SMAX - The Will to Power - mod

[bvanevery]

The self-destruct is for narrative purposes.  It lets you understand what happened with Recon Rover Rick.  It is not well thought out as a game mechanic, and obviously they didn't write AI for it.  That's generally what happens when designers, artists, and writers sit around thinking stuff up for games.  Nobody considers whether the AI programmer is ever going to get around to handling that, and whether it will make the AI play piss poorly compared to a human.

Bunkers are another instance where obviously the AI has no brains at all.  That's why they're not allowed in my mod.

[Nexii]

Overnerfing things with inane costing is worse than letting them be overpowered. It takes choice/strategy away in a much less enjoyable way.

[Alpha Centauri Bear]

This is a game of features. As, essentially, any other game. You throw more features in it and it becomes more fun. Supposedly. 😅
In this regard it is not politically correct to just drop something from the game without it being clearly unusable. Other than that it is all very subjective. Similar way we can drop probes or psi. Will it make it less strategical? Not likely. However, it'll make it less feature reach and maybe less fun for some.

There are a lot of stupid features for sure. We cannot waste our time on each and every one of them without strong reason. I never thought that either BP or COC are of some threat to anyone. They never changed game much. At the point they are available the row economical power decides the win.

However, if we are moving it much earlier that may make a difference. Therefore this discussion.

[bvanevery]

Overnerfing things with inane costing is worse than letting them be overpowered. It takes choice/strategy away in a much less enjoyable way.

I'm the game designer.  I'm not personally interested in your "enjoyment" when it means a Happiness progression cannot be enforced.  This is a game, with rules, and victory conditions.  Not a sandbox.  People use it as a sandbox, but I don't design it for sandbox.  Nobody needs design for sandboxing, they just need a construction set of gewgaws to put in their virtual world.  This is why Minecraft Alpha was a lousy game, yet plenty of people loved it anyways.

As the game designer, I do know better than you do.  I wouldn't make a career out of it if I believed otherwise.  Just as I can have better judgment than you do, about painting, or writing.  Wouldn't do those things if I believed otherwise, if I thought the job should be outsourced to you the player.  I'm very much opposed to "player driven" game experience.  The designer is supposed to design, anticipate, and enforce these things.  Failing to do so is abdication of game designer responsibility.

Notice my icon to the left.  The choices are not an accident.  I'm the designer.

[Alpha Centauri Bear]

Overnerfing things with inane costing is worse than letting them be overpowered. It takes choice/strategy away in a much less enjoyable way.

Agreed. That is my rule of thumb. When I like to fix some irregularity I try to fix it with minimal changes so it doesn't not flip this problem to opposite side.

I tend to compare ROI and break even point for different facilities. Break even point in 20-40 turns is an acceptable range. Anything outside of this range is imbalanced to either direction. This is not an exact science but just a way of analysis.

Example for Brood pit original cost of 80/2. It has two benefits: native units building and POLICE. Let's say we have 20 mineral production and producing mind worm in 4 turns.
Native units bonus: +1/4 for cost +1/8 for lifecycle = 3/8 of production cost = 20 * 3/8 = 7.5 minerals per turn. But only while building natives.
Police bonus: -2 drones  (which saves about 4 energy = 2 minerals).
Total profit while building natives = 7.5 + 2 - 1 (2 energy maintenance) = 8.5 minerals / turn.
Break even point = 80 / 8.5 = 9.5 turns. Insane return. However this is only when we are building natives.
At the same time when we do not build them we get: 80 / 1 = 80 turns. Pretty dull.
So if we are building it for just in case and do not build natives all the time but say 25% of the time then we get it on average: (10 * 1 + 80 * 3) / 4 = 60. Which does not seem very excessive.

Of course, this if very-very rough attempt on estimate. However, in lack of other reasoning this seem to be an acceptable ballpark estimate.

Conclusion: it is not OP unless you keep building natives here to the end of the game which is unlikely.
Actually same type of consideration is applicable to other morale boosting facilities as you don't crank units all the time.

Disclaimer. This is just a description of my train of thoughts. I don't pretend it is in any way an ultimate truth nor I want to persuade anyone.

