I crunched a lot of numbers to investigate the current state of balancing in more detail. In general my feeling after the update was really good. But some small things raised my eye brows. Now I think it’s better to have objective numbers than just a subjective feeling, plus I thought it would be interesting.
I’ll explain the methods and reasoning behind my calculations in a spoiler tag because it’s quite lengthy and you may not care. The TLDR is that I use worker requirement as a universal currency to compare „prices“ so whenever you see numbers, they basically mean „how many workers do I need to produce 1 unit per minute of something“ (including power, terraforming and fulfilling citizen needs).
Methodology
I choose workers as the common currency for everything for two reasons: In my opinion workers are the real limiting factor for everything you do (much more so than building material for example) and also because they work so well as a currency in the sense that everything can be given a price using required workers. Another possible metric that I might also look into later is „traffic produced“.
For each good I looked at how much it costs without any upgrades and also with maximum upgrades. The numbers given in the main text are usually with max upgrades.
Let’s do a very simple example of that: A small sand mine produces 1 sand every 5 seconds. I rebase everything to the minute so it produces 12 sand per minute. It needs two workers. With this basic setup the price for sand is 2/12 = 0.167 workers.
Let‘s look at the upgraded version: a big sand mine with industrial robots. It produces 2 sand every 5 seconds, 24 per minute. It needs 2 workers minus 1 robot for that, so 1 in total. The cost of sand is 1/24 = 0.0417. It’s dirt cheap (pun intended).
This example lacks the cost for the power we need to generate to run the buildings. Which brings me to the next point.
So to calculate the most efficient way to produce something (a.k.a the minimum price), what did I consider? These are the things I factored in:
- Use large mines instead of small ones (also see main text)
- Use atmospheric processor (see main text)
- Use methane pumps rather than fermentation (see main text)
- Industrial robots (but see below)
- Waste Recycling (see main text)
- Soil Enrichment
- Spaceship bonuses
- Use power from Methane power plants because it’s currently the cheapest except for wind (see main text)
- All buildings running at 100% occupancy
And also very important, these are the things I did not consider and why:
- Cost of industrial robots. I assume they are for free but they are actually quite expensive. Introducing them would create a loop in my spreadsheet which it probably can’t handle.
- Proximity bonus. There is no universal optimal setup so I cannot just choose a number for this.
- High-Tech tools. For the most part, they would just cut the cost of everything by 33%. So it’s a universal cost-down of everything that makes no sense to consider when comparing different goods. In reality it’s more complicated than that. Since the bonus now adds rather than multiples, it’s not just a linear 33% price cut with buildings having different amounts of proximity bonuses. Also just like industrial robots, the tools are not for free at all.
I’d post my sheet here if it weren’t in such an embarrassing mess 
If there’s overwhelming demand to see it then I’ll clean it up and attach it.
In general, the improvement in efficiency varies between each resource but most have a factor of 2-3, which is quite nice for balancing, I think. It means that actually rebuilding parts of your city and industry as new tech rolls in is really worth it. This is also thanks to the new spaceship upgrades, which give many products a 50% bonus.
Next I’ll look at the major resource sinks in the game and then at some specific findings regarding balancing.
Construction Material
Concrete and steel are really cheap. Concrete is 0.07 and steel costs 0.1, a little more. Since it’s the slightly more advanced building material that’s totally fine.
Adamantine costs 87.6 which makes it one of the most expensive resources in the game. Kind of makes sense and this is probably how most people feel about it, too. Now that it finally has uses it’s not just a toy anymore and you have to consider how many workers you want to invest into it. You probably want more than 1 per minute, so be ready to spend a few hundred workers just on that.
Overall the building material is very satisfying, if not even too cheap in the latest patch.
Research
The cost of each science pack is as follows:
| Color | Optimized cost |
|---|---|
| Blue | 1.1 |
| Green | 5.3 |
| Yellow | 29.8 |
| Red | 189.5 |
As you can see the cost goes up by about a factor of 5 with each tech level. Also the numbers you can find on the upgrades are getting pretty high later on, but the total amount of tech in yellow and red is not as high as in green. Considering that you don’t have any optimizations at the start, the curve actually starts out flatter. Raw blue costs 1.8, raw green 9.0. That makes the jump to red science really, really tough. Red science packs are the most expensive thing after magnetic field generation. But of that you only need 1 unit. Of red science you may want more than 1 per minute. Given its ingredients it would be cheaper if it was made out of pure gold Personally I’d reduce one ingredient (not sure which) a little bit.
Still the research is working quite well with these numbers, I think.
Habitat Requirements
As you are painfully aware, your pesky inhabitants get increasingly picky and hard to satisfy. Sometimes you may wonder if it is even worth satisfying their needs because it seems to cost more to produce all the garbage they want than the amount of workforce you additionally get shipped to the planet. Here are numbers of per-capita effort to make you rest easy:
| Habitat level | Raw cost | Optimized cost |
|---|---|---|
| 1 | 0.027 | 0.014 |
| 2 | 0.075 | 0.023 |
| 3 | 0.137 | 0.043 |
| 4 | 0.19 | 0.075 |
| 5 | 0.35 | 0.14 |
| 6 | 0.67 | 0.23 |
Looking at the optimized values, effort for citizen does a little less than double with each habitat level, which I think is really well balanced. Makes me wonder if the devs also calculated these numbers and did that on purpose. Even the ridiculously expensive VR Edutainment (it’s not quite made out of gold but probably silver or so) doesn’t spoil the chart. At the beginning you will be working mostly with the raw cost however. But as you reach level 4 or 5, some serious optimizations should start coming in and you may even find that going to the next level is not much effort compared to how much it took you to get thus far.
I think the habitat requirements are pretty spot-on in the current version. There’s not much I’d change (except maybe with the ratios of the different goods, in particular how demand for the same good goes up and down through the tier levels).
Important note on these numbers: This is with medium difficulty. Easy needs half of that, Hard needs 50% more. I think these are very reasonable steps. In the old patch I think it was -30%/+30% only? Funnily enough, if you play on hard and don’t optimize then level 6 is going to be tough on you as demand will reach about 1 worker per inhabitant.
Spaceship Components
Since starship components are not something you produce by the minute, I can’t actually give meaningful numbers here regarding cost and balancing, just based off of my calculations. While the upgrades are really strong, I think they are probably not too overpowered. The requirements seem a little wonky sometimes and also some combinations of materials are a little weird from game progression point of view, but I don’t have specific suggestions what needs to change. To give an example, in order to upgrade your motor output you also need reinforced iridium and AI control units, both of which are on a whole different tech level.
My biggest gripe with the current way spaceship components work is that, for each good whose production rate you are increasing, you first have to spend exactly that good. This is the case for almost all upgrades. That feels a little lame.
Terraforming
I calculated how much it costs to run each of the buildings for one minute and these are my results:
| Terraforming | Optimized cost |
|---|---|
| Dust Particle Filtering | 8.3 |
| Fertilizing Drones | 14.7 |
| Ground Acidity Regulation | 26.6 |
| Ammonia Extraction | 71.9 |
| Magnetic Field Generation | 341.7 |
Since the radius of each building is also increasing in that list from top to bottom, the numbers are roughly inverse proportional to how many buildings of each type you are going to build. That means you need to spend roughly the same amount of workers for each kind of terraforming. Ok very roughly, but still I must again congratulate the devs for how well these numbers seem to work out. (Disclaimer: I haven’t actually played the patch until teraforming yet myself, so this is theoretical only for me.)
Also if you consider these numbers you can get a feeling for how many workers you’ll need to terraform the planet if everything goes super smoothly, and I think it’s in a good region.
And now to my nitpicking about specific items 
Large Mines
Large mines are not really good. They weren’t fantastic before the patch and they still aren’t. Not completely horrible or useless, but productivity-wise they are not a big upgrade.
In some cases, big mines are even worse than small ones. Let’s look at the copper mine for an example.
| Property | Small copper mine | Large copper mine |
|---|---|---|
| Copper per min | 12 | 24 |
| Workers | 2 | 3 |
| Workers with robot | 1 | 2 |
| Power used | 2 MW | 6MW |
The large one produces twice as many while using less than twice as many workers, which is good. But once you add industrial robots, the worker requirement is also double. So on that side they are breaking even. But the large mine needs more than twice as much power, making it actually worse than the small mine. Now this is neglecting the cost of the robots, since the small mines need 2 in total to reach the same output as 1 large mine, but still I feel the large mine could be more useful. If you somehow get to use these mines for proximity bonus, then using 2 small ones is likely better than 1 large one because you get +20% instead of +10% only.
The main benefit of large mines is the reduced foot print. So whenever I get the tech, I am happy that I can get more out of my mineral deposits. But once faster transport is available and far-away spots are not so far away anymore and the laser can clear the rocks on the spots I am already mining from, large mines lose a lot of their appeal. If that is intended then fine, I guess. But I wish large mines would really feel a little special over small ones, but also that there are always places for both.
Atmospheric Processor vs. Oxygen/Carbon Processors
The carbon and oxygen processors are just really bad. If you can use atmospheric processors, then you are off much better. The dedicated ones use 50% more workers and 150% more electrical power. Now in the old patch they were great, because they used the same amount of workers and you would want your oxygen and your carbon divided into different places because they work in different industry clusters. But the new patch is throwing this design requirement completely out of the window because pipes are so d*** strong now. No efficiency malus and they are really cheap. That means you can just put your atmospheric processors where your carbon needs to be and put pipes from these areas to whoever needs oxygen.
So as long as your demand for carbon and oxygen is mostly balanced, there is no need for the dedicated processors at all. They become even worse once robots enter the stage (+67% more workers needed!). If your carbon/oxygen demand is lopsided then for the difference you can consider using the dedicated ones, but this is going to change quite a lot as you expand and build new industries, so getting this right is going to be super annoying and provides only a very small benefit.
Interestingly enough: How lopsided your demand for either is going to be probably depends on the difficulty you are playing on. Since oxygen only goes into products that are consumed by habitats, the citizen demand difficulty is determining your oxygen demand, whereas carbon is also used for a couple of other resource sinks. I actually like that it is that way.
Long story short, I think the oxygen and carbon processors should only require 4 workers like the atmospheric one. Give it a higher building cost and keep the ridiculous power requirement to counterbalance their usefulness and it’s fine. But 2 additional workers is absolute overkill.
Methane
Methane pumps are great, methane from fermentation is useless.
| Source | Optimized cost |
|---|---|
| Pump | 0.087 |
| Fermentation | 0.346 |
That’s a difference of almost factor 4! The fermentation needs a serious buff, or the pump a serious nerf. Or a bit of both. For the fermentation, I didn’t even consider any costs for producing food waste, since it occurs as a byproduct of good meal production. The fermentation plant just doesn’t produce enough methane with too many workers in the factory.
Now in the previous version I’d try to cover all my methane demands (at least in mid-game) from fermentation because transportation from the canyon was too much of a hassle. But now that we have such strong pipes, this concern is out of the window and I’m all for pumps. Maybe if methane pipes were a lot more expensive, then it’d also work out better for the plant… (which is a pretty expensive tech by the way)
Power Generation
These costs will be in “how many workers do I need to keep up 1 GW for 1 minute”.
| Power | Optimized cost |
|---|---|
| Wind | 0 |
| Solar | 26.7 |
| Methane | 11.7 |
| Nuclear | 18.6 |
| Fast Neutron | 16.8 |
As you can see, wind is obviously the best, since it is for free. But I think you can agree that it is not very convenient in mid and late-game, so I didn’t use it for cost calculations.
Solar was hit quite hard during the balancing update, but not quite to the point of uselessness. Especially with the upgrade it can play a role in mid-game.
Methane is the cheapest right now, because it produces a whopping 400 MW per plant and methane is really cheap to produce as its ingredient. I think it needs a nerf.
Nuclear power and fast neutron reactors are just for show. They are too expensive to operate with the power output they produce. In particular the requirement for enriched uranium is not cheap. I think simply increasing the times of the consumption cycles should do the trick to make these economical (and a nerf of methane).
Another thing about power is to consider how much of total production cost is used up by power. That number varies quite a lot for each good but most fall in the range of 0.5 to 4%. I think that is quite reasonable.
Vegetables vs. Meat
This is more of a joke than anything related to balancing so don’t take this too seriously. Without soil fermentation, meat is a little cheaper to produce than vegetables. This is far from the truth in our current food industry and even in the future with vat grown meat this won’t be the case, I reckon. Or else you’d also vat-grow the vegetables 
On the other hand, I suspect in a few hundred years in the future in a space colony they won’t be eating meat at all anymore.
Robots, Tools, Recycling
The cost for an industrial robot is 2.1. They last 20 minutes and will reduce the worker requirement by 1 for that time. In other words, each building needs only 1/20 robot per minute, so the price for each building becomes roughly 0.1. This is the cost I did not include in the calculations above.
With robots I was always wondering whether they were worth it and if so, by how much. Even without any optimizations, their cost is only 6.0, so yes, definitely worth it (they are worth it if it’s less than 20.0). The good thing about robots is that it doesn’t matter where you send them to work. They just reduce the amount of workers by 1, so any workplace that can be upgraded is equally good.
High tech tools cost 2.4 and also last for 20 minutes. But how much do they save? As I said, I didn’t consider them in the calculations because under my premises, they would just cut all cost by an equal amount. However in practice, there is one thing to consider: High tech tools are much more useful if employed in buildings with high worker requirements. For example, a holo display factory (without robot) needs 15 people. Add a 50% bonus from high tech tools and the same output can be generated by just 10 people in that factory, a saving of 5 people for the duration of 20 minutes. That is really strong. But put the same tool in a large copper mine (which needs 3 workers) and it will only save you a single worker, which is the same cost reduction as an industrial robot, for a slightly higher investment.
Also interestingly enough, the raw price of high tech tools is lower than that of industrial robots: 5.4. The reason for this is that robots get a spaceship component which boosts their productivity by 33% and thus creates a cost-down of 25%. High-tech tools don’t have that (unless I’m blind ).
Recycling is also interesting to look at and analyze the numbers. Before, I always assumed that recycling makes a product cheaper by 17% per step in the production chain, leading to ridiculous price decay for high-end products. But this is actually not the case. Most goods overall benefit rather little from recycling, usually only 2-5% per step in the chain. The reason is that source material is quite cheap and a decent chunk of a good’s cost comes from the factory making it. Higher tech production buildings also have higher labor costs. Of course on long production chains you still rack up a nice total bonus, but by far not as much as I had always assumed.
Red science reigns king here with 25.8% total cost reduction if you apply recycling all the way through its production chain. But if you look at something simpler like home appliances, they get a meager 7.0% only. Now on the one hand side recycling centers cost nothing in the metric I use, i.e. they are for free. So can’t complain about a free bonus, can you? Actually I would like to anyway I think changing their effect from 1 in every 6th round to 1 in every 5th round wouldn’t be too bad and create more incentives to actually build them and ruin your aesthetically perfect industry clusters with specks of recycling centers dotted all over.
Wow that was a lot of text. Hope somebody found this useful 