Another Economy Deep Dive (aka Machine Empires aren't horrible)

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

Gamma Knife

Recruit
16 Badges
Jan 7, 2019
5
0
  • Crusader Kings II
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris - Path to Destruction bundle
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Age of Wonders III
  • Surviving Mars: First Colony Edition
  • Age of Wonders: Planetfall
Long time forum scroller, first time poster here. I've been attempting my own deep dive into the economic models of the various empire types to try and determine the "true" cost of various resources for normal empires as compared to hive mind and machine empires. I had heard that machine empires were considered borderline worthless in 2.2 because their main advantage (extra alloys per worker) had been given to the other empire types as well, and because energy was a more expensive pop maintenance resource than food. However, in my most recent playthrough I found that the AI machine empires were not only faring well, but tended to dominate against their biological neighbors. After building a lengthy spreadsheet, I've determined that machine empires are still on par with the other empire types, and are actually superior to non-gestalt biological empires under certain conditions.

In building an economic model to compare the various empire types, I first needed to find common denominators by which to compare. Sure, everything has an exchange ratio on the galactic market, but those are always in flux and ideally the market should eventually shift such that the market reflects the true values of resources, rather than driving them. Furthermore, how do you measure non-tradeable finite resources like building slots, district slots, pops etc.?

I eventually determined that rather than measuring economic value in goods like food, energy, etc., the TRUE value of produced resources are based around common fixed resources in the game: Pops, District Slots, Building Slots, and to a lesser extent Housing. Housing can be created via districts or buildings, so there's really only 3 resources (pop,building slots, district slots) with housing achievable via two exchange rates. I then modeled all of the basic resources to determine their true value in these resources. E.g., to NET one unit of extra food from a farm district in a regular empire, you need 0.2 pops, 0.1 energy, 0.05 consumer goods, and 0.1 farm districts. Notice that this is the net value, the housing costs and farmer food maintenance costs are baked in to the housing provided by the farm district and offset the net food output. Similarly, calculate out the costs for energy, minerals, and consumer goods.

The second consideration is the recursive costs. Take the above example: each unit of extra food costs 0.05 consumer goods. Those goods are produced by artisans, which themselves require extra food grown by farmers who require consumer goods made by artisans, and so on and so forth. I did a first or second order approximation of these recursive costs as appropriate. First or second order approximations seem to be acceptable as the additional costs imposed by each "loop" are frequently significant on the first pass, but fall off pretty dramatically after the first iteration.

Going back to the food example, you could then break down all of the costs to net one unit of extra food as something like: 0.24 pop, 0.01 housing, 0.005 building slots, and 0.12 district slots. For each empire type, I came up with base values as follows (note, these are rough approximations with several assumptions, your mileage may vary):

Regular Pops Housing Bldg District
Food: 0.24 0.009 0.005 0.12
Energy: 0.36 0.013 0.0065 0.19
Minerals: 0.35 0.011 0.0057 0.16
Consumer goods: 0.73 0.2 0.1 0.26
Alloy: 1.46 0.4 0.2 0.86
Research: 1.89 0.77 0.38 1.08

Hive Mind
Food: 0.22 0.073 0 0.07
Energy: 0.33 0.11 0 0.11
Minerals: 0.32 0.11 0 0.09
Alloy: 1.27 0.65 0.17 0.62
Research: 1.58 0.86 0.25 0.83

Machine
Energy: 0.4 0 0 0.2
Minerals: 0.27 0 0 0.18
Alloy: 1.27 0.33 0.17 0.57
Research: 1.63 0.5 0.25 0.67

The first takeaway is that both Hive and Machine empires appear to be more efficient in utilizing pops, building, and district slots into resources. By contrast, Machine empires appear to be the most housing efficient and Hive Minds are the least housing efficient. This is due to the fact that Hive mind non-city districts provide 3 basic jobs but only 2 of the 3 required housing to fill those jobs. This adds an additional housing requirement to basic resource production that simply does not exist with the 2-job, 2-housing districts of regular and machine empires. Offsetting this, of course, is the ability to convert extra building or district slots into housing via city districts or luxury residences. Therein, hive minds shine with extra housing provided per slot. Reducing the above charts to a 3-resource system (converting housing costs away with building/district costs), Hive Minds are more efficient across the board when utilizing city districts than their regular counterparts, while lagging behind in district/building efficiency somewhat when warrens are required for housing.