[Nexii]

BP and COC are decent, not amazing. I have them both at 10 rows, 2 maint and that feels a bit on the pricy side. BP does have a significant cost reduction, the higher your minerals the quicker the payback - it's much like a Genejack or Robotics this way. But it's not very flexible. Default native life is kind of just fodder for empath/trance. So a lot depends what other modding exists there. Cost of native life, cost of empath/trance, even the base cost of infantry all play in. Plus both come at around the time stronger economic facilities become available, like Tree Farm, Genejack, Fusion Lab. A BP or COC doesn't help you grow faster so in a sense military facilities have to have faster payback (due to their limited scope) than economic ones.

[Alpha Centauri Bear]

@lolada

As Nexii pointed out BP is exceptionally good to produce natives. However, natives are just subset of military which is a subset of production in general. So it has very limited applicability. It's +2 POLICE is also a nice addition that allows you to somehow easier maneuver on SE table. Yet, it should be built in each base for that benefit. So I doubt it is that overpowered even for original 80/2. I agree it is pretty useless to increase maintenance as you would need to dismantle it if you stop building worms. I propose something like 120/2.

As for moving it earlier I have following available options.
Centauri Genetics (L10) - current assignment
Secrets of Alpha Centauri (L9) - a little bit overloaded already but still possible
Bio-Engineering (L9)
Advanced Subatomic Theory (L8) - doesn't align with native techs but is pretty empty now
Eudaimonia (L7) - also almost empty
Gene Splicing (L6)
Centauri Meditation (L6) - nice match but overloaded already
Retroviral Engineering (L6) - could be thought as a more or less match

[bvanevery]

It has two benefits: native units building and POLICE.

3.  Lifecycle bonus, cheaper native units, and POLICE.  It is also a late game facility coming after all sorts of other facilities for the same purposes have already been granted.  There's no point offering late game facilities that are cheap, it would just be "here's your new stuff" for having merely learned the tech.  The economy of the game is already overheated by the time it comes into play.  You don't tell rich people they only have to pay $5 for something.

There's no point having 4 levels of factory in this game either.  Yet it does.  In this respect, the game is poorly designed.

[dino]

I dislike a bit the concept of making everything super expensive to balance, makes early game super slow and things are still OP later. Better to nerf OP stuff a bit instead, imo.
So, below some alternative solutions, some may be too difficult to ever implement, but who knows... I'm just an idea man you know ( everybody hates those ).

Satellites:
Max bonus without Aero Complex 1/4 base size and 1/2 with Aero, rounded down, balances both satellites and Cludbase Academy.

Advanced terraforming:
1) Condensers only increase raininess, no 50% food bonus, already implemented in Ytzii's patch and 4,4 Boreholes,
2) 1,1,2 forests -  wood is not a useful material in s-f setting, but can be burned for energy ( or you could nerf it even further to 1,1,1 but it'd be too much imo ).
3) Remove a mining platform bonus from EcoEng and give it as aquatic faction bonus from the start instead of the current OP one, it'd balance both sea minerals in general and aquatic factions.

Crawlers:
1) Gather resourcess with 1 point penalty, already in Ytzii's patch.
2) No disbanding for 100% minerals to rush SP, but make one crawler transport 4 resource points between bases  ( or 6 ).

With these nerfs beelining toward BUILD techs from the beginning would not be absolutely essential anymore and a factions that beelined military techs instead would remain a significant threat for a long period of time. Would make for a much more interesting and varied games. It would also significantly slow down late game.

Probe teams:
1) Add energy cost to tech stealing ( like half the cost to research +/- probe modifiers ), hijack base mind control function and dialogue window for that.
2) Give tech stealing/mindcontrol bases cost adjustable multipliers.
3) Do not grant a free prototype from stealing a unit, unless stealing faction has the required tech.

Natives:
Make each additional lifecycle granted by facilities ( biology lab and brood pit ) cost one additional mineral row for worms, spore launchers and sealurks and 2 additional rows for locust and isles.
Add 1 more row for each two "free" lifecycle bonuses granted by SP, SE, or faction bonuses for spore, sealurks and 1 row for every "free" lifecycle for isles and locust.

[lolada]

Brood Pits and Cover Ops Centers

I think cost should be no more than 3-4 turns in that stage for a good base - otherwise it won't be built. Its ok to build Command center 6-7 turn early on.. but on turn 200 there's no way i am using 6-7 turns in many bases building those facilities and still having to wait to build worms afterward. I can spam ton of worms instead from every base in those 12 turns and just finish the job. Police rating is not that valuable this late in the game - especially when you need to build police units as well.

