4X The Unsurpassed Brian Reynolds' Alpha Centauri thread

[MilesBeyond]

Big issue I have with SMAC is that even with Kyrub's excellent mod, and a variety of rules both hard and self-imposed (Ironman, Blind Research, no crawlers or planet busters or boreholes) it's still just too damn easy. I'm not exactly a master of TBS either. That being said, Kyrub's mod really did make a big difference. I remember when I first loaded it up and Operation lol needlejet was halted in its tracks by the AI responding with AA units. Still, though.

I don't know what the solution is, either :/


As for ICS... the reason ICS is feasible is because cities come into existence producing enough income to be self-sustaining. In other words, even with the penalties caused by overexpanding, 1 pop bases are still going to be a net gain to your empire, even without using crawlers. IMHO a Civ 4-style approach would be a lot better - each base is a net drain on the economy until it reaches a point where it becomes self-sustaining. A nice balance where massive empires are absolutely possible, you've just got to make sure you have the infrastructure to maintain them.

Other gripe I have about SMAC: Tech pace. By midgame or thereabouts, you're blazing through techs so fast that you'll hardly even have the chance to take advantage of half of what's in there. This is with Tech Stagnation, of course (does anyone play without that?).


To me, SMAC is the best 4X largely due to its atmosphere, writing, etc. The gameplay, while good, has a lot of niggling issues that slowly combine to diminish my enjoyment of it. There's just so much that's broken - and even without using any of it the AI still poses minimal threat.

 

[Average Manatee]

Did a bit of an analysis comparing two different builds, and it appears that you are mostly right. I compared a scrap of land 5 by 5 tiles with 4 fully upgraded bases and 9 specialist-based ones.

The following assumptions were made:
- There is a maximum amount of borehores (16) around the bases.
- Builder bases mine 4 borehole and 10 forests each. ICS bases all mine 2 boreholes each (even though there are only 16 boreholes, not 18), and have the rest of their pop turned into Engineers, mining the food through crawlers.
- Labs and Econ are split 40 by 60 percent (to imitate the Engineers and simplify the count).
- The builder style puts emphasis on Hybrid Forests and facilities enhancing mineral output. The ICS style omits those.
- There is a satellite network that the builder doesn't really need to support their bases. This slightly favors the ICS numbers when they aren't using crawlers.
- The mineral output is converted to energy at a 2-to-1 ratio. That's what you pay to hurry production if you need something built.

Spoiler

Overall, the Builder churns out 3 times more minerals than the ICSer does (might be handy for costly units such as locusts and/or rockets), while the ICSer earns 2 to 3 times more energy/research points, depending on whether the Builder uses Free Market or not. Still, the ICSer outperforms the Builder by 10 to 25 percents.

However, if the ICSer needs to feed themselves the old-fashioned way instead of collecting food through crawlers, their performance drops to a point where they are almost on par. The builder does not depend on crawlers at all.

All the more reason to abolish crawlers.

You're kind of making a wonky comparison here, of a late game builder building from scratch. What matter isn't the end result, but how quickly you get there. Also some mistakes: Specialists don't need rec commons, and they still benefit from tree farms/hybrid forests quite a lot. You also seem to be comparing energy per base, but omit the fact that the specialists build has over twice as many bases, and not really counting how much more terraforming time it takes up.

The correct way to think about it is: you need 30 former turns to make farms/condersors, and 60 energy (colony pod) + 90 energy (children's creche) + 180 energy (3x crawlers) to provide a return of about 25 energy (technicians/librarians) or 40 energy (engineers) per turn. You don't have to wait for tree farms to research and build or any of that stuff. It's about the immediate, quick payoff from a 2x2 square. And every bit of energy you get out of the base can either be used for building more bases or building facilities in your current base to enhance energy.

Going ICS with trees/boreholes, you really have so many humps. Your initial research is complete shit, barely helped by Weather Paradigm unlocking boreholes (you're completely fucked without it, but we'll assume you get it). Even so, you can only limp anemically to tree farms. In the meantime you can't do anything to boost your base efficiency (what, you're going to buy a +50% economy energy bank that costs 1 maintenance when your base only produces 2? Great joke). In the meantime the more ICS you do the more you hurt your efficiency, causing more need for population control, tech detours to policing, and potentially slowing you down. And for what? 2 forests + 1 borehole will give you 10 minerals and 10 energy. Even with our generous assumption that 10 minerals = 20 energy, this means your income is 30 energy (actually closer to 20-25 probably due to facility maintenance and efficiency). Compare that to the incredible speed at which the above works (we reach 25 energy income from almost the start of the game, with only a children's creche!), and it's really no contest.