The second thing I noticed from this chart was.... what on Earth, Machine Empires are actually still very good?! Even though they are the absolute WORST at generating excess energy, requiring the most pop/district utilization to create one unit of surplus, they run away with mineral production, match hive minds for their alloy production, and still manage to beat regular empires at research. All the while, they remain incredibly housing and building efficient. At first I thought there was a typo somewhere in my spreadsheet. How can they be so efficient when their energy costs are higher than any other energy cost, they use energy in place of food, and their energy cost is worse than any equivalent food cost? It turns out that the culprit is in the recursive costs, as discussed earlier. Take an energy district for a regular empire. To run that district you need 2 pops, food for those 2 pops, consumer goods for those 2 pops, food and consumer goods for the pops making the food and consumer goods, extra housing for the people making the consumer goods, energy maintenance for the farm district and consumer goods buildings, it's turtles all the way down. Hive minds, you don't have the consumer goods cost, but you do have the food recursive costs AND have find the extra housing in there somewhere. But machine empires? Nada, nothing, zilch. 1 energy district requires 2 pops, no extra housing, consumes 1 energy in maintenance and 2 energy in drone food (both consumed from what is produced at THIS energy district), and nets you 5 energy. You only get 5 energy from these 2 pops, but there are NO recursive costs. You have a very simple, clean equation that 2 pops + 1 district = 5 energy. The math is similar for mining districts: 2 pops, 3 energy in maintenance, and that's it. Between the lack of food, the lack of consumer goods, and the complete removal of any recursive costs, the machine supply chain becomes incredibly efficient all the way to alloys and research points. Sure, you're off to a slow start with energy, but it's not as slow as it looks on the first pass, once you factor in the recursive costs. To counter balance, there's no compound interest holding you down and your path from start to finish is much shorter.

This analysis has one huge caveat: I'm comparing empires assuming that energy and consumer goods are coming solely from generator districts and industrial buildings: no trade value. This means that regular empires appear to be at a huge disadvantage, but that's because their benefits from trade value are not factored in. Trade value is a much harder resource to model in this instance for several reasons. First, trade can turn into energy, energy and consumer goods, or energy and unity (a non-economic resource). Second, assuming that you go energy and consumer goods for the shortened consumer good supply train, that's one resource input turning into multiple resource outputs, compared with the other way around which is much easier to work with. Finally, as end products, the ratio of energy to consumer goods varies wildly depending on whether your spending those resources to produce minerals, alloys, or especially research points. Solving a single equation for raw resource cost then instead turns into a system of equations depending on your relative output of minerals, alloys, and research points.

Nevertheless, I gave it a shot. I took a look at two possible scenarios to help reduce the complexity of the equation. First scenario: what if I'm swimming in trade value? I get a ton free from city districts, trade agreements, starbase modules, and stack huge multipliers. If energy and consumer goods were basically free (no generator districts or civilian industries required) and my limiting factors were instead food/mineral production, district and building slots, housing etc., what would my economy look like? Turns out, as you'd expect, a trading economy runs away from the rest.

"Free" Trade Pops Housing Bldg District
Food: 0.2 0 0 0.1
Energy: 0* 0* 0* 0*
Minerals: 0.3 0 0 0.13
Consumer goods: 0* 0* 0* 0*
Alloy: 1 0.33 0.16 0.62
Research: 0.6 0.5 0.25 0.55

Objectively better across the board, especially in research. Obviously, this is not a realistic scenario. You never have infinite trade value. Rather, this models what resources cost when they are consumed using energy and consumer goods generated up to your "trade I get for free" limit, whatever that may be. More realistically, you have to look at the real cost of generating trade value, typically in clerk jobs generated from city districts and commercial zones.

Turns out, trade value produced from city districts doesn't produce net value. 2 clerks producing 4 trade value (I'm assuming you have the prosperity tradition) provide 2 energy and 0.5 consumer goods.... which do nothing for you but offset the maintenance cost and consumer goods cost of the city district and clerks, respectively. It doesn't even factor in the recursive costs to feed the clerks. In short, city clerks do not actually generate value for you, they merely offset the costs of converting a district slot into extra housing (and even then they don't cover ALL the recursive costs).

So how about commercial zone clerks? They're even worse, if such a thing is possible. 1 building slot and 5 pops produce 10 trade value at the cost of 2 maintenance. That creates 5 energy and 2.5 consumer goods... but 2 energy and 1.25 consumer goods are immediately eaten in upkeep. Then you have to factor in the cost of 5 food recursively... and before you know it you've spent 1 building slot and 5 pops for almost nothing. Any energy or consumer goods you actually net out of the deal ends up being insanely more expensive than if you'd simply built a few generators and a civilian industries. Granted, trade value can get some pretty crazy multipliers so it's not quite so bad, but it's still miserable and a calculation with multipliers included is outside the scope of this post.