So, that said, good bases in my last game at a time i got brood pits had around 50 minerals - i think i had one at 68 it was eco-damage cap at time.. many other weaker ones were 20 - 30. If you price BP between 100-150 minerals then it will get between 3-6 turns thats roughly ok - very best bases might get in in 2, weaker don't have to build it. You can probably go with 120 minerals if you find that ok, and no more than 3 maintenance. Its excessive to go over. Bvanevery wanted to move the facility out of use - i kind of don't like that argument, I'd like to see it used.

@tnevolin Maybe try putting it in Secrets of Alpha Centauri (L9) for start. It would be an upgrade and won't risk much being op.

I wouldn't put CoC in same bracket with BP. I built two in that game just to have stronger probes to win combat. Two bases can spam ton of probes and they don't cost support - worms do and you can't have them on fungus all the time. There's also no point to have CoC in many bases. One can have more expensive CoC in both minerals and maintenance.

Overpricing

About overpricing in the mods - there's no much point to it - you are both probably hurting AI. For example i find Aerocomplexes overpriced at 120/3 (for the time they appear) and just don't build them - or build them in 2-3 best bases eventually. AI wastes ton of time building them in most bases. Then i try to snatch Cloudbase or if i don't get it i don't really care. Its not that hard to spam more Food satellites or planes instead. I think its actually benefical for AI to have them cheaper as they are good defensive facility for them. Same goes for Naval Yard as they work as Perimeter Defenses.

For example overpriced special projects. The only really good reason to overprice something is that if you allow it early - but don't want to have it built early - like Command Nexus. If you put The Command Nexus at 1000 minerals and its so good - than the player will get it if he really wants it. I can just spam Crawlers and get any expensive special project - yes you made me invest in Crawlers but i get the op thing. But poor AI will waste tons of time on it and lose progress - Bvanevery said AI uses Crawlers to rush - but in any case they are not nearly as effective as player. For example Cloudbase Academy is expensive - that just means its easier to get for a player .

Special projects i can't get are actually cheaper ones - that AI builds before me. Thats important. Or one could teach AI to use multiple supply crawlers to rush very expensive projects. In that sense, 50 min base building special projects shouldn't take more than 15-20 turns. I wouldn't price later special projects more than 1000 minerals unless they are very best. The later project appear the faster it should be built. If its broken - then ok - but move it anyway to tech 10+. Very late best bases top 100 minerals.. and you won't be playing much with such power - its game over.

[lolada]

I dislike a bit the concept of making everything super expensive to balance, makes early game super slow and things are still OP later. Better to nerf OP stuff a bit instead, imo.

I agree with this - but examples below are mostly not relevat to this mod ^^.

Max bonus without Aero Complex 1/4 base size rounded down and 1/2 rounded up with Aero, balances both satellites and Cludbase Academy.

No need for this in will to power mod. Satellites come late and bases are big already - its just more needless drag of spamming satelites. For vanilla it would be logical.

1) Condensers only increase raininess, no 50% food bonus, already implemented in Ytzii's patch and 4,4 Boreholes,

Condenser are quite unique in its working if used normally - on bonus resources. A base ends up having one good nutrient tile and it can grow more and faster. I've seen people abusing it making rows of condensers. T-Hawk really exploited it well in his games. Now i hate ruining feature because someone exploits it. I'd just leave it as it is or maybe try experimenting with 8 eco damage like borehole - that would stop the spam. It might screw up AI ?!
1 condenser, 1 borehole per base being the goal - more and ecology is in trouble. I find it hard to use or place 2 borehole per base. If people still want to abuse terraforming let them do it.

2) 1,1,2 forests -  wood is not a useful material in s-f setting, but can be burned for energy ( or you could nerf it even further to 1,1,1 but it'd be too much imo ).

Forests as 1-2-0 are fine early mineral source. With Tree farm its 2-2-0 thats something but not great. It really makes you want have that borehole for energy. It also makes mines more interesting; solar collectors are actually much more interesting when forests are not strong - gotta move that research somehow. Then choice between 3-1-2 tile or 2-2-0.. You can use one for growth and then switch to more minerals when you don't want growth or energy. Anyway, making it 1-1-1 tile is kind of crap its good for nothing.

