Author Topic: Population growth model and terrain rebalance/nerfs  (Read 1150 times)

0 Members and 1 Guest are viewing this topic.

Offline lolada

Re: Population growth model and terrain rebalance/nerfs
« Reply #30 on: September 16, 2020, 09:46:25 AM »
Its mixed there you are partly right in theory, but it doesn't feel right in gameplay terms to me.. Like you are right that long-term food prevails - in theory leading to stronger cities, but does that really matters, when you use forests and overrun opponents who don't have minerals? Then you get double the amount of bases and it snowballs. Then as a player especially you can't really grow as much as you would like. Your logic is more applicable to AI, see for example Lal here with size 12+ bases and more space to grow.. he went to 15-17 pop a bit later.. while i get stuck at 7-10. I need to extract as much minerals as possible to compete - build units, secret projects, infrastructure - since its impossible for me to grow more.

Player simply values minerals much more because minerals lets you do stuff. Nutrients are needed just to a point. Basically in ideal situation one could go high nutrients grow and then switch to forests and minerals.. but it usually does not go that way due to drones, SE and safety. High pop and energy lets you research, but you get stuck unable to build expensive units. Player go around and build glass cannon while AI gets stuck build super-duper probe unit for 12 turns that it then loses attacking 0-1-1 probe.

If you could use energy to build units... ans grow more... its very inefficient at 4 energy per mineral rushing. Thats why player rushes infrastructure (at 2 energy per min..) and hard builds units.
And I don't get that 1 forest per 2 farm+solar logic. Smart thing is to work as much good tiles as possible. I can't choose to work 6 farm+solars or  3 forests.. i end up with 6 pops for example.. cant get more due to drones and can't grow so i have no use for food. Then I want to work as much min/energy as possible with those 6 pops while being neutral on food. So basically the way it plays out.. food is important early up to a point and then you want to switch to minerals and 2-2-1 tiles are ideal since they are food neutral.
Later when you want to grow more you could really use more food i guess - thats why i don't forest everything, but keep 3+ food tiles "alive".. The way to keep more farm tiles from replacing by forests - would be to enable Soil enrichers earlier or Condensers earlier do that farms don't get replaced in the first place. Its really impractical to replace forests later because they cover terrain - PRACX has those layers so one in theory could do it, but its way more micro than i am willing to do. Then there are multipliers.. like Recycling tanks, Genejack factories.. there are nor multipliers for food.

Anyway its complex balance issue - I don't see how its good to rely on Farms+Solars if you don't have multiple mines around and some extra nutrient resources to actually use them. Thats exceptionally good base to have - its rare. Even then you get stuck with eco damage from mines and forests help there. You also can't say you can use Soil enrichers - because by the time player gets them its mostly game over. I use them over farms that i have left.

tldr... Tree forests  2-2-1 are great and they kill lots of farms/solars from the map. Is it good or bad i don't know, I prefer to see those mirrors and condensers they are cooler high-tech environement than forest almost everything. I'd almost like to have special improvement that adds 1 min to farm+solar without eco dmg.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 4739
  • €2302
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC2 Hall Of Fame AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: Population growth model and terrain rebalance/nerfs
« Reply #31 on: September 16, 2020, 03:13:57 PM »
I'd almost like to have special improvement that adds 1 min to farm+solar without eco dmg.
"Rock fruits?"

Offline lolada

Re: Population growth model and terrain rebalance/nerfs
« Reply #32 on: September 16, 2020, 04:02:10 PM »
Quote
"Rock fruits?"
Multivitamins and minerals - we have it now  ;stupid Magnesium, Calcium that stuff. Its already in forests so they get 2-2-1.

Re: Population growth model and terrain rebalance/nerfs
« Reply #33 on: September 16, 2020, 05:50:07 PM »
Its mixed there you are partly right in theory, but it doesn't feel right in gameplay terms to me.. Like you are right that long-term food prevails - in theory leading to stronger cities, but does that really matters, when you use forests and overrun opponents who don't have minerals? Then you get double the amount of bases and it snowballs. Then as a player especially you can't really grow as much as you would like. Your logic is more applicable to AI, see for example Lal here with size 12+ bases and more space to grow.. he went to 15-17 pop a bit later.. while i get stuck at 7-10. I need to extract as much minerals as possible to compete - build units, secret projects, infrastructure - since its impossible for me to grow more.