Parting thoughts:
1. I think the paradigm of measuring resource production as a function of pops, district slots, building slots, and to a lesser extent housing, is the way to go. District and building slots represent the available planet resources but require an investment in minerals to properly develop/exploit. Pops represent a time-gated method of exploiting the resources generated by building and district slots, and vary wildly in cost based on growth rate, resettlement policy, and game year. Housing provides an upper limit on the pop resource, and is convertable from district and building slots via two exchange rates (cities vs housing buildings) that can vary between empire types.
2. Based on this model, there is an opportunity to directly compare empire type production capacity and efficiency, as well as the ability to differentiate different empire playstyles. Right now, Hive Minds are king when it comes to pop and district efficiency and growth, but lag behind when housing becomes a limiting factor. For flavor, I'd be interested in seeing perhaps Hive Minds become even more zerg-like. Rapid growth, high maximum population, but lower individual pop efficiency. This might entail a slight nerf in per-pop efficiency, a matching buff to housing efficiency (merely to offset the per-pop nerf), while retaining the rapid-growth expand-or-die playstyle. They should be good early, but worsen as the game progresses if other empires can catch up.
3. Machine empires, amazingly, still fit their intended niche. They suck at producing energy for its own sake, but excel in producing the finished resources you actually care about. I think they should play the opposite of a hive mind. Slow as molasses growth, but incredibly efficient with what they have. I also believe there is room stylistically for them to excel at using the building resource efficiently: They should somehow always get more bang for their buck using the limited building slot resource. This can be by having their buildings produce more good per worker (as they used to be with alloys), or simply by being more efficient with a smaller population, therefore not hitting that 75pop building cap nearly as quickly or as often. They should be the worst early game, but snowball if they have the chance to catch up.
4. Regular empires... trade value needs some love. If you are swimming in trade, you are unstoppable. But actually GETTING that trade value is a major trap. And once you have to actually work for trade or collect energy/consumer goods manually, regular empires lag far behind. City clerks don't net positive resources, commercial clerks barely recover more than their base cost. And that doesn't even begin to consider the starbase trade hub/hangar bay/patrol costs. I think the niche for regular empires is that they are the "average" all around in total pops, pop growth rate, district efficiency, and building efficiency. However, their unique playstyle comes in trade value as a supplementary/alternate energy/consumer goods producer. When they have the trade to spare, they are above average, when they lack the trade, they fall behind. I think the playstyle is actually there, but in need of a numbers tweak. We need more "free" trade from non-commercial zone sources, more value from the existing trade sources to make them net positive, less micro and overhead in producing/collecting/transporting/protecting trade to the capital, or some combination of all three.

tl;dr
1. Machine empires are actually very good. Sure they suck at producing energy for its own sake, but in terms of producing what you ACTUALLY WANT, shorter supply chains and lack of recursive costs cause them to come out ahead.
2. Hive minds are in a really strong place right now. They grow faster than anyone, allowing them to more quickly consume the "true" resources (pops,district slots,housing slots). Eventually, they get a housing problem once their planets start to fill up. If they have spare city districts to use, they remain ahead, but become starved for building slots when they need to use building slots for housing. I think this fits very thematically with the Hive mind playstyle. They outpace everyone when they have room to expand, but stagnate easily when they've filled up the available space.
3. Regular empires are objectively the best when they are swimming in extra trade value, but objectively the worst when producing resources "manually" or by expending resources to generate trade. When generating trade, city clerks really don't actually help, but only (not quite) offset the cost of creating the city district to produce housing. Commercial zone clerks are almost a complete waste of time. Pops are better spent resettling almost anywhere else.


Edit:
I'm including modified numbers in this post for gestalt consciousness empires, baking in housing costs and factoring in the increased amenities cost as addressed by Promethian. Hive mind empires come across as noticeably worse on per-pop and per-building efficiency, but still excel on per-district efficiency. Now look to be worse at mineral and energy production across the board, but the shortened supply chain coupled with the adjustments to Hive Mind numbers show they are the clear leaders in Alloy and Research production.

Per sillyrobot, I've also adjusted the trading empires numbers a little to reflect the slight Consumer goods production created by city districts. Apologies, I can't seem to format the tables properly.

Trade Empires:
Pop Bldg District
Food 0.24, 0.005, 0.12
Energy 0.36, 0.006, 0.19
Mineral 0.35, 0.006, 0.16
Alloy 1.68, 0.2, 0.95
Research 2.31, 0.38, 1.25


Hive Mind
Pop Bldg District
Food 0.29, 0.018, 0.11
Energy 0.43, 0.027, 0.15
Mineral 0.42, 0.027, 0.13
Alloy 1.68, 0.27, 0.87
Research 2.1, 0.38, 1.16


Machine
Pop Bldg District
Energy 0.55, 0.04, 0.21
Mineral 0.37, 0.03, 0.19
Alloy 1.44, 0.2, 0.65
Research 1.89, 0.31, 0.79
 
Last edited:

Promethian

Field Marshal
44 Badges
Feb 2, 2016
2.592
787
  • Stellaris
  • Age of Wonders III
  • Tyranny - Tales from the Tiers
  • Stellaris: Synthetic Dawn
  • Surviving Mars: First Colony Edition
  • Cities: Skylines - Parklife Pre-Order
  • Cities: Skylines - Parklife
  • Stellaris: Distant Stars
  • Shadowrun Returns
  • Surviving Mars: First Colony Edition
  • Cities: Skylines Industries
  • Stellaris: Megacorp
  • Prison Architect
  • Surviving Mars: Digital Deluxe Edition
  • Cities: Skylines - Campus
  • Age of Wonders: Planetfall
  • Age of Wonders: Planetfall Deluxe edition
  • Age of Wonders: Planetfall Season pass
  • Stellaris: Lithoids
  • Stellaris: Federations
  • Empire of Sin
  • Stellaris: Necroids
  • Stellaris: Nemesis
  • Stellaris: Leviathans Story Pack
  • Warlock: Master of the Arcane
  • Cities: Skylines
  • Cities: Skylines - After Dark
  • Cities: Skylines - Snowfall
  • Stellaris: Galaxy Edition
  • Stellaris: Galaxy Edition
  • Tyranny: Archon Edition
  • Stellaris: Digital Anniversary Edition
  • Cities: Skylines - Natural Disasters
  • Stellaris - Path to Destruction bundle
  • Cities: Skylines - Mass Transit
  • BATTLETECH
  • Surviving Mars
  • Cities: Skylines - Green Cities
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Crusader Kings II
  • Stellaris: Ancient Relics
  • Age of Wonders: Planetfall - Revelations
  • Stellaris: Distant Stars Pre-Order
Your analysis doesn't take amenities into account. Especially with the nerf on the test patch of each drone producing 4 instead of 5, the amenities cost of maintaining society is very high for gestalts. Regular empires get amenities very efficiently from their ruler pops (usually) and get lots attached to their unity sources.
 

Gamma Knife

Recruit
16 Badges
Jan 7, 2019
5
0
  • Crusader Kings II
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris - Path to Destruction bundle
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Age of Wonders III
  • Surviving Mars: First Colony Edition
  • Age of Wonders: Planetfall
@Promethian
Thanks, I'll factor that further in. I had assumed (somewhat incorrectly) that amenities were a non-issue for gestalts as they seem to be for regular empires, since clerks and gene clinics are everywhere for regulars. Some of the amenities cost is already baked in my spreadsheet: for example, maintenance drones already attached to housing buildings already factor in their pop/upkeep to work that job. This will be important when determining the number of extra building slots allocated to amenities-only buildings for gestalts. I'll take a look, and perhaps copy over more of my raw data.
 

sillyrobot

General
Jul 18, 2015
1.859
3.584
"2 clerks producing 4 trade value (I'm assuming you have the prosperity tradition) provide 2 energy and 0.5 consumer goods.... "

Shouldn't that be 2 energy and 1 consumer good? Is this a failure in the text or does it affect the model?
 

Gamma Knife

Recruit
16 Badges
Jan 7, 2019
5
0
  • Crusader Kings II
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris - Path to Destruction bundle
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Age of Wonders III
  • Surviving Mars: First Colony Edition
  • Age of Wonders: Planetfall
"2 clerks producing 4 trade value (I'm assuming you have the prosperity tradition) provide 2 energy and 0.5 consumer goods.... "

Shouldn't that be 2 energy and 1 consumer good? Is this a failure in the text or does it affect the model?

That is a failure of the text, and slightly affects the model. You are correct in that it should be 2E, 1CG. What this means is that now city districts are net generators of 0.4CG (1-0.5 from the clerks - 2*0.05 from the clerks' farmers). This isn't much individually but will snowball with a large number of city districts. The affect this will have is increasing the "free trade" floor for trading empires, meaning that they stay optimal through to a larger economy before needing to create generators/civilian industries and falling behind. This doesn't factor in whatever the starbase costs will be, and commercial industries are still pretty bad, but this helps.

I'll collect a few replies for now and edit my original post in a while.

Edit: Quick napkin math. Assuming you gain your required housing all through cities, I subtracted out the free CG costs you get when creating finished goods alloy and research. Alloy cost, you knock your CG cost from 0.17CG per alloy to 0.12CG per alloy. Research drops from 1.25CG to 1.18CG. Total savings is about 2-3% for the various "raw" resources across the board, not as significant as one might hope.
 