Fungus is also alternative to forests - keeping some for energy (fungus become 2-1-2 midgame and then 2-1-3). It leads to some tradeoffs - do you want minerals or energy, its hard to get both. Forests in vanilla give both and are basically the ultimate improvement. I'd like to play more with these forests - i kind of randomed twice in psi/fungus factions.

What could be maybe done is make Tree Farms and Hybrid forests somewhat cheaper. Hybrid forests come a bit late so its not smart idea to forest everything in the mod.

3) Remove a mining platform bonus from EcoEng and give it as aquatic faction bonus from the start instead of the current OP one, it'd balance both sea minerals in general and aquatic factions.

- AI uses lots of sea bases it would be a shame to nerf minerals there.

Crawlers:
1) Gather resourcess with 1 point penalty, already in Ytzii's patch.
2) No disbanding for 100% minerals to rush SP, but make one crawler transport 4 resource points between bases  ( or 6 ).

No need for 1. Making Crawlers expensive solves this (they are 120 min). Noone will really build crawlers on anything except mines/condensers. Boreholes i consider best for base otherwise you lose 6 resources.
2) Yeah i would like this changed if possible - or if AI could learn to rush projects. That could be maybe obnoxious ^^. Anyway making projects too expensive is not good solution. I usually play with self-made rules.. like no crawler rushing. But when projects are that expensive there's no other way. If AI don't rush then its free taking for the player. Btw if projects are made  cheaper - then one could place 6-8 energy credits for rush on them to stop abusing that. AI can't rush buy secret projects apparently.

Probe teams:
1) Add energy cost to tech stealing ( like half the cost to research +/- probe modifiers ), hijack mind control function and dialogue window for that.
2) Give tech stealing/mindcontrol bases adjustable multipliers.
3) Do not grant a free prototype from stealing a unit, unless stealing faction has the required tech.

It would be interesting to nerf tech stealing - its op. At least AI is doing good work with probes, they defend and attack well. I would really nerf mind-probing bases its so obnoxious when you lose entire base + units in and around it.  AI could actually spam less attacking probes - they seem crazy with it.

Natives:
Make each additional lifecycle granted by facilities ( biology lab and brood pit ) cost one additional mineral row for worms, spore launchers and sealurks and 2 additional rows for locust and isles.

I don't like this one. Isle of the Deep are very expensive - i think they are 120 min, Locusts as well. Sealurks are offensive unit at 80 minerals - and thats expensive as well. Sea/coastal bases are not exactly mineral rich. They die to stupid thing like formers and foils and probes, especially in enemy territory. And they often die vs combat units as well - if they get hurt AI counterattacks and wipes them with sometimes much cheaper ships (due to reactors). I had Drones spamming tons of cheap 3-3-4 ruining expensive Sealurks - you just can't get numbers to win naval war. Its dumb to tisk Isle of the Deep in combat - they have one purpose here and thats transport in dangerous waters.

Locusts don't really get extra combat advantage and they are expensive. They are good due to their movement/felxibility but thats about it - mindoworms are better and cheaper attacking unit. Locusts also get somewhat countered by AAA and interceptors.. mindworms don't - AI don't really spam trance units that much. By this time you also have magtubes for movement issues. If i spam locusts thats for micro convenience - i am too lazy to transport worms.

[lolada]

@tnevolin

So that changes it:
The Citizens' Defense Force = 500 -> 400
The Maritime Control Center = 800 -> 600
The Cloudbase Academy = 1200 -> 1200
* Command Center cost/maint is 6/2.
* The Command Nexus cost is 40.
* The Citizens' Defense Force cost is 40.

I like these changes. I would consider nerfing Command centers to +1 morale. It makes SE choices stronger and goes with logic of making morale valuable in your mod. As for Naval Yards and Aerospace complexes they act as defensive facilites for AI. Player is not dumb enough to spam them - making them cheaper would help AI i think.

What are your thoughts on expensive projects - doesn't it make it just easier for player to get projects? That logic i mentioned in posts above - it gives more time to player to use supply crawlers.

What do you think about:

a) making projects cheap enough that first who gets it has best chance to build it. that would often mean AI gets it - not the player. projects are all over the place so presumably zakharov would not build every project. he's a wimp anyway. and i think AI can be told not to build more than n projects at a time?
b) disable supply crawlers project rushing completely (and make projects reasonably priced for building
c) is it possible to teach AI to plan/use multiple supply crawlers to rush secret project?
d) is it possible to teach AI to rush secret projects with energy credits?

The goal of all of these would be to level the playing field. The thing is that player can pick up any secret projects at will in late midgame. Making projects more expensive is counter productive - it just makes it certain that player will get it.

[dino]

> I agree with this - but examples below are mostly not relevat to this mod

It's very relevant since the primary tool it uses to deal with issues is to increase mineral cost. And exe modding is also very within tnevolin's capabilities.
Some of my "revolutionary" ideas became a staple of this mod for me, like no collateral damage for example.

It's also relevant to me, since I use mostly exe tweaks and increased armors from this mod and play something very close to vanilla otherwise.
I suspect that a very low popularity of this mod is caused by wild changes to vanilla tech tree and few other aspects and most fans would rather play more familliar SMAC but with more challenge.
Early game is awful with all these huge costs for everything and make people bounce, less brute force approach and more subtlety would make this excellent mod good imo.

As discussed in a thinker thread, nerfing advanced terraforming and crawlers would help AI compared to the human player who can alway abuse it better.
I want less minerals and other resources at play in late game which would slow down late game without issues in early game, like increased costs generate, or without the need to move interesting toys into a very late game tech tree levels. The ultimate goal is to make AI competent with the same rules as player, without heavy cheats, ideally to the point that it is competent at one level below transcend.

> Forests as 1-2-0 are fine early mineral source.

Mines on rolling are already fine early 1,2,0 mineral resource, but ridiculous 2 mineral forest is so good that it's absolutely pointless to build mines, especially on rolling.
It's a leftover from civ games, that didn't had boreholes and mining platforms, or that many rocky tiles for a good mine, to generate minerals.
It's secondary, but from a versimilitude perspective wood is not a viable production material in a s-f setting, unlike in medieval times, but energy rich plants can be burned for energy.
So I'd imagine forests here as mostly fast growing energetic plants plantations and I think 1,1,2 with 8 turns cost would be perfect.

- AI uses lots of sea bases it would be a shame to nerf minerals there.

Removing EcoEng platform bonus is suggested to balance proposed land advanced terraforming nerfs.

> No need for 1. Making Crawlers expensive solves this (they are 120 min).

No need for expensive crawlers with gathering penalty and 1 mineral forest and it would hurt human player way more than AI.
Super expensive crawlers can be abused by human sniping, or stealing them from AI's making them waste resources this way.
And most importantly, huge costs for basic things are a true blight of early game in this mod.

[lolada]

In these kind of the discussions goal needs to be formed otherwise we talk past each other. For example i am for .exe modding if modder can do it - sure it helps solve some big issues.

I suspect that a very low popularity of this mod is caused by wild changes to vanilla tech tree and few other aspects and most fans would rather play more familliar SMAC but with more challenge.
Early game is awful with all these huge costs for everything and make people bounce, less brute force approach and more subtlety would make this excellent mod good imo.

I don't think you are right about reasons for low popularity thing at all. My impression is that Alpha Centauri is kind of forgotten game with little audience. Even less people use mods and even less can install the mod thats not simple double click on .exe. Then even less want any kind of challenge or change to rules - its certainly hard to guess their wishes. Most players want it different a bit but not hard at all. Modders maybe know how many people use the mods - but i bet its very little. Each mod would get a fraction of players anyway.. there are several mods out there. And not like this mod is marketed around much - bvanevery for example did much better job at  it - and good for him - thats how its done. Still its a bit said that such a good game don't have more players - remake would be cool.

    

As discussed in a thinker thread, nerfing advanced terraforming and crawlers would help AI compared to the human player who can alway abuse it better.

Its not exactly that simple with Thinker AI I believe. The AI is so good with borehole, condenser spam that player has to use it -_-. AI also spams lots of forests. Induktio really did a good job there. AI also spams supply crawler a lot and uses them even better than player.
How? Well to match AI's fanaticism for this spam you have to spam as well and it gets tiresome - every game you need to spam boreholes, condensers and supply crawlers and every game is kind of similar and micro intensive and it gets tedious. That said not having to micro ton of crawler - its a huge plus imo. Because it really gets tedious to move them around to optimal tiles as bases grow. In my opionon, Crawlers that are rarely and strategically used are much better for the game than 30 mineral vanilla spam versions. Bvanevery sad that he also moved them very late in tree.
On topic of nerfing condensers, boreholes for example - its just a feature.. you have to rebalance everything. Now modders do this already, Tim made things more expensive. If you nerf adv. terraforming then you have to make everything cheaper. Its same thing, two different solutions.

Forest 2 minerals are ridiculous, it's better than mine on a rolling terrain and it's a leftover from civ games, that didn't had boreholes and mining platforms, or that many rocky tiles for a good mine, to generate minerals. I think 1,1,2 would be perfect.

For example i don't like this in this mod and here's the reasoning. Idea is we want terraforming to be used (all of it including mines and solar collector) and fungus to be used a bit as well. So thats starting point.

What goes wrong? In vanilla forests are so good with tree farms (2-2-1) that there's no point using Farm-Mine ever, and Farm-Solar is usually worse as mineral > energy. Once you get hybrid forest, you can remove all these for op 3-2-2 tiles. People like it - so there's vanilla for that. Its actually playable to just spam forest all and tech for tree farms and not bother with terraforming besides adv. terraforming.

Now forest need their place - you propose 1-1-2. But that competes out a lot of use of solar collectors. It also competes with fungus tiles. Fungus in the mod is 1-1-1 and 1-1-2 for quite some time. Then it gets to 2-1-2 its pretty sweet.. your forest would be 2-1-2 with Tree farms. Tree farms have other great bonus - can be upgraded and gives eco, psych and can remove eco damage. Its one solution ok - its not bad. One can use mines and boreholes for minerals.

If you have forest as 1-2-0 then they have obvious weakness; no energy and its not easily fixed. So you need to mix it up - and start to really like that farm + solar collector. Fungus is also really nice to use for energy. Later depending on terrain, if crap you may want to forest everything and have strong mineral, low energy forest base. It kinda leads to base diversity better than 1-1-2. Vanilla forests are op and kill diversity - they are 1 ultimate solution.

I want less minerals and other resources at play in late game which would slow down late game without issues in early game, like increased costs generate, or without the need to move interesting toys into a very late game tech tree levels.

I see this - its two things actually. One - tech is too fast and i agree - every freakin civ game had this issue. I play with tech stagnation and blind research for that reason - i want to play lategame more, its cool. So thats my solution. Modders did make tech way more expensive than vanilla - but its still snowball without tech stag. Games are plenty long with tech stag and you can try everything - thats my recommendation. Now if modders try to make every game like that - other people might hate it - many like faster games.
Other thing is - less minerals is crap - i wouldn't want that nerfed. I keep wanting more minerals in Will to power. First you can't get much it takes time and then you are restricted by ecology and population as well. Mineral boosting facilities come a bit late and are expensive to produce.
Costs also prohibit you from building those sweet 16-10-1 units and its such a shame. I really like that these are made cheaper in mods. In vanilla its totally dumb to build these units.. way better to buld 16-1-1 and 1-10-1 or just 1-1-1 and upgrade. Thats gamey. Anyway it needs a balance, too cheap and its way too much micro and war.. too expensive and its a drag.
On units side will to power might be slightly too expensive at times - but they can be rushed - later there is so much money available that you don't know what to do with it. As for secret projects - i don't like that they are so expensive - it doesn't make sense. I don't think that fulfills reasonable purpose.

[dino]

Less late game minerals, less builder micro, does it realy improve your game experience to have 20 - 30 units per base, as opposed to 10 - 15 ?
Also tons of minerals makes building facilities, a no brainer late game when you can build anything in 1-3 turns.
Balancing costs for mid-late game mineral surplus and stalling early game is the major issue for me with WtP.

1,1,2 forests are different from fungus, since they don't require planet rating and can be exploited while running free market.
And with 2 minerals forests, mines are completely pointless, you want an early  game 1,2,0 tile ? You have it already, a mine on rolling.
1,2,0 forest is still way superior to mine, with potential future upgrade, and borehole is superior to mine on rocky, why keep mines completely useless ?
Thinker with few lines edited in the source code can be easily adjusted to build less forests and more mines to facilliate changes.
For exmple flat non rainy - forest, otherwise either farm+mine, or farm+collectors.

Thinker also already generates way more nutrients than it can utilise, crawled 4,0,0 condensers won't hurt it and it should just be adjusted to build them only when most nearby tiles are non rainy.