Game Design, Programming and running a one-man games business…

Updated Democracy 4 to build 1.28. Lots of super cool balance tweaks.

Soooo… I just uploaded the latest build of Democracy 4 everywhere, after 2 days of non stop testing. Lucky I did that because I found a few little visual anomalies that I could fix straight away, like some missing text in one dialog, and some occasionally ‘leaky’ underlined buttons in other places. Anyway…apart from some (minor) bug fixes, this was mostly a big run through all the stats in the game balancing things. First here is the complete list of changes:

1) Removed duplicate impact on security from mandatory microchip implants (2 appearances on security screen)
2) Reduced Technology impact of driverless car laws by 50%.
3) Doubled the cost of Stem Cell policy.
4) Reduced Technology impact of Technology Colleges by 25%
5) Boosted Technology impact of Rare Earth Refinement by 100%.
6) Reduced Technology impact of Technology Grants by 25%.
7) Illegal immigration situation is now slightly harder to trigger, easier to solve.
8) The Private Space Program,Armed Religious Communities and Doctors Strike situations are now harder to trigger.
9) Egalitarian society situation no longer influenced by Education, but needs Gender Equality and low Generational Wealth Gap.
10) Reduced link between Rare Earth Refinement and Productivity.
11) Reduced impact of Alcohol Consumption on Productivity.
12) Reduced default level of Productivity.
13) Slowed most inputs to the Corporate Exodus situation by 50%.
14) Slightly easier now to trigger the Teachers Strike.
15) Plastics Tax and Packaging Tax are now also inputs to the Black Market situation.
16) Homelessness,Antibiotics-ResitantBacteria,Class War and Food Banks situations now trigger more easily.
17) Dilemmas rebalanced so some are more frequent, others less frequent.
18) Bug fix for theoretically fixed situations like EU monetary policy stopping or starting.
19) Media Monopoly situation now reduces Democracy.
20) Fixed bug where some values on the ‘disposable income’ screen were inaccurate, or garbled (esp for South Korea).
21) Car Emissions limit’s impact on motorists income is now scaled by inverse of EV transition.
22) Cost of body cameras now scaled to numbers of police,community police & armed police.
23) Fake news now increases chance/effect of Contagious disease situation and boosts racial tension.
24) Fixed links for minister icons on main screen to correctly show inputs and outputs from them and their sympathy groups.
25) Education now takes time (2 years) to feed through into a change in Productivity.
26) Air travel now slower to respond to GDP, but boosts pollution more.
27) Worker productivity impact on GDP now a gentler curve.
28) GDP impact on unemployment now a gentler curve.
29) GDP has a steeper impact on pollution, homelessness and car usage now.
30) Corruption and average minister effectiveness now shown at the bottom of finance screen. Corruption affects policy costs/income now.
31) Corruption and minister effectiveness now shown at bottom of finance history for each policy, mouseover shows cost impact of each.
32) Corruption slightly boosts tax evasion now.
33) Rebalanced many inputs from state services to state employee membership and unemployment to make them more accurate.
34) Voter turnout is now generally higher in all situations.

Quite a list this time… but that’s because a lot of it was small statistical nudges here and there. Actually this is one of the most time consuming things about developing a game like this. I don’t have any fancy new particle effects to show you, or any new horse-armor options or new fireball spells or laser upgrades. All I can offer you is “After careful consideration and spreadsheet analysis, Corruption is X% more likely when Press Freedom is below Y”. This is actually stupidly important because its getting all those numbers right that makes the game playable and fun and not a mess.

The only big noticeable changes that players will see rather than *experience* are that bureaucracy is now a thing (triggered by having a LOT of policies, or not so many if you are the weirdly bureaucracy-obsessed Italians), and I added some extra data at the bottom of the finance charts to show you that corruption also affects policy effectiveness along with ministerial effectiveness.

In theory Democracy 4 gets easier to balance over time, because we are slowly homing in on the ‘perfect’ game balance, by pruning obvious dumb edge cases as they are reported or spotted by me, and also I get more and more data from players on what happens too often, what happens too rarely. For example each update should have you experiencing a larger and larger percentage of the dilemmas and events in the game. They may *feel* random, but they are actually being triggered by various in-game parameters. Balancing those so that they show up at appropriate times, but you also get to experience most of them is really tricky…

As always, don’t forget to voter on priorities from the games main menu screen, and like I always say, leaving a positive steam review is a really big help, as is telling your friends about the game if you are enjoying it. Its also really great to get helpful feedback. The best possible feedback is stuff like “When playing country Y as a conservative, it feels too easy to keep voter group X happy by doing Z, even though it doesn’t upset anybody else” or whatever…


3 thoughts on Updated Democracy 4 to build 1.28. Lots of super cool balance tweaks.

  1. This sounds a bit like balancing chess. Are there too many options? Maybe some scripts with sample games that can be replayed with new effects might give you faster results?

  2. Actually coding scripts to play the game is almost impossible because of the sheer number of choices. It seems like any game can be play-simulated by an AI but… omg the number of actions a player can take each turn quickly becomes astronomical after a few turns.
    Also… the game is pretty popular, so we can collect a lot of anonymous data from real play sessions and analyze what real humans choose to do, rather than exploring all possible random permutations of play with some AI.

  3. Yeah makes sense. With that many permutations it’s an atoms in the universe numbers problem. I imagine the anonymous data helps a lot. I was thinking more of an analytical blue/green check. With 100 games which ones fail pre-term limit under various conditions? What’s the vote ratio? If you change one choice (less health care?) or add an extra option (breathing privilege tax) or change to ‘hard mode’ what does that do for the end result? Similar to an auto clicker but it’s game plays you can mess with.
    If you take a certain game where you get 90% and then add more challenges do you lose? Do you go down to 60%?

Comments are currently closed.