Also, the 1 mineral = 2 energy assumption completely breaks down the instant your base runs out of things to invest minerals in. Which turns out happens very quickly when you ICS with forests (see aformentioned "build energy bank when you only have 2 energy"). When you ICS with forests 90% of your minerals quickly turn into deadweight that can only be used to spam out units to rush an opponent, unable to actually be invested in your bases. Specialists are clean, 100% energy reliant that can always find somewhere to spend money that provides a return and are infinitely expandable without efficiency issues.

All this said, it's probably true that the best strategy is foresting the first 2-4 bases and then going specialist. I won't pretend that literally building only farms right from the start is ideal.

 

[Nevill]

Also some mistakes: Specialists don't need rec commons, and they still benefit from tree farms/hybrid forests quite a lot.

I assumed two workers sitting on boreholes, so I had to install basic means of riot control. I have excluded tree farms to keep the amount of investments in infrastructure (mineral cost) roughly the same.

Your initial research is complete shit, barely helped by Weather Paradigm unlocking boreholes (you're completely fucked without it, but we'll assume you get it)

I am not sure why you need to unlock boreholes if you do not lift resourse limitations, first.

I am also not sure when you think it is a good idea to go full ICS. You need condensers prepared to place your crawlers on (which is a setup that does not seem to me like it is much faster to build than a forest+boreholes one) and enough money to mass-hurry Creches for it to take off and become self-supporting. Just wondering what your timelines are. Any saves I can look at to see this in action?

Regardless, I think your example just reinforces how vital crawlers and pop-booming are to ICSing, which is one of my original points. If they are more effective than I thought - well, so be it.

 

[Average Manatee]

I assumed two workers sitting on boreholes, so I had to install basic means of riot control. I have excluded tree farms to keep the amount of investments in infrastructure (mineral cost) roughly the same.

Yeah, I don't really use boreholes until after the expansion stage. I suppose its better for a 100% optimized endgame, but again growth is more important.

I am not sure why you need to unlock boreholes if you do not lift resourse limitations, first.

Ah yes, not needing to care about resource limitations. I forgot another reason specialists are better

I am also not sure when you think it is a good idea to go full ICS. You need condensers prepared to place your crawlers on (which does not seem to me like it is much faster to build than a forest+boreholes setup) and enough money to mass-hurry Creches for it to take off and become self-supporting. Just wondering what your timelines are. Any saves I can look at to see this in action?

Pretty much immediately, though you still have to adapt slightly to the terrain until you have a mound of crawlers, and singling out resources bonuses really speeds you up. The disadvantage of forests is that ICSing slows your research and you have to take a break to pop boom. With specialists you are always pop booming, and always a leader in research.

Regardless, I think your example just reinforces how vital crawlers and pop-booming are to ICSing, which is one of my original points. If they are more effective than I thought - well, so be it.

Thing is, I don't really think ICS is that OP, just the specialists are. Normal ICS with almost pure mineral-heavy starts really isn't that great unless you plan to impact rover rush with all those excess minerals. Ultimately crawlers are the only thing that dissuade ICSing. A 2 forest/1 borehole base doesn't need no crawlers, crawlers are what a 1 base covering 20 squares needs to keep up with the 5 ICS bases covering 20 squares. Given that for much of the game bases are sitting around 6-9 size, crawlers are the ONLY reason to leave more than 6-9 spaces for each base (and I would consider 6-9 spots per base still ICSing).

 

[Nevill]

Pretty much immediately, though you still have to adapt slightly to the terrain until you have a mound of crawlers, and singling out resources bonuses really speeds you up.

Any saves?

I play without crawlers for the most part, and in a vast majority of my games I sit at 2 pop (sometimes 3) at 90% of the bases until I am ready to PopBoom, which usually comes not too early, in 2180s or so. But I don't remember my games being very different even when I used crawlers.

I don't get how you are supposed to constantly pop-boom from the get-go. Specialist bases don't produce minerals. How do you build enough formers and crawlers to support the expansion? To me it seems like there needs to be a good industial support for this strategy before you can switch to full-time ICSing. I certainly can't imagine how to do it after your 4th base.

 

[Average Manatee]

Sorry, I didn't mean "start ICS pop booming from 4 bases". I meant "ICS with forests for minerals initially, around base 5 start putting down farms and crawling them". You'll slow grow your first few cities to 5 pop so that you can switch them to all-specialist and accelerate your research. You'll want to overload your cities as much as possible with like +10 food apiece, then once they are at your hab cap switch the excess food to another city. You'll be able to grow once every 2-4 turns or so. Once you have a few of those your tech rate will recover and you can gain the ability to pop boom.

You are aware of the little crawler exploit where you can tell it to crawl a resource even when its out of moves, yes? That's a major part of being able to be flexible with the crawlers.

 

[mastroego]

Well guys, you sure seem to know your SMAC.

BTW, harvesting just doesn't require movements points. No cheat imho :D

 

[Average Manatee]

So I tried a game out to show an example, and it completely fucked up. Come to find out, Yitzi's patch (which I use) actually implements a harsh nerf to this. http://alphacentauri2.info/wiki/New_alphax_variables_in_Yitzi's_patch

There is a "Drone rules code" setting (first working in version 2.1), allowing the rules controlling drones and talents to be tweaked:

A value of 32 means that Lal's bonus talents are multiplied by facilities in the same manner that Psych is, allowing these bonus talents to be powerful even once talents are fairly "cheap" due to multiplier facilities.
A value of 16 means that the University's bonus drones and conquest-based drones can create "superdrones", making the University significantly harder to play on higher difficulty levels (or when using a substantial number of specialists) and increasing the value of Police State when warmongering (or digesting recent conquests) in such cases.
A value of 8 implements a number of changes meant to limit drone control to the methods designed for it (making all factions harder on higher difficulty levels, though the Peacekeepers and Hive less than most): Firstly, if drones are turned into specialists with no talents in the base, other drones will be turned into "superdrones" to compensate. If all non-specialists are "superdrones" and there are supposed to be even more drones than that, it will make "phantom drones" that will not appear but will offset new talents before superdrones do; "phantom drones" can be removed by offsetting them with talents, or via police/facilities/projects if there are no drones (regular drones or superdrones) remaining. If drones are reduced by facilities, and then added again by pacifism, each "phantom drone" will combine with a pacifism drone to make a superdrone. Thirdly, superdrones count double, and "phantom drones" once, when counting drones to determine whether there will be a golden age or drone riots. (It will thus be possible to have drone riots even in an all-specialist base.) Finally (as the flip side to "phantom drones"), it will be possible to have "phantom talents", allowing golden ages even in bases where specialists outnumber non-specialists.
A value of 4 means that the non-lethal methods special ability and punishment spheres will not work at -1 or lower POLICE rating.
A value of 2 means that drone-suppressing facilities and projects are less effective at suppressing "superdrones" than regular drones (effectively taking one point of drone control to turn the superdrone into a regular drone before using another point to remove it), and will suppress "phantom drones" first. Police are unaffected by this setting, and will control superdrones just as well as normal drones.
A value of 1 means that drone-suppressing facilities can suppress pacifism drones.
To use more than one rule, the values for the desired rules can be added to get the desired value for the setting.

Apparently this was either on by default or I turned it on and forgot about it. Either way, it's a pretty good change. Though you can't see the "extra" drones, so it looks like you are getting drone riots in perfectly normal areas.

Anyway, once I turned that off I did this:

Forgive me for not going turbo-autist 2x2 cities, it's so soul draining to micro and I didn't have the heart for it a second time, so I went 3x3. This is me on the cusp of spamming satellites. I think 3x3 is actually a bit faster to get to this point since you run Hab Complexes earlier and get more use out of building multipliers even if you have slightly less base income, but you don't get the insane advantage that 2x2 does from satellite cheese, which upgrades them to hab complexes and the insanity that that pop density allows. Of course satellites are still useful for growing new cities faster and you can switch to 2x2 afterwards and stuff, but meh. Fuck the tedium.

At the moment I have about 12 bases producing all over 170 gross econ/labs apiece, for a total output of over 2k energy. I have 201 council votes, which is still beaten by Lal who landed in a monsoon forest

Well guys, you sure seem to know your SMAC.

BTW, harvesting just doesn't require movements points. No cheat imho :D

Yeah, not sure whether there's any consensus on whether its an exploit or not, but its a trick that lets you save massive amounts of resources by being able to migrate crawlers while still harvesting.

 

[Irenaeus II]

This thread is way out of my league. This information is esoteric to me.

I long for the days of simple Rambo fights in the jungle.

 

[Average Manatee]

You haven't seen reality-warping silliness until you've abused the "merge colony pod with base to add pop, even past hab limit" trick. If you think a 16-size base worth of engineers looks sick, imagine a 127-size.

 

[MilesBeyond]

Also, I like MilesBeyond of making a