

3 Likes
15

Add bureaucracy simulation then - every policy would have bureaucratic load, would be scaled like x^2 or something like that.

1 Like
16

I think the solution there is to make it more obvious that changes do indeed occur, even if they are slow. Some sort of projection mode “where will we stand in x turns if nothing changes (and probably not including any catastrophes)” and why is that where we’ll stand" would be good. So long as it’s easy enough to see even slow change, this should be fine.

This is a different effect. Note, people in charge do have the ability and ideally (but clearly not always) the expertise to rely on statistics to see some of the future repercussions. That’s what the player gets to see.
Meanwhile voters may well ignore relatively slow-moving effects (except for a brief spike right when it’s implemented, emulating news-coverage about it) but react rather strongly to fast-moving ones (because they are instantly felt)

This definitely does happen (and is chiefly because it’s currently hard to discover that you already did enough and a problem will soon solve itself) but truthfully, about 80% of the time, going either min or max on a policy just seems to be the right choice. It’s honestly not often all too interesting to explore the in-between. - More complicated effect and pricing curves (ballooning costs and diminishing returns) would certainly alleviate this though.

1 Like
17

Not only that. If you play a patient long game, you can, in fact, get rid of debt completely, saving a TON on interest payment, which ends up netting you a lot of extra room to actually get in quite expensive policies while remaining at a balanced budget.
This has gotten somewhat harder, but I don’t think it’s actually impossible yet. (And maybe it shouldn’t be impossible either. I’m not sure about that)

1 Like
18

Well there are countries with no debt, and actual sovereign wealth, so its definitely something that should be possible. I think the game should perhaps give you more hints about this, maybe a financial expert popping up saying “Hey you DO realize that if we lowered the debt this big payment would shrink right?” or whatever :smiley:

2 Likes
19

Here are some interesting player stats for the latest version (1.15). This is the median values at an election after turn 40:

Popularity 87.95 %
GDP 88.61 %
Unemployment 32.59 %
Debt 26.72 %
Poverty 8.31 %
Health 95.29 %
Education 97.18 %
Crime 2.91 %
Violent Crime 2.96 %
Technology 92.61 %
Equality 62.33 %
Global Socialism 35.24 %
Global Liberalism 84.21 %

Obviously popularity is pretty high but I think thats just due to the other variables. This is after maybe 2 elections so we are filtering out players who got kicked out of office before then…

My thoughts:

Education and Technology are WAY too high too soon. maybe some of those effects need to be slower, or implementation times of policies longer (or both).
Unemployment is actually quite high…but clearly thats not causing enough problems? not sure how to fix that without making the first election harder.
Maybe both crime and violent crime and too low.

Comparison with the same stats but after turn 21:

Popularity 86.25 %
GDP 89.25 %
Unemployment 32.1 %
Debt 26.68 %
Poverty 8.85 %
Health 90.38 %
Education 96 %
Crime 3.35 %
Violent Crime 2.68 %
Technology 91.09 %
Equality 58.76 %
Global Socialism 31.33 %
Global Liberalism 84.35 %

Now a table showing turns >14 >21 and >40

image

Definitely no signs of technology pushing unemployment up enough in the late game. Education definitely fixed too quick. technology maybe too quick too…
hmm…

5 Likes
20

Yes. I agree with your conclusions. I can also generally deal with crime strait off the bat, and automation has not come to bite me, which it definitely should have in the long games I have played.

1 Like