Last edited:

Bouchart

Field Marshal
85 Badges
Nov 29, 2014
2.643
1.193
  • Crusader Kings II
  • Europa Universalis IV: Common Sense
  • Stellaris: Apocalypse
  • Teleglitch: Die More Edition
  • Victoria 2
  • Warlock: Master of the Arcane
  • Europa Universalis IV: Dharma
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis IV: Cradle of Civilization
  • Cities: Skylines
  • Age of Wonders: Planetfall Deluxe edition
  • Europa Universalis IV: El Dorado
  • Mount & Blade: With Fire and Sword
  • Pillars of Eternity
  • Majesty 2 Collection
  • Crusader Kings II: Horse Lords
  • Europa Universalis IV: Cossacks
  • Europa Universalis IV: Mare Nostrum
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris: Leviathans Story Pack
  • Age of Wonders III
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Imperator: Rome Deluxe Edition
  • Shadowrun: Dragonfall
  • Stellaris: Distant Stars
  • Stellaris: Ancient Relics
  • Imperator: Rome
  • Surviving Mars: Digital Deluxe Edition
  • Europa Universalis IV
  • Shadowrun: Hong Kong
  • Shadowrun Returns
  • Europa Universalis IV: Rule Britannia
  • Europa Universalis IV: Golden Century
  • Age of Wonders: Planetfall
  • Prison Architect
  • Stellaris: Megacorp
  • Cities: Skylines - Parklife
  • Stellaris: Federations
  • Europa Universalis IV: Art of War
  • Divine Wind
  • Europa Universalis III: Chronicles
  • Leviathan: Warships
  • Magicka
  • Majesty 2
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: Charlemagne
AI can't play machine consciousness at all. I've seen Determined Exterminators with a barebones fleet a hundred years in with mass unrest everywhere, without ever entering a war at any time.
 

supreme

Sergeant
25 Badges
Oct 6, 2010
95
182
  • For the Motherland
  • Stellaris - Path to Destruction bundle
  • Stellaris: Leviathans Story Pack
  • Pillars of Eternity
  • Semper Fi
  • Hearts of Iron III: Their Finest Hour
  • Hearts of Iron III
  • Rome Gold
  • Hearts of Iron IV: Death or Dishonor
  • Stellaris: Nemesis
  • Stellaris: Federations
  • Stellaris: Ancient Relics
  • Imperator: Rome
  • Stellaris: Megacorp
  • Stellaris: Distant Stars
  • Stellaris: Apocalypse
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Synthetic Dawn
  • Hearts of Iron IV: Together for Victory
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV Sign-up
  • Stellaris
  • Mount & Blade: Warband
  • Rome: Vae Victis
  • Crusader Kings II
Machine Empires are only a challenge early game, they need to be played very wide and grab as many planets as possible and resettle for 5 pops on each to get the assembly plants running. Once you do that you become unstoppable mid game. But ye as an above poster pointed out the AI machine empires are not functioning properly.
 

RyuujinZERO

Corporal
15 Badges
Jun 8, 2016
38
4
  • Stellaris
  • Stellaris: Galaxy Edition
  • Stellaris: Galaxy Edition
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Stellaris - Path to Destruction bundle
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Stellaris: Distant Stars
  • Stellaris: Megacorp
  • Stellaris: Ancient Relics
  • Stellaris: Lithoids
  • Stellaris: Federations
  • Crusader Kings II
  • Stellaris: Synthetic Dawn
Interesting read, I'm definitly inspired by this to re-examine the idea of playing a robot empire next playthrough.

This is a big ask, but how does Agrarian Idyll fit into this picture for non-gestalts?

Agrarian Idyll completely reshapes the districting for organics by giving +1 housing (+2 for farms), giving farmers +2 amenity production, which means it's quite reasonable to to house an entire working population including buildings without any need for city districts (Especially if you have housing need reduction traits stacked). Seems like it'd significantly re-write the organic balance.
 

Secret Master

Covert Mastermind
Moderator
95 Badges
Jul 9, 2001
36.655
20.097
www.youtube.com
  • 200k Club
  • Europa Universalis III Complete
  • Victoria: Revolutions
  • Europa Universalis: Rome
  • Semper Fi
  • Sengoku
  • Ship Simulator Extremes
  • Sword of the Stars II
  • Victoria 2
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • Rome: Vae Victis
  • Warlock: Master of the Arcane
  • March of the Eagles
  • 500k Club
  • Cities: Skylines
  • Crusader Kings II: Holy Knight (pre-order)
  • Pride of Nations
  • Mount & Blade: Warband
  • Mount & Blade: With Fire and Sword
  • Crusader Kings II: Way of Life
  • Pillars of Eternity
  • Crusader Kings II: Horse Lords
  • Crusader Kings II: Limited Collectors Edition
  • Crusader Kings II: Conclave
  • Hearts of Iron IV: No Step Back
  • Europa Universalis III: Chronicles
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: Sword of Islam
  • Commander: Conquest of the Americas
  • Deus Vult
  • Europa Universalis III
  • A Game of Dwarves
  • Divine Wind
  • Europa Universalis IV
  • Europa Universalis IV: Conquest of Paradise
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III Collection
  • Heir to the Throne
  • Europa Universalis III Complete
  • King Arthur II
  • The Kings Crusade
've seen Determined Exterminators with a barebones fleet a hundred years in with mass unrest everywhere, without ever entering a war at any time.