You are right that my theory was one sided. No one is channeling all nutrient surplus into growth. It is distributed between growth and supporting poor nutrient mineral production tiles: forests, rocky mines. However, even in their mineral supporting application farms + mines are better than forests. You can review it yourself at different stages (TF, HF, enrichers, etc.) with different worker placement distribution. So farm bases won't grow twice as forest bases in practice. They'll probably grow about 25% faster and will have about 25% more minerals at the same time. I don't yet have math to support this but your statement that 2-2-1 forest kills 3-1-2 farm bends truth too much too as it neither yield is superior to other. It may feel this way for some but the difference is too subtle to be obvious.
Tim

Offline EmpathCrawler

Re: Population growth model and terrain rebalance/nerfs
« Reply #34 on: September 17, 2020, 02:19:39 AM »
You are right that my theory was one sided. No one is channeling all nutrient surplus into growth. It is distributed between growth and supporting poor nutrient mineral production tiles: forests, rocky mines. However, even in their mineral supporting application farms + mines are better than forests. You can review it yourself at different stages (TF, HF, enrichers, etc.) with different worker placement distribution. So farm bases won't grow twice as forest bases in practice. They'll probably grow about 25% faster and will have about 25% more minerals at the same time. I don't yet have math to support this but your statement that 2-2-1 forest kills 3-1-2 farm bends truth too much too as it neither yield is superior to other. It may feel this way for some but the difference is too subtle to be obvious.


lolada is having a similar conversation to one I recall us having before about tile yields. I think you are too focused on comparing raw tile yields outside of their context while ignoring a lot of quirks and perks from forests and their potential knock on effects to your economy. Every forest spread is former-turns saved, to use T-Hawk's term. I think that's why lolada and I agree based on our intuition. A size 7 base working 2-2-1 forests is going to produce at least 16 minerals and 9 energy per turn with potentially very few former-turns spent, won't starve, and will be pretty solid economically with a Genejack Factory and Energy Bank. Solid enough to do what they're doing in that screenshot: conquering fat bases from the AI to compensate for their bad position.


A lot more math has to be done here to make an informed determination.

Offline lolada

Re: Population growth model and terrain rebalance/nerfs
« Reply #35 on: September 17, 2020, 12:41:50 PM »


Ok, I will try to do some testing to see if i am playing wrong and how much exactly. Just to correct you - I said farms destroy 2-1-2 forests, not 3-1-2 those are worth keeping - but they are rare. 2-1-2 are common and they get destroyed. So you end up with lots of forests and few good 3-1-2 farms that you can build soil enrichers on.

Lets try to setup things a bit:

1. At very start obviously you want 2-1-2, 3-1-2 to grow and forest 1-2-1 are only ok if you can't grow more.. this is until you get sources of more food

2. Mines are used only when you can't grow and you HAVE to have good food to support them.. 4-1-2 or 2x 3-1-2.. (you actually removed 1f from rec tanks).
Mines are somewhat rare as well as good rainy tiles.. btw i rarely use 3-0-1 or 3-0-2, if i really can't grow differently.. i rather go for 2-1-2

3. Then Tree forests come.. its great facility.. +eco +psych can't be underestimated.. i would consider it building for larger bases even if i had no forests at all. When you count it eco dmg reduction.. its like the best facility in the game.

- here i value 2-2-1 tiles quite more than 2-1-2.. so i forest them. I do keep 3-1-2 or better so i can grow more later.
- i will also consider foresting 3-0-1, 3-0-2 tiles for 2-2-1 if i need minerals

4. Then late on come soil enrichers and i just use them on farms i have.. thats my usual routine.

---------
Now odds here are adv. terraforming. I do use Condensers when I have access to them and basically i turn 2-1-2 tiles into 3-1-2. Then i place Mirrors if there's good spot. Issue here is that bases without forests have crap minerals unless you have  mines around. From my XP and settling pattern, quite often there is not enough mines. So lately i am using more and more Mining platforms on the sea to compensate.
Btw - have to do this with mining platforms - because i think Boreholes are ruined - 0-2-4 is just not good enough tile. I'd rather have 2-2-1. So in newest WTP versions there are no boreholes in my land - if i see arid flat tile - its going to be forest. Then its hard to justify building any condensers when almost everything is forested. I suppose I could build condenser anyway - then recheck terrain and farm/solar again forested tiles.

So two problems here..
1. I often get condensers late, can we get them a bit earlier (compared to Tree forests)?
2. And there is lack of minerals on maps - maybe i should play on rocky planets, i will consider that. Boreholes used to fill mineral role, now they are not worth
it. How do we get minerals? Gotta go for forests.


