Still balancing... a blog post

A behind-the-scenes look at how the game gets balanced for maximum playability and challenge:


Is it really so that winning should be harder? The LDP has dominated Post-Independence Japan politics. The INC dominated post-independence politics for 54 years in India. And for the UK, the Conservatives have dominated UK politics for over a century, cumulatively.

I understand that it’s a game, so there has to be a challenge, but perhaps election difficulty could be specific to some countries (US) but downplayed for others. Even so in the US, the Republican Party has dominated since its formation (most Presidents were rep) and in Germany CDU + CSU has dominated in post-war Germany. Incumbency and stabiliry are huge positive factors for reelection, hard for any opposition party to overcome. Maybe people are complaining because your simulation is too realistic. And that they want it to be more gamey. Which I have no problem with. It’s just funny that’s all.

It’s like how Apple had to make music selection pseudorandom to make it seem more random. So you have to make your game pseudoreal for it to seem more real.


But maybe the end game can be made more difficult without making the elections difficult, that just feels like penalization for playing well and solving problems to me, and a disincentive for being a good and attentive leader, what do you say?

(Or perhaps, asides making elections more difficult)

1 Like

To solve this a bit I would sugest something like masked problems, those kind of problem that are minor, but once you solve the main problem may get harder. For example, in a society where crime is high having lots of police and strong penalties may seem correct, but once you solve crime keeping that police level may leave to overpolicing or police brutality, wich may hurt liberal opinion. A bit like how once you correct the problems on the economy is the eviromment what starts creating problems.


I can tell you right now why the win rate steadily climbs. The voters in this game don’t have any memory or sense of context. They react to a current snap shot and that is it.

If you want a clear example of this, play Japan and try to please liberals. You can make many liberal decisions, and liberals will still be mad at you. This is because Japan starts with a lot more conservative policies than other countries, and the liberal voter block don’t see how you’ve changed the coutry compared to what it was like when you took control, they just see a current snap shot.

With 2 terms worth of turns behind me I can make a much better snap shot for the voters to look at than is possible with only 1 term’s worth of turns and therefore political capital.

1 Like

Citystate II approval works like that:
The right and left have only temporary happiness from policy - when you change it it shifts, then goes back to neutral level.
Simulation and situation equivalents are permanent.
Same with taxes, so maybe taxes would be exempt from this, or be simulated as tax load simulation.

So to represent this in Democracy 4 all happiness values from policy would be temporary - first turn is full effect then decline to 10% of full value after some turns.
For example banning abortions is -100% to Liberal happiness, but after 10 turns it would be only -10% to Liberal happiness.
There are way too much of happiness increasing policies.
Maybe cynicism should start earlier and be more dynamic?

You already have inertia, so implementing happiness decay wouldn’t be too hard.

1 Like

I removed the death penalty in Japan, but I also made crime zero, among other things. I got reelected, just a brief example, but making liberals happy in Japan (atleast gradually), doesn’t lose you the next election.

Why should lose happiness with a policy they like?

Because its becoming status quo - a background.
There are still feelings but just tuned out.

1 Like

Glad to see another blogpost! It was very interesting to read and had several points I agree. I’ll just add some thoughts here.

  • It seems Free market & Liberal democracy is still a norm. One of the major factors here might be initial conditions of the given countries but it would be nice if the game can add challenges to divisive governments.
  • I think the game needs more moving parts. Most simulations don’t have any random-walk behaviors and remain static most of the time unless players are trying to stir them. Some social issues can be a result of long-term stress accumulation on the social structure. For example, Austrian business cycle theory says that malinvestment & overconsumption happen in the time of the boom, eventually leading to the bust. While this doesn’t mean the GDP sim should be cyclic (and get out of player’s control) or the game should embrace Austrian perspective, this can be a good way to introduce long-term (or 3+ term :P) trouble.
  • Election-time stat summaries by nb of terms are interesting to read. But I want to address the possibility of Survivorship Bias. Stats at 2 & 3+ terms were collected from those who made it through the last re-election. Only 54% of the 1st term stat samples were able to pass the first hurdle and they might not be the same with the other 46% in terms of understanding of the game mechanics. In other words, players who can’t improve the stats in the 1st term might have not given a chance to send their 2/3+ term stats. The same might have happened between the 2nd term samples and 3+ term samples. So I’d like to take these numbers with a grain salt.
  • Regarding the winning stat, I’m generally for the idea of dynamic difficulty but adding election wins as a factor might bring 2 issues. First, sounds redundant with the year sim. I find it hard to say consecutive wins are different from the number of years at office if not the second issue. That is, it will encourage players to extend their executive term length to minimize their election win count.
  • I’m still for stronger complacency effects and it would be nice to have a separate option or mod enabling such.

Thoughts on comments

  • While the game has some unrealistic parts (like demanding the same efforts to reduce the crime stat from 50% to 45% compared to 5-to-0, the income sim allowing half middle-class and half wealthy, or the country remaining mostly static for like 50 years), I honestly don’t find voter behaviors that unrealistic. Keeping the leader in the office for 50 years is acceptable if he made so much progress like any experienced players can do in the game. Plus to that, there are only 3 possible endings - end of career, end of life, or getting bored of clicking the next turn. I can’t help but agree that harsher endgame could feel punishing. I do think that the game should remain similarly difficult after 7 consecutive election wins. However, since nb of policies aren’t multiplying themselves, starvation of untapped policy choices is inevitable. Players will find optimal points and stay there unless the surroundings force them to move.
  • I’m for the idea of ‘masked problems’ brought by javive11. I’ve created a very experimental mod on my own which tweaks approval effects of policies with memories of situations. Situations, even solved ones, leaving a mark on political landscapes can be a good way to introduce 2+ term troubles or present different gameplay experience every time they play.
  • It might be nice to separate temporary & permanent effects on policies. While I’d like to see such distinction in economic stimulus policies too, I’m always for more complacency-like effects in approval. There are several policies which are likely to be considered a norm several years after the introduction. Additionally, in a society where some voter groups have an absolute majority, no party would even dare to cross them and they won’t feel grateful for the government (this can be a condition for the enhanced complacency to kick in).

I don’t know about this. It feels like a disincentive to please your voters, what would incentivize pleasing your voters then?

Simulations and situations would be ones with constant strength of happiness influence

  1. Yeah, I think it’s pretty clear, Cliff will (probably) not shift, I don’t expect Soviet Democracy (for ex) to replace Liberal/Conservative Democracy in game anytime soon. The DLCs might bring about some change, who knows?

  2. Yeah, I think that the events and situations try to induce some pseudorandom walk behaviour in game, we could probably use more such behaviour (it be fleshed out). Are there any figures to back up Austrian Business Cycle Theory?

  3. Yeah, Survivorship bias is a real issue, which needs to be addressed.

  4. Indeed.

  5. Nothing of substance to add.

  6. True.

  7. Yes, I’d like to see this.

  8. You never know when a radical party may sweep into power. Unfettered abortion is considered a norm in US society (atleast for Liberals/Libertarians), but what if there’s a conservative revolution, or a soft one, like with how the US Supreme Court has a conservative supermajority (for the next couple of decades? I definitely think courts and opposition need to be fleshed out more, opposition might be handled in upcoming DLC, but what about the courts?)?


With regards to balancing, I did suggest a new situation which might be a good balance in longer games:

1 Like

It seems that this person is having the exact opposite problem. They’re having people hate them regardless of what they do, atleast where the group they wish to please is concerned.