2qlu17.jpg
 

Gamma Knife

Recruit
16 Badges
Jan 7, 2019
5
0
  • Crusader Kings II
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris - Path to Destruction bundle
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Age of Wonders III
  • Surviving Mars: First Colony Edition
  • Age of Wonders: Planetfall
Interesting read, I'm definitly inspired by this to re-examine the idea of playing a robot empire next playthrough.

This is a big ask, but how does Agrarian Idyll fit into this picture for non-gestalts?

Agrarian Idyll completely reshapes the districting for organics by giving +1 housing (+2 for farms), giving farmers +2 amenity production, which means it's quite reasonable to to house an entire working population including buildings without any need for city districts (Especially if you have housing need reduction traits stacked). Seems like it'd significantly re-write the organic balance.

So for organics I'm already disregarding the amenities equation, as you typically come by enough from your capitol building, gene clinics, and the odd clerk pop. Similarly for idyll, I'll just assume that the amenities bonus provided by the bonus offset the amenities you'd otherwise gain off of your city clerks, and then some.

Counterintuitively, I think what this does is make agrarian idyll refined goods powerhouses. Food/minerals/energy are already created essentially housing-free, as the districts provide housing for their workers. Housing costs really factor in when creating consumer goods, alloys, and research. Granted city districts are no longer worth it, but it basically means that you get "free" housing from the low-tech districts you'll already have. This really factors in when producing specialized goods, such as alloys that require 0.4 housing per unit produced.

If you want to see what I mean, take organic empire from my first list (the one with pops, housing, buildings and districts). Now, zero out the entire housing column and compare it with any empire in my second list (the one with only pops, buildings, and districts). So long as you never need to build luxury residences or city districts to meet your housing needs, that is your comparison. OP.

But as already mentioned, you'd have to play pacifist. Which defeats the point of swimming in all of those alloys in the first place.

And yes, definitely worth it to attempt a machine empire. It may feel like its playing poorly as you're always swimming in generator districts yet starving for energy, but if you put a human in place of an AI for now, you'll fare much better than current wisdom suggests. I'll give it a shot, if ever they fix the lag.
 
Last edited:

Less2

Field Marshal
Jan 20, 2016
3.737
5.039
Have you tried running slaves and/or robot servitude in your spreadsheet? Those are the big, insanely good moneymakers that push normal empires way ahead. I'd also love to see your spreadsheet if you are comfortable posting it. I have mine here: https://forum.paradoxplaza.com/forum/index.php?threads/economic-spreadsheet.1140978/. It's a bit more bare-bones and doesn't consider recursive costs or housing costs, but it calculates expected production of alloys or research for 1k pops optimized to the task, which I figure is the best way to evaluate the overall competitiveness of a build since the ideal food/energy/minerals is just whatever keeps you stable while you pump out ships or research.

1. Machine empires are actually very good. Sure they suck at producing energy for its own sake, but in terms of producing what you ACTUALLY WANT, shorter supply chains and lack of recursive costs cause them to come out ahead.

Umm, hate to throw a wrench into this, but shorter production chains are BAD. This is because production modifiers get applied to each portion of the chain. It's not unusual to start the game with 20-30% production bonuses and by the end game you have more like 100%. I'm guessing you went with only base production, not production with modifiers?

Commercial zone clerks are almost a complete waste of time. Pops are better spent resettling almost anywhere else.

Glad another mathematically-minded person agrees with me here :D. I seem to always get a lot of flak when I state this.
 
Last edited:

Gamma Knife

Recruit
16 Badges
Jan 7, 2019
5
0
  • Crusader Kings II
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Europa Universalis IV: Rights of Man
  • Stellaris - Path to Destruction bundle
  • Europa Universalis IV: Mandate of Heaven
  • Surviving Mars
  • Stellaris: Synthetic Dawn
  • Age of Wonders III
  • Surviving Mars: First Colony Edition
  • Age of Wonders: Planetfall
Have you tried running slaves and/or robot servitude in your spreadsheet? Those are the big, insanely good moneymakers that push normal empires way ahead. I'd also love to see your spreadsheet if you are comfortable posting it. I have mine here: https://forum.paradoxplaza.com/forum/index.php?threads/economic-spreadsheet.1140978/



Umm, hate to throw a wrench into this, but shorter production chains are BAD. This is because production modifiers get applied to each portion of the chain. It's not unusual to start the game with 20-30% production bonuses and by the end game you have more like 100%.

