Do we really need carrying capacity? Why not use housing directly?

[mial42]

Better to remove logistic growth entirely. It's an obtuse system that serves no useful purpose and causes several major problems. That carrying capacity is ~80% redundant with housing just makes it worse, but is not the main cause of the problems with it.

 

[Annihilat0r]

Better to remove carrying capacity entirely. It's an obtuse system that serves no useful purpose and causes several major problems. That it's ~80% redundant with housing just makes it worse.

Isn't that exactly what I said in my OP?

 

[mial42]

Isn't that exactly what I said in my OP?

I should have been clearer. I meant "remove logistic growth entirely." Edited original comment for clarity.

 

[Ludaire]

But it remains very vague just what this is simulating. How does a housing district improve a planets carrying capacity? Why does a farming district reduce it? It seems unnecessarily confusing.

Because fewer people can live in a district's worth of farmland than a district's worth of city. And undeveloped is assumed to be somewhere in the middle or the average. I don't see what's so confusing about "this area of land can probably support X people, but if we turn it into a super dense city, it's more and if we turn it into a much more population sparse farm, it's less."

I'm an advocate for significantly reducing the number of pops and jobs the average planet can support. This deals with both performance and economic problems (currently, there is an overabundance of everything - space, resources, building slots).

Sure, and if they do that rebalance, I'd be perfectly fine with it. I would still hope that we have different planet sizes and that the planet sizes matter enough that you see a size 25 and go "Ooh, awesome! That's really great find!" instead of "Eh, I already have five large industrial/research worlds, so I guess it'll be a 10 district rural growth place instead of anything exciting." Not needing to build any or at least not as many city districts is a nice way of handling that.

 

[Brael]

I should have been clearer. I meant "remove logistic growth entirely." Edited original comment for clarity.

Pop growth isn't sustainable, if it's linear per planet, but exponential as more planets are filled and then drops off to a linear increase once all living space is gone. Unless other mechanics are used to taper pop growth, logistic curves are necessary in order to keep pop growth within decent performance requirements.

So the solution is to either slow pops, implement mechanics that cause huge devastation and kills massive numbers of pops on a routine basis, or move away from the pop system entirely.

Of these, a lot of players like the third option but it's a core part of the games mechanics so it's a really tough sell to devs to change what the game is about. The second option is incredibly unfun, which leaves managing pop growth as the most fun solution

What should probably be done, is to leave things as they currently are, but add constraints that limit how many planets can be settled and when. A lot of the current feel bad involves planets just being too unpopulated because people expand a lot faster than pop growth does. So, if planet acquisition is slower then you avoid a lot of the issues that spread population around too much and lead to current issues.

 

[mial42]

Pop growth isn't sustainable, if it's linear per planet, but exponential as more planets are filled and then drops off to a linear increase once all living space is gone. Unless other mechanics are used to taper pop growth, logistic curves are necessary in order to keep pop growth within decent performance requirements.

So like the mechanic (existed in 2.*) which stopped growth after a low level of overcrowding? Logistic growth massively increases pop counts, so logistic curves are doing the opposite of what you're suggesting. Not to mention that assembly exists.

So the solution is to either slow pops, implement mechanics that cause huge devastation and kills massive numbers of pops on a routine basis, or move away from the pop system entirely.

Of these, a lot of players like the third option but it's a core part of the games mechanics so it's a really tough sell to devs to change what the game is about. The second option is incredibly unfun, which leaves managing pop growth as the most fun solution

What should probably be done, is to leave things as they currently are, but add constraints that limit how many planets can be settled and when. A lot of the current feel bad involves planets just being too unpopulated because people expand a lot faster than pop growth does. So, if planet acquisition is slower then you avoid a lot of the issues that spread population around too much and lead to current issues.

The obvious solution is to slow pops, but in a way that actually slows pops (unlike logistic growth) and doesn't cause huge issues with the game. There's a number of obvious ways of doing this, one of which (using 200 instead of 100 as the growth/assembly threshold) I've implemented in my mod reworking pop growth and conquest. Or the much-suggested halving of jobs and housing. Combine that with removing or limiting habitats (habitats are responsible for most of the issues; without them the endless expansion of growth sources isn't a thing) and you're 90% of the way there (which I know, because I play with my mod and it works to keep pop counts down and performance up). Logistic growth does not solve this problem, it makes this problem much worse.

