Stellaris

Stellaris

Administrative Upgrades
Nightskies 11 Mar, 2019 @ 3:58am
Upkeep and balance (2.2.5)
It's been discussed in comments, I think it's worth a discussion.

First, a glance on the gains of the mod's building earnings. Assuming the empire has some Sprawl beyond the Administration Cap, one tier 1 admin building effectively gets the following per month:
1.2% of total empire Research of each type, 2% of Unity, 4% of leader upkeep Energy, 2 clerk jobs (4 amenities, 4 trade), 1 Energy + 1 Society upkeep. An easy first step into expansion.

Now a early-mid game empire. I plugged in the numbers to my 150 extra sprawl empire. 9 tier 2 buildings produce the following, each:
(4.8%) About 32 Research of each type, (8%) 18 Unity, (16%) 4.3 Energy, 3 clerk jobs (6 amenities, 6 trade), 2 Energy + 10 Society upkeep. Mostly covers some developed worlds.

An late-game empire, 480 excess Sprawl example. Tier 3 buildings, each:
(9.6%) About 250 Research of each type, (16%) 333 Unity, (32%) 18.4 Energy, 3 clerk jobs + 1 Admin (3 Unity, 14 amenities, 6 trade), 19 Energy + 60 Society. Covers most developed single-world sectors and you shouldn't need more than this.

And just for completion. Tier 4.
(19.2%) Tons of Research, (32%) Now What Unity, who cares% leader upkeep, 5 clerk jobs + 1 Admin (3 Unity, 18 amenities, 10 trade), 19 Energy + 200 Society upkeep. Covers your three-planet homeworld sector and the greedy hyperlane trolls collecting tolls.

These numbers should have clear meaning. But just to be extra clear -> In the early game with extra sprawl, they're hands-down the worst thing to build except the silo, but by mid game you should be building tier 2 on as many worlds as you need until you reach the Sprawl level and only upgrade if there are no more planets available. They're basically empire-wide self-additive multipliers that stack with all other multipliers for Research and Unity. And the larger your empire is, the larger the multiplier can get. Only they don't apply to special projects. You should know how big those Sprawl debuffs are- these buffs are equally big. With this, it is easy to completely counter the Sprawl.

SOCIETY UPKEEP: In the endgame, the extra Society-producing sources offset the upkeep a fair bit. If wisely spread across your empire, you might not even notice Society advancing any slower (perhaps even faster than Engineering). If prolific with these in early-mid-game, you might hurt yourself with those special projects, namely species altering. In the end, the Society upkeep ends up not actually being a cost, but a Research check.

UNITY UPKEEP: Suffers the same as Society upkeep. As long as it's moderated, you'll just make more Unity than the upkeep. A lot more- having a large flat cost just delays when that happens.

Goal Check: Do you want a natural feeling way to deal with spread, or do you just want to get rid of it? I like the feature, myself, but would like a way to mitigate it, perhaps eliminate for an actual cost (not a resource check).

Easy Solution: What does the game do with Every Other Empire Wide Buff? Limit 1, and it's pretty expensive. It shouldn't be able to eliminate wide Spread like this, but it is easy and doesn't break the late game.

Solution 2: Drastically reduce the Admin Cap and change upkeep to more Energy + AMENITIES. I'd recommend 5 cap to 6 energy 8 amenities, 10 cap to 14 energy 22 amenities, 20 cap to 22 energy 48 amenities, 30 cap to 30 energy 70 amenities. Perhaps crystals would be a good addition, and a bit less amenities. No limit to # on planet, so you could build a true administration world teeming with entertainment districts or just sprinkle them through the colonies. It would be as vital to the nation as the first Tech World and Capital World are, just later down the line.

This is high-intensity office work smoothing the lives of billions in multiple dimensions and environments demanding state-of-the-art fences security and monitoring systems. It identifies individuals suitable for demanding jobs to help make the most of the minds of the nation. I suggest Amenities because Administrative Cap doesn't affect it, but it directly affects productivity on that specific colony as it interferes in lives but improves the greater good. You could build one on every colony, but that many amenities will call for some additional redirected labor and building space, very seriously impacting economy and reducing specialization (except for a Resort World). At some point you might even elect NOT to reach cap in interest of more specialized economy-focused planets. Heck, maybe even keep that administrator job or two, but I think the clerks should go.

Remember, 30 cap = 9% empire-wide research that stacks with everything, even more Unity, and a noticeable bit of energy. That cap should cost something.

Still, Solution 3. Shift to starbases with imperial administration Modules, with a large energy upkeep and unity/social research cost. This is a linear progression but starbase space is in high demand. The cost here is more a choice- largely against fleet cap, trade lanes and defense.
Last edited by Nightskies; 11 Mar, 2019 @ 4:05am
< >
Showing 1-13 of 13 comments
tonechild  [developer] 11 Mar, 2019 @ 5:47am 
That's a lot to take in and I appreciate the effort done calculating how this mod affects the game balance (very much something I am interested in) - before we dive in, I want to mention I have a plan to add modules to stations that increase admin cap, so any solution we come up with should consider that as a very real option (maybe it is researched or maybe not available first), another thing that seems very enticing is adding a specialized world for admin cap and balancing it against ameneties liek you've suggested, that seems very cool.

You've mentioned a lot of percentages and I'm trying to figure out how you came up with them. The way this post reads I almost feel ashamed for asking but can you explain how you came to those conclusions? What variables are you factoring in to get those? What could I myself do or anyone else involved in the discussion do to come up with the same percentages.

Now, after we get on the same page with percentages, I'm willing to walk through the rest of the post. I have indeed read the entire post twice, but I'm failing to draw the same conclusions mostly because I'm having difficulty coming up with percentages. Anything like a spreadsheet could be helpful, if this seems like too much work or effort on your part then I totally get that and I might just try to dive into this more with some suggestions, I'm willing to put up the spreadsheet and do just as much work as well.

Lastly, I just want to point out my goal with this mod:
1. it should have a risk / reward, meaning - completely nullifying the debuff should not be possible without a fair tradeoff.
2. There should be multiple ways to manage the sprawl using this mod, so that players can balance the tradeoffs effecively with their playstyles.. What I would consider game breaking is making it too easy to formulate how to nullify the the sprawl debuff and nullify the tradeoff at least quickly.
3. Not sure how to balance end-game, as in the end-game I just have researched everything and I'm either stomping everyone or I'm in a complete stalemate. To me, the early/mid-game is where my decisions have the biggest impact, where as in late/end-game it feels like everything I do has little impact on the game, because I've either snowballed out of control or I'm in a stalemate and the game is running slow. This is mostly an excuse to not try to spend too much effort balancing the end-game, that being said, I am open to balancing for end-game so don't let my apprehension discourage you.

Last edited by tonechild; 11 Mar, 2019 @ 6:11am
Nightskies 11 Mar, 2019 @ 12:47pm 
No sweat, actually. The math is pretty easy. Each point of Sprawl over cap has the following effects on empire. Can be seen by hovering over Sprawl in game. I checked that the tooltips were right, they pass (although they round a little weird).
+0.3% Tech Cost
+0.5% Tradition adoption cost
+1% Edict cost (not universal)
+1% Leadership recruitment and upkeep
So these don't directly add to research and unity, but they do directly affect how fast techs and traditions are completed, and effectively act as a research/unity multiplier by changing the cost of techs and traditions. So, 30 points of sprawl directly increases tech cost by 9%, and inversely, adding 30 points of admin cap negates that, reducing cost (and therefore faster to complete) by 9%. The numbers from my empires were assuming I went all out and negated the sprawl, which would be about as hard as building a tier 2 Temple on every world. Not hard, in other words. It's a fixed ratio, so it doesn't matter if the empire has bare-bones Research or has a bunch of tech worlds- it's always effectively doing a fixed % of total empire output. As posted above!

Drawn out math, let's take a later game example. 55,000 research tech, in that 480 sprawl example, ends up with a cost of 134,200. With 2800 post-multiplier Research (that is, say, 1500 Job and Station output multiplied by empire modifiers and leadership values), the tech normally takes 48 months. But with 480 Administrative Cap, it goes back to 55,000 cost- nothing else affects tech costs. Now the same 2800 Research gets it done in 20 months. 240% faster. To get 480 administrative cap, you'd need either 30 tier 2 admin buildings, or 15 tier 3. This example happened to be what I was researching with Social, so looking at the upkeep, I'll now show how the upkeep affects it.
The base research before multipliers was about 1500. 30 tier 2's eat up 300 research, dropping base research to 1200, which drops total output to 2240. I had 30 worlds, so that is easily achievable. Now that Society tech, still costing 55,000 instead of 134,200, gets done in 25 months. 192% faster. Those 30 facilities each contribute 1/30th of that- effectively the difference between 134,200 and 55,000- 79,200 Research in those 25 months. That's 105 Society per month per Administration building at tier 2. Even if they were instead 15 tier 3's, it would impact base research a lot more, but still ends up getting the job done in 49 months- not really slowing down Society. It does do a % of total empire output- so the larger the empire, the more work the Administration effectively does (hey, the math makes sense). If I had to use tier 3's, I'd have a lot more Research.

Let's see... I'm in the same boat with totally snowballing late game. This empire is a good example of that- rather inefficient as I got lazy after victory was all but assured as the 400 lb gorilla in the playground of children. If it had the admin cap, though... sharks with lasers man... but I did notice this seemed to hit it's prime in the early-mid game. Totally on board with that, the current numbers most excite at that point where the snowball gets going. In the end, end-game balance as a whole can't be helped with this kind of mod, but if balanced, it won't exasperate it at least and the game can be enjoyable longer.

Hmm, as for a spreadsheet... first, let's see if we're on the same page after that.

Incidentally, the tipping point an empire should start building these, is when it has about 400-500 total research before multipliers (shown at the top of the UI). At that point the empire can afford the Society cost without crippling it, and it should be in full swing in the empire at 600 total output. My early-mid game example was at 1000.

I was thinking more about the proposed Amenities, and I think it should also keep the society cost or unity. Because of the above paragraph. The only reason tier 1 buildings aren't recommended is because they're too weak in the early game- compared to other buildings.
Last edited by Nightskies; 11 Mar, 2019 @ 1:12pm
tonechild  [developer] 12 Mar, 2019 @ 9:20am 
I think I get what you're talking about now.

So here is what I was thinking initially:
1. You had percentage values that indicated the sprawl debuff.
2. You had percentage values that indicated the negation of the sprawl debuff

However, it seems you were only talking about the sprawl debuff which is the following when you are +1 over the admin capacity:
+0.3% Tech Cost
+0.5% Tradition adoption cost
+1% Edict cost (not universal)
+1% Leadership recruitment and upkeep

Bear in mind, we have megacorps that apply a 50% increase there.

Percentages get me confused sometimes especially when they are compounded.

Ok, so I'm thinking of building a spreadsheet that directly takes the sprawl debuff, then we just need how everything adds to the sprawl itself, which can probably be taken from the wiki.

Now, a lot of consideration and rightly so has been taken into account for the "research check"

This is where it gets tricky I think.. so you have this:

"Now the same 2800 Research gets it done in 20 months. 240% faster"

That is true, but we're comparing a certain amount of sprawl excess which could vary, so while the speed is 240% faster, it's not a hard number - it could be 200% faster, or 300% faster, depending on other mechanics, right? So what would be the mean I guess?

So here's what I think we should take into account (keep me honest here, we want to take into account all the important factors) for a spreadsheet

First iteration:
1. the cost of 1 point over admin capacity against tech and tradition, ignoring edict and leadership cost (we do have edicts and leaders that increase research speed but lets ignore those for now)
2. the amount of points of admin cap used for a single planet, and for a starbase.
3. come up with an average for early, mid, and late game - using the years passed in game as our metric
4. then come up with the admin cap increases our buildings apply.

We just need to have some assumptions, like: what should the size of the empire be for "early" etc, as the more planets you have the more buildings you can add.

Then we can find out how much society cost the buildings should have based on averages - which I think would be the easiest way to balance this. If we want to apply unity as a cost, we could plug that in as well, I was thinking 0.1, 1, 10, 100 from tier 1 to 4. But that's if we decided that unity should be used instead, and I'm on the fence there.
Nightskies 14 Mar, 2019 @ 10:02am 
1 -
At first I thought the effect of Sprawl was complicated, but it's really simple. Reducing tech cost is effectively increasing total Research output, same with Traditions (and by extension, Ambitions) to Unity assuming no Administration Cap is wasted. Each point is worth 0.3% of the empire's Research, 0.5% Unity, not counting Edicts and Leadership costs. Of course, Megacorps are 0.45% and 0.75%, respectively. Finding that out was complicated though! Admin Cap is like other multipliers to empire total output, it just varies through the game. Not like Job multipliers, or planets, or mines, which are considered to come up with the base value of an empire's output, which upkeep then draws from before doing total output modifiers. I spent quite a few hours now with some minor discoveries to make sure of that. Well, I just had to make some formulas, which really was kind of wasted time, but it did assure me of these statements. It really is effectively a simple flat % of total output, not counting special events or projects. Some people think Administrative Cap is not fully effective if there is still excess Sprawl, but it's not true- the excess Sprawl affects all research and unity multipliers to the same degree as the effect of Administrative Cap.

And so, as shown in the first post, the effective output of the Admin buildings as a % of total output. The examples were just for reference of live empires for comparison to other buildings and really the % is really what to pay attention to... I didn't have a good example for a 4th tier on hand. My empires have taken over the Large galaxy by that point!

2 -
Each system, colony, and branch office adds 2 to Sprawl. Each district on a colony adds 1. Starbases do not add to Sprawl. So a typical 18 district colony is 20 Sprawl. A fully developed, large world can be worth up to 28 Sprawl. A fully developed Habitat, 10 Sprawl. A full Ringworld megastructure is worth 212 Sprawl. I average a 1/3 to 1/4 ratio of colonies to systems, but it could easily go either way depending on play style, like building tons of habitats or just sticking to ideal worlds without terraforming. Based on my playstyles, we're looking at about 32 Sprawl per developed world and it's surrounding systems. Maximizing district efficiency with extreme planet specialization at the cost of total output could drop that, but that's a LOT of micromanaging.

3 -
Okay! Now a chart for some noteworthy numbers are handy. I'm starting that today. It'll be blank until I get to it tonight. Feel free to do some formatting if you want, I'll happily throw in the numbers :)
https://docs.google.com/spreadsheets/d/1x4UnXySoqkCbRJi-o2EzMBEPjYJ7KeNFWtu0HjKxPVc/edit?usp=sharing

4 -
Okay!

I have some more thoughts on Unity vs Social Research later. I'm leaning toward Unity but see the merit of Social, even though they both currently would have very similar effects on gameplay. As is, being a research check, it doesn't really matter unless someone upgrades too early. But those people deserve the fate that's coming.
Nightskies 14 Mar, 2019 @ 2:37pm 
I might not have been clear, but you're right that the "240% faster" very much varies like you said. It goes from unnoticeable to magnitudes faster. I'll include examples in the chart, and hit on factors later.
Nightskies 14 Mar, 2019 @ 10:02pm 
Alright, I got as many numbers as I can fill for now from two empires- my other empires don't have any useful information due to incompatibility. If you can, add some in the 2nd sheet (at the bottom) or 3rd for Megacorps. To get total Research, open the tech tab in game and hover over the +% number on the right side of each type of research. It'll show total output and the raw output. If one is much larger than the others, check for a "Stored Research" modifier, and subtract that value.

If you're looking for the 240% example, it got lost in the autosaves and I'm using the average of Physics/Society/Engineering, and Engineering always drags the average down.

Looking at Unity vs Society. For one, I favor Unity because it doesn't actually have any empire-wide modifiers that apply after base is calculated, the base earned is equal to total output. So upkeep of Unity is more directly understood and felt. Second, Research can be stacked up more easily than Unity- you can start a new tech world anytime, but Unity can hardly be stacked on a second world after you place the Ministry of Culture. While some empires can earn a lot more Unity, anyone can just plop down more labs. Thematically speaking, they both seem plausible- the governments and people inclined to produce Unity seem more likely to sport a robust administration than a scientific-oriented one. An empire short on Unity seems like it will suffer just as one short on Society Research.

Alright, more tomorrow!
Nightskies 15 Mar, 2019 @ 1:27pm 
What other numbers or calculations might be useful on the chart? Perhaps colony count?

You might have noticed the Eaton Nir empire was built tall until 2360-ish.
Nightskies 16 Mar, 2019 @ 9:14am 
I'd really like to see some numbers plugged in from an empire going gung-ho on wide play, regardless of the use of admin buildings. I've also got some gaps in the two examples I put in, so someone with a little more through set of data could be useful.

Of course, we could also just plug in some estimations. I switched protection to the second sheet so anyone can just plug in #'s in the data input.
Last edited by Nightskies; 16 Mar, 2019 @ 9:15am
tonechild  [developer] 16 Mar, 2019 @ 9:29am 
I just finished a game and am annoyed with myself for not recording the numbers. I played vanilla without any mods so I could get some steam achievements. When the game ended, I had around 480 sprawl, and my admin cap was at 190. I had a scientific nexus, and I had a ton of research buffs. Around then, 200k research took around 40 months to complete or so.

If I had no sprawl debuff, what would my research even be for the 200k items? I might load that last game up to see what the debuff percentage was. But it seemed like it would be amazingly fast and I'm OK with that xD.

So looking at the spreadsheet, more in a few minutes...
Nightskies 16 Mar, 2019 @ 10:09am 
The 200k research was probably already including the Sprawl debuff (which was 87% given that sprawl and admin cap, for a base cost of about 105k), so you probably had about 5000 total output of whatever research type it was. Looks to be really close to the last line of my Xirmian example, but the Xirmians had to deal with excessive useless tech from mods and didn't have as much late-game tech progress as seen in the smaller Admin Cap.
Long-comment-san 1 Sep, 2019 @ 4:15am 
Damn this thing is OP. Basically it eliminates the sprawl at some very-very low cost. And gives you tons of jobs. I HONESTLY believe, that max admin capacity per planet should be something like 8-12. Because if you get more than 15, you mostly compensate for 50-100% of the planet sprawl. And you want sprawl. Having almost zero sprawl lets you expand almost infinitely without any kind of cost and penalty to technological growth. This is a big deal in the early game when you typically dont have adm capacity research and expand agressively and you have a lot to research, so say +50% sprawl can even offset the benefit of having more territorry and resources. These building should imho serve to reduce the planet sprawl, but definetely not eliminate it as a concept.
Long-comment-san 1 Sep, 2019 @ 5:26am 
Yeah, I manually reduced to 18 adm capacity at max, reduced jobs by around 50%, lowered upkeep a little and.. hear me out. I have a 400 star galaxy with 0.25x habitable world's. I have roughly 20 colonies and 470 sprawl. Modded mod buildings give me 146 sprawl reduction. +60% to tech cost. Honestly, this is a HUGE reduction in sprawl. Even after me nerfing.
What I want to say is, that if I played x1 habitable worlds, after all the nerfs, I would totally eliminate the sprawl. I think 18 adm capacity at max level is way too much..and your idea is 64?;D
Optimal is something like 12 at max level... My humble opinion, though!
tonechild  [developer] 2 Sep, 2019 @ 2:16am 
Thanks for your input Sliderpro! I can't even use this mod because its so OP. I've just been reluctant to make any changes.. I might either reduce this version, or release an alternative.
< >
Showing 1-13 of 13 comments
Per page: 1530 50