Longer production chains are very good when calculating forward. Better to multiply your minerals before you convert them to alloys, etc. They're bad when you have a long list of inputs that feed off of each other recursively. Alloys requiring consumer goods made by pops that require consumer goods plus food made by farmers that require food running on farms that require energy made by workers that require consumer goods and food....the list goes on. The recursive elements on supply chains actually serve to increase the cost of producing to the next step, and you'd need a certain multiplier just to break even on those costs. Longer supply chains are amazing when you have clear cut inputs and outputs, less so when the inputs have cross-dependencies. Machine empires have one of the cleanest inputs with generator districts that only require pops and a tiny amenity input, but are otherwise self-sufficient.

I haven't done slaves or robot servitude, nor have I tried it with different living standards, etc. I'll bet that they really help normal empires, and it's part of the flavor/strategy normal empires need to avoid being the boring "middle of the road" strategy. I think playstyle-wise, regular empires are in a good place right now with their diversity. But since not every empire is a slaver or robot producer, I think it's fair to compare the value of machine empires against the vanilla template for determining their viability. I'll try throwing in some test multipliers tomorrow and see what it spits out.
 

Less2

Field Marshal
Jan 20, 2016
3.737
5.039
Longer production chains are very good when calculating forward. Better to multiply your minerals before you convert them to alloys, etc. They're bad when you have a long list of inputs that feed off of each other recursively. Alloys requiring consumer goods made by pops that require consumer goods plus food made by farmers that require food running on farms that require energy made by workers that require consumer goods and food....the list goes on. The recursive elements on supply chains actually serve to increase the cost of producing to the next step, and you'd need a certain multiplier just to break even on those costs. Longer supply chains are amazing when you have clear cut inputs and outputs, less so when the inputs have cross-dependencies. Machine empires have one of the cleanest inputs with generator districts that only require pops and a tiny amenity input, but are otherwise self-sufficient.

If Machine Empires have a 1x cost for maintenance/upkeep/etc and normal empires have a 1.5x cost, but the former have a 1-chain and the latter have a 2-chain, then MEs with a 100% production bonus effectively have a .5x cost while normal empires have a .375x cost. It works both ways in favor of production bonuses, and against production penalties. But penalized production almost never happens with the new economic system, only in heavily mis-managed planets lacking amenities or newly conquer slaves that lack pops to police them.

I haven't done slaves or robot servitude, nor have I tried it with different living standards, etc. I'll bet that they really help normal empires, and it's part of the flavor/strategy normal empires need to avoid being the boring "middle of the road" strategy. I think playstyle-wise, regular empires are in a good place right now with their diversity. But since not every empire is a slaver or robot producer, I think it's fair to compare the value of machine empires against the vanilla template for determining their viability. I'll try throwing in some test multipliers tomorrow and see what it spits out.

Robots at least are open for every ethic in the game, along with being almost required for Egalitarian*. Thematically they are suitable for every ethic other than Spiritualist. They are also right at the beginning of the tech tree. Slaves are open for authoritarian or xenophobic. So I'd say that most empires are probably suited for one or the other and in terms of making a powerful economy there's no reason not to have one or both.

It'd be different if Hives or MEs had their own really powerful civics or something and the majority of gestalt empires had some way of acquiring -50% housing and -50% amenities pops, but they don't. There's a big gaping hole in potential pop power there.

*btw, if you want calculate the cost for egalitarian's Utopian Abundance and report back with exactly how awful you find it. Cause it's really dang bad.
 
Last edited:

Kaigen

Sergeant
21 Badges
Mar 15, 2018
85
0
  • Stellaris: Federations
  • Stellaris: Lithoids
  • Stellaris: Ancient Relics
  • Stellaris: Megacorp
  • Stellaris: Distant Stars
  • Pillars of Eternity
  • Stellaris
  • Stellaris: Galaxy Edition
  • Stellaris: Galaxy Edition
  • Stellaris: Apocalypse
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Stellaris - Path to Destruction bundle
  • Stellaris: Galaxy Edition
  • Stellaris: Synthetic Dawn
  • Crusader Kings II
  • Magicka
  • Crusader Kings II: Sword of Islam
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: The Republic
  • Crusader Kings II: Legacy of Rome
Counterintuitively, I think what this does is make agrarian idyll refined goods powerhouses. Food/minerals/energy are already created essentially housing-free, as the districts provide housing for their workers. Housing costs really factor in when creating consumer goods, alloys, and research. Granted city districts are no longer worth it, but it basically means that you get "free" housing from the low-tech districts you'll already have. This really factors in when producing specialized goods, such as alloys that require 0.4 housing per unit produced.