 

[Andy_Dandy]

Sounds terrible what Thread Starter suggests. Building housing is an easy non-brainer, and carrying capacity is an elegant way to balance diferent planet classes like Gaias and habitats.

 

[Ferrus Animus]

So the solution is to either slow pops, implement mechanics that cause huge devastation and kills massive numbers of pops on a routine basis, or move away from the pop system entirely.

Or option 4: reduce pop calculations.

 

[Farbros]

Definitely agree with the op, it would do a much better job, the numbers of pops will be limited to the number of planets/ habitats, and they are limited with influence, which is limited to the number of year the game goes. So if someone wants to have a huge galaxy with huge number of pops - they can create a game until 2600, someone who wants to have good performance, can create a game until 2400 and have a nice performance, without need of all of these limitations.

 

[Brael]

Or option 4: reduce pop calculations.

You can't do that without eliminating pops as a mechanic, which is undeniably far more computationally efficient and pretty easy to fake that pops still exist. But, this is likely an absolutely massive refactor and seems to be against what the devs want to do with the game.

 

[mial42]

You can't do that without eliminating pops as a mechanic, which is undeniably far more computationally efficient and pretty easy to fake that pops still exist. But, this is likely an absolutely massive refactor and seems to be against what the devs want to do with the game.

There's still some low-hanging fruit, like faction demands using any_owned_pop as of 3.0.2. That might be fixed in 3.0.3 (they did add a species migration control trigger like I mentioned in that suggestion, although IDK if there's any causality there), but there's also stuff like happiness and faction membership/approval being checked daily (only needs to be checked once a month, maybe once a week if you want more responsiveness). And many of the events that don't use is_triggered_only = yes could be rewritten. No idea how big the gains from doing that would be and I support generally fewer pops anyways, but it's there.

 

[Ferrus Animus]

You can't do that without eliminating pops as a mechanic, which is undeniably far more computationally efficient and pretty easy to fake that pops still exist. But, this is likely an absolutely massive refactor and seems to be against what the devs want to do with the game.

Reducing, not eliminating. The game has a ton of legacy inefficienies from the time the game had less pops and mroe from way too many checks done when 2.2 implemented them without regard for CPU burden. Like mial42 lists.

In addition there could also be a limiter implemented, making pops not calculate everything all the time.
I posted this elsewhere, but imagine limiting standard calculations to 1 pop per planet per day. Most stuff about a pop doesn't change often, so there's no need to have it checked daily or monthly. Like a pop wont need different housing, amenities or change it's faction stance that often. And such a limit would reduce calculations in a big empire with say 50 planets and 2000 pops by a factor of 40!

Add some exceptions like for aassimilations, genemodding and so on to make it work better in response to events, but I think we can reduce the current lag issues a lot by simply addressing the mechanics behind them.

 

[Annihilat0r]

Let's not derail this into another performance thread please. Personally, I am actually very happy with the performance of the game in 3.0. I started this thread, because I was frustrated with the pop growth due to not noticing the concept of planetary capacity at first. I think most casual players will miss it, since it is a very obscure mechanic.
Housing is something that players are familiar with, that follows a simple logic, and that is prominently displayed in the UI - thus in my opinion much better suited as an anchor for the S curve.

 

[Olterin]

I started this thread, because I was frustrated with the pop growth due to not noticing the concept of planetary capacity at first. I think most casual players will miss it, since it is a very obscure mechanic.
Housing is something that players are familiar with, that follows a simple logic, and that is prominently displayed in the UI - thus in my opinion much better suited as an anchor for the S curve.

Well, given how housing feeds into planetary capacity... they're not that decoupled. I do agree, though, in a sense - namely that Planetary Capacity, now being the primary mechanic interacting with pop growth, should replace housing in the UI. Possibly even entirely - have districts modify base planetary capacity, have pops take it up, modify all traits that currently interact with housing to instead interact with planetary capacity taken up.

 

[NinjaBus]

Isn't the obvious solution to the relationship between capacity and housing just to have empty, unblocked, districts provide 1 or 2 housing? It sidesteps the whole issue.

Merge the growth effects of capacity into the housing, and have available districts provide some so you don't start with 0.

 

[tobias.mb]

I do think using available housing instead of total housing makes more sense.
However I actually like the idea of getting some capacity from empty districts. That way newly colonized worlds can grow well, even if you don't immediately build all the districts.... (esp. important early game)

 

[DrNukeLear]

I’ve been playing with a pop balance mod which completely rejects the new mechanics. Using housing to slow and halt pop growth is very effective. I’ve reduced housing by 20%, implemented exponential growth (more pops in colony = faster growth), removed the empire scaled pop growth modifier, cut base pop growth and my performance is currently faster than 3.0.3 for the first fifty years. So yes. Basing planet growth limits around housing to curb and then stop growth works.

 

[Trokkin]

I’ve been playing with a pop balance mod which completely rejects the new mechanics. Using housing to slow and halt pop growth is very effective. I’ve reduced housing by 20%, implemented exponential growth (more pops in colony = faster growth), removed the empire scaled pop growth modifier, cut base pop growth and my performance is currently faster than 3.0.3 for the first fifty years. So yes. Basing planet growth limits around housing to curb and then stop growth works.

I wonder what happens if the curve was actually applied to colonies at low pop count, so that a 2 pop colony would have a penalty of ~95% growth no matter how much capacity there is. I predict a similar effect you had. Although tbh, at 50 years you probably have more pops in 3.0 than 2.8, you'd need to make comparisons at 100 and 150 years, cause that's where 3.0 was aiming at with their new growth mechanics.
Also interesting twist could be brought if high stability decreased pop growth speed, as it does corellate IRL, where developed countries with high quality of life has significantly lesser birth rates.

 

[HFY]

I wonder what happens if the curve was actually applied to colonies at low pop count, so that a 2 pop colony would have a penalty of ~95% growth no matter how much capacity there is. I predict a similar effect you had. Although tbh, at 50 years you probably have more pops in 3.0 than 2.8, you'd need to make comparisons at 100 and 150 years, cause that's where 3.0 was aiming at with their new growth mechanics.
Also interesting twist could be brought if high stability decreased pop growth speed, as it does corellate IRL, where developed countries with high quality of life has significantly lesser birth rates.

The original Carrying Capacity mod did that -- growth at low population was very, very slow.

You needed immigration to get a reasonable population, and at the start of the game that meant either moving pops off your homeworld -- you were either focusing your pops on building up your Capital, or you were spreading them to the colonies as an investment.

With 3.0 code changes, this would be viable: you could just turn off jobs on your Capital and let the pops auto-migrate, painful but free; you could use your free building slot on a Robot Assembly Plant or on Clone Vats, and assemble your way up; or you could manually resettle to get your breeding base up to ~10 or so, not free and thus differently painful.

Not sure how well the AI would handle it in 3.0, but it seems like auto-migration might save them.

 

[Trokkin]

The original Carrying Capacity mod did that -- growth at low population was very, very slow.

You needed immigration to get a reasonable population, and at the start of the game that meant either moving pops off your homeworld -- you were either focusing your pops on building up your Capital, or you were spreading them to the colonies as an investment.

With 3.0 code changes, this would be viable: you could just turn off jobs on your Capital and let the pops auto-migrate, painful but free; you could use your free building slot on a Robot Assembly Plant or on Clone Vats, and assemble your way up; or you could manually resettle to get your breeding base up to ~10 or so, not free and thus differently painful.

Not sure how well the AI would handle it in 3.0, but it seems like auto-migration might save them.

Yep. So, this would effectively remove the rush to settle any new colonies before any troubles with overpopulation, or before you see rare resources on a planet. (well, if they fix robot pops contributing to growth curve and probably increase robot tech tier by one) That would very well change how the game is played, shifting focus from colonization and conquest fever to developing a good economy.

And yeah, auto-migration is a blessing for everyone.