Side note:
Also, sea mining platforms producing 2 minerals from the get go would be interesting. Food could be cut down to 2 to balance. Its change sea mechanics a bit.. but its likely for the better; sea bases are NOT fun.

To use mines - one need 4 food source, or two 3 food sources - that's hard to get. So it often even makes sense to level the terrain and just use forests instead of mines. The problem is once you level the terrain its forever. Eco damage is issue also once you get multipliers.. thats another go for forests instead of mines.

Once you get soil enricher thing change - so its easier to get more food to support mines, if you have any left. i didn't even consider Hybrid forests as i find them too expensive in general, they are very late game.

I am circling here.. to sum up:

I feel like there's little space/time for condensers and forests take over. Thus mines are used less. And boreholes are now ruined so mineral issue is worse - benefiting forest spam. I suppose i could improve my terraforming by going with 3rd pass, checking over terrain and farming forests later once i get Condensers. I don't do this because everything is forest so i don't see base terrain - i guess i can use PRACX. Then maybe another check on Soil enrichers.

Am i going totally wrong somewhere here? I don't know at one point in midgame its all too much micro and i just don't want to search for which forests to remove and where exactly to place Condenser when it randomly? change tiles around.

Re: Population growth model and terrain rebalance/nerfs
« Reply #36 on: September 17, 2020, 02:54:35 PM »
2. Mines are used only when you can't grow and you HAVE to have good food to support them.. 4-1-2 or 2x 3-1-2.. (you actually removed 1f from rec tanks).

Temporary experimentation. I've returned it back.

Mines are somewhat rare as well as good rainy tiles.

The frequency of appearance would depend on map parameters but by default there should be about 1/3 rocky tiles and about 1/3 rainy tiles. Rocky tiles should be also, more or less, evenly distributed but rainy tiles, though, may be clustered and not evenly distributed.

1. I often get condensers late, can we get them a bit earlier (compared to Tree forests)?

That may vary depending on version but latest WTP uses the same tech sequence as vanilla: Ecological Engineering (condensers) -> Environmental Economics (TF). So you should get condensers before. In other version that order may be swapped but they are still in relatively same time bucket and no more than one tech level difference.

2. And there is lack of minerals on maps - maybe i should play on rocky planets, i will consider that. Boreholes used to fill mineral role, now they are not worth
it. How do we get minerals? Gotta go for forests.

You mean rocky tiles? Personally I never experienced lack of rocky tiles. Sometimes, some coastal bases could happen to get 0-1 rocky tiles due to limited land coverage but it is like 1 in 20 bases. On average they get like 3-5 rocky tiles. I wouldn't say bases systematically lack them.

To use mines - one need 4 food source, or two 3 food sources - that's hard to get. So it often even makes sense to level the terrain and just use forests instead of mines. The problem is once you level the terrain its forever.

Leveling is irreversible but do you actually need to level for forests? Isn't it enough non rocky tiles around for it?

I feel like there's little space/time for condensers and forests take over. Thus mines are used less. And boreholes are now ruined so mineral issue is worse - benefiting forest spam.

Yea. Balancing boreholes is a pain. We can return them back to 4+ minerals and let them obsolete mines or we can give them more energy like 6 or 8 if you believe it'll increase their value. I am not fixated on either.

Am i going totally wrong somewhere here? I don't know at one point in midgame its all too much micro and i just don't want to search for which forests to remove and where exactly to place Condenser when it randomly? change tiles around.

Let me reiterate it once more. Forest is a very good improvement. It's yield is so close by value to conventional terraforming and with some other benefits there is no surprise people actually prefer it. That is completely fine if you want to use it all around. My calculations are not to avert everybody from it. I was to show that even if forest is slightly below in output comparing to best contemporary conventional improvement it is still quite often a very good choice - meaning it is perfectly balanced as is. There is no need to modify it! Except maybe terraforming time which I rightfully doubled and maybe it even need to be tripled but this would be disputable.

If you still want to compare yields to yourself to see if you are using right strategy then I'd suggest to compare average minerals-energy output for conventional and forest bases assuming all nutrient excess is going to support miners. You would have following more or less confined periods:

basic:
2-0.5-2 vs. 1-2-1

condensers/echelon and TF:
3-0.5-3/4 + mines/boreholes vs. 2-2-1

enrichers and HF:
4-0.5-4/5 + mines/boreholes vs. 3-2-2

Tim

Offline Nevill

Re: Population growth model and terrain rebalance/nerfs
« Reply #37 on: September 17, 2020, 04:18:08 PM »
The value of tiles and terraforming is connected to the stage of the game when they are made available.

I found that it works decently well with vanilla improvements, if you adjust a few values and timings.
With forest build times extended, boreholes at 0-2-6, and condenser giving its 50% improved yield there is, in fact, a choice between mass foresting and condensers, mines and boreholes.

Forests fill the niche of a poor man's mine now, yes... that, and energy production. But the latter is subpar.

I place condensers at Gene Splicing, which is a Tier3 tech, and Tree Farms at Planetary Economics, which is Tier6 (using a SMAC tech tree). This allows the player to place a couple strategic condensers early in the game to improve areas particularly affected by draught, but they won't have the former turns to do this on a mass scale and obsolete differences between starting conditions. That comes later, and by the time Tree Farms are available one has a nice array of options to choose from. Forests spreading is nice to have, but it's uncontrollable, and with no Tree Farms they are inferior to better tiles... so my solution is to delay tree farms until former turns are no longer an issue.

Quote
I can't choose to work 6 farm+solars or  3 forests.. i end up with 6 pops for example.. cant get more due to drones and can't grow so i have no use for food.
Quote
To use mines - one need 4 food source, or two 3 food sources - that's hard to get. So it often even makes sense to level the terrain and just use forests instead of mines. The problem is once you level the terrain its forever. Eco damage is issue also once you get multipliers.. thats another go for forests instead of mines.
This is why I believe extra food from condensers is important, and wanted the 50% functionality restored. A lot depends on allowing bases create 4-food and later 6-food tiles on demand. It allows some freedom as to what to do with it, be it placing a worker on borehole/mine, or creating a specialist. Which also alleviates the problem of drone control.

This is why orbital facilities are in their place at midgame where the food production stalls and population growth runs into drone issues.

Something needs to be done about growth rates. I'll probably experiment more with reduced intakes/increased yeilds.
« Last Edit: September 17, 2020, 04:44:50 PM by Nevill »

Re: Population growth model and terrain rebalance/nerfs
« Reply #38 on: September 17, 2020, 05:04:15 PM »
That's one big topic. So I may not answer each and every question but these are few I can.

There are about 25 (!!!) different land improvements in this game if we count different yield modification like forest + TF, mining platform + Adv. EcoEng, sea facilities, different fungus yield stages, etc. It is not possible to balance them all perfectly and to everybody's satisfaction. I even consider vanilla setup where boreholes override mines satisfactory enough. Forest in this regard doesn't seem to be completely broken as nobody actually covers all their territory with it even with forest facilities. It is more of other way around. People build forest facilities if they already had forest on infertile land but they don't convert fertile conventional terraforming to forest just because facilities become available.

Regarding condenser nutrient bonus. This encourages placing condenser at every tile as it adds 2 nutrients on top of rainy-farm-enricher and if one place condenser everywhere all tiles will be rainy, obviously. That is madness. Besides favoring just a single specific tile (center) in its 3x3 affecting area disrupts their placement strategy forcing player to value this central tile effect more than the overall moisturizing. If anything it should add nutrient bonuses uniformly for whole area but this would be even bigger madness, apparently.

WTP latest nutrient model with farm adding 2 nutrients seems to solve this mid game stagnation problem. That also effectively reduce an "all forest" approach attractiveness even with TF and HF. Yet it still supports the value of mixed farms+forests approach. So 1) forest are still valuable as mineral-energy source, 2) that modification left forest yield intact which is good for forest lovers.
Tim

Offline Nevill

Re: Population growth model and terrain rebalance/nerfs
« Reply #39 on: September 17, 2020, 06:31:54 PM »
Quote
This encourages placing condenser at every tile as it adds 2 nutrients on top of rainy-farm-enricher and if one place condenser everywhere all tiles will be rainy, obviously. That is madness.
No, because:
a) you only get +1 Nutrient for the first half of the game. Hardly a lot.
b) a tile with condenser is otherwise unworkable. You can't place a mine or a solar on it, making it inferior to its rainy neighbours that can and will get additional enhancements (arguably this also applies to echelon mirrors too, and why they don't see a lot of use - you need to waste tiles)
c) you won't have the former-turns to place them at every tile
d) you need resourses other than food

What you said was theory, here is practice. A screenshot from an MP game with Tayta played at the beginning of the year.

I was playing Edenists, one of his custom factions that has +2 ECON but starts with -2 INDUSTRY. No lack of credits, but minerals are hard to come by. Even then I had a healthy balance between farms, condensers, solars and mines. I had to lean mine-heavy just to offset the penalties, so the left half of the map is pure condensers + mines, while the right half is more energy-oriented.

Re: Population growth model and terrain rebalance/nerfs
« Reply #40 on: September 17, 2020, 07:49:12 PM »
I'll expose configuration option for this. Below are just my thought those may be as well biased.

b) a tile with condenser is otherwise unworkable. You can't place a mine or a solar on it, making it inferior to its rainy neighbours that can and will get additional enhancements (arguably this also applies to echelon mirrors too, and why they don't see a lot of use - you need to waste tiles)

Placed right it generates up to 9 more food. It is a second best improvement besides 12 FOP vanilla borehole. It doesn't seem it needs extra boost. People still build it even if they can get only +4-5 nutrient by it. And 4 extra nutrients = 2 more mines to support.

Why does it matter that condenser tile is inferior to surrounding ones? Nobody forces you to work it. Work these surrounding tiles instead! 🤨
They benefit from both moisturizing and existing improvements! That makes inability to build an improvement on condenser really minor thing.

a) you only get +1 Nutrient for the first half of the game. Hardly a lot.

Yes, maybe not every tile because other resources are needed. However, you would not deny that it is exceptionally tempting to place it on nutrient resources disregarding its moisturizing use. I just don't find these two functionalities (moisturizing and multiplying) playing well together. Each dictates its own placement strategy confusing player.


If you are in dare need of plain nutrients - why don't add as much as you need to base tile and be done with it? Would benefit all bases equally both human and AI.
Tim

Offline lolada

Re: Population growth model and terrain rebalance/nerfs
« Reply #41 on: September 17, 2020, 08:55:34 PM »
Old condensers were quit obnoxious i really disliked it. You couldn't really choose not to spam them because AI spams them and every AI ends up with 15 pop bases like Lal in that screenshot above. Not building them means lagging badly behind. And then they cause ton of eco damage and AI gets ruined.. AI still has massive problems with eco damage and worms and fungus covering their land.

Similarly, Boreholes were so strong that it was dumb not to have at least 1 per base. Boreholes at 0/4/2 was interesting.. its a good mix of energy and minerals - I would place some of these on arid flat tiles. Its just too much eco damage in general to build them only for 0-2-4.. At 0-2-6 its more interesting for energy strats. Still, 0-4-2 looks like something you would really want for crap bases with no minerals - its hefty for eco dmg and former turns, but you get that "mine" on flat tile. Its not op because it can't be spammed and i am cool with limiting them on no slopes.

I don't think Condensers need buff - how about making them easier to build, in fewer former turns. You said Adv. Engineering comes before Tree forests - but i guess trading can turn everything around. I kind of don't jump on spamming Condensers like i jump on Tree forests - maybe its my playstyle.

Anyway, I like this state of terraforming more than old op Condensers, Boreholes so i would not revert those back... How about just maybe delaying Tree forest one tech in future ^^.

Re: Population growth model and terrain rebalance/nerfs
« Reply #42 on: September 17, 2020, 09:09:44 PM »
Boreholes are all defined in alphax.txt. So I don't even need to expose this parameter. Probably go with one of the below.
1. 0-2-6, 18 turns. Decent energy output and some minerals for not too much time.
2. 0-4-2. 24 turns. Slightly better to mines due to extra energy but about double work.

First choice is more desirable as it has clear focus. Second one still competes with mines and confuses both human and AI.
Tim

 

* User

Welcome, Guest. Please login or register.
Did you miss your activation email?


Login with username, password and session length

Select language:

* Community poll

SMAC v.4 SMAX v.2 (or previous versions)
-=-
18 (6%)
XP Compatibility patch
-=-
9 (3%)
Gog version for Windows
-=-
77 (29%)
Scient (unofficial) patch
-=-
29 (11%)
Kyrub's latest patch
-=-
14 (5%)
Yitzi's latest patch
-=-
85 (32%)
AC for Mac
-=-
2 (0%)
AC for Linux
-=-
5 (1%)
Gog version for Mac
-=-
10 (3%)
No patch
-=-
13 (4%)
Total Members Voted: 262
AC2 Wiki Logo

* Random quote

In the years since our arrival, we have foolishly disrupted so many of Planet's ecosystems that entire species may vanish without our ever having understood, or even known them. We must halt this plunder, and halt it immediately, for our own survival as a species depends on our ability to strike a balance on this world.
~Commissioner Pravin Lal 'Mind Worm, Mind Worm'

* Select your theme

*