Agrarian Idyll looks really good for refined goods production right up until you take into account that they can't have the Arcology Project ascension perk. You're right, AgId means not having to build housing buildings ever and very rarely building city districts; you generate enough incidental housing from your resource districts to allow for a lot of advanced production buildings. But upgrading those buildings requires rare resource upkeep, and arcology districts skip all of that and have their own sufficient housing to boot. So while AgId has to deal with the increasing costs of adding additional specialist jobs in a given amount of space, an Ecumenopolis simply has a flat line. The same applies to gestalt empires in the final analysis: They don't get Arcology Project, and so ramping up alloy production becomes increasingly expensive over time.
 

permeakra

Major
66 Badges
Nov 20, 2017
651
349
  • Cities: Skylines - Green Cities
  • Tyranny: Archon Edition
  • Tyranny: Archon Edition
  • Tyranny: Gold Edition
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Cities: Skylines - Natural Disasters
  • Crusader Kings II: Monks and Mystics
  • Stellaris - Path to Destruction bundle
  • Cities: Skylines - Mass Transit
  • Surviving Mars: First Colony Edition
  • Crusader Kings II: Holy Fury
  • Tyranny - Bastards Wound
  • Crusader Kings II: Reapers Due
  • Crusader Kings II: Jade Dragon
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Surviving Mars: Digital Deluxe Edition
  • Cities: Skylines - Parklife
  • Stellaris: Distant Stars
  • Shadowrun Returns
  • Shadowrun: Dragonfall
  • Stellaris: Megacorp
  • Surviving Mars: First Colony Edition
  • Cities: Skylines Industries
  • Stellaris: Nemesis
  • Cities: Skylines - Campus
  • Crusader Kings III
  • Stellaris: Necroids
  • Stellaris: Lithoids
  • Victoria 2
  • Cities: Skylines
  • Crusader Kings II: Way of Life
  • Crusader Kings II: Horse Lords
  • Cities: Skylines - After Dark
  • Crusader Kings II: Conclave
  • Cities: Skylines - Snowfall
  • Stellaris: Federations
  • Stellaris: Galaxy Edition
  • Stellaris: Galaxy Edition
  • Cities in Motion 2
  • Crusader Kings II
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sword of Islam
  • Majesty 2 Collection
  • Sword of the Stars
  • Sword of the Stars II
Right now, Hive Minds are king when it comes to pop and district efficiency and growth,
I'm running a hivemind right now.

Hiveminds have problem with "crime" and amenities. They might have efficient resource districts and with hive worlds they don't suffer from resource shortage ever, but their hunter-seeker drones are less efficient than enforcers and they don't get advanced crimefighting buildings, having at most -40 per building, while normals may reach -125 per building. They also lack efficient amenity producing buildings, having only maintenance at +25 amenities per building (with traditions), while normals have 4 free amenities from city districts and +40 from a buildings. And no, unlike normals, hiveminds don't want any meaningful level of "crime". It can hit their stability too easily.

All in all, hiveminds are better suited for small planets and habitats, but less suited for huge planets and ringworlds as they simply have problems packing enough sentinel posts and maintenance depots with their increased number of pops-per-resource district.
 

AlanC9

Field Marshal
16 Badges
Mar 15, 2001
5.081
320
Visit site
  • Hearts of Iron III
  • 500k Club
  • Europa Universalis III: Collection
  • Pillars of Eternity
  • Stellaris
  • Cities in Motion 2
  • Crusader Kings II
  • Crusader Kings II: The Old Gods
  • Europa Universalis III
  • Divine Wind
  • For the Motherland
  • Hearts of Iron III: Their Finest Hour
  • Heir to the Throne
  • Majesty 2 Collection
  • Semper Fi
  • Magicka 2
*btw, if you want calculate the cost for egalitarian's Utopian Abundance and report back with exactly how awful you find it. Cause it's really dang bad.

Does it matter? I thought the point of UA is to run it when material wants are a thing of the past and you're just trying to manage a post-scarcity economy.
 

Less2

Field Marshal
Jan 20, 2016
3.737
5.039
Does it matter? I thought the point of UA is to run it when material wants are a thing of the past and you're just trying to manage a post-scarcity economy.

It was an arguably valid option before, it's really not a valid option at all now. It will kill any economy for incredibly minor production bonuses.
 

AlanC9

Field Marshal
16 Badges
Mar 15, 2001
5.081
320
Visit site
  • Hearts of Iron III
  • 500k Club
  • Europa Universalis III: Collection
  • Pillars of Eternity
  • Stellaris
  • Cities in Motion 2
  • Crusader Kings II
  • Crusader Kings II: The Old Gods
  • Europa Universalis III
  • Divine Wind
  • For the Motherland
  • Hearts of Iron III: Their Finest Hour
  • Heir to the Throne
  • Majesty 2 Collection
  • Semper Fi
  • Magicka 2
That's why I said post-scarcity. The point of UA isn't production; you run it when you're past caring about that. Going from +600 CG to +200 CG isn't much of a problem when they're just piling up anyway..
 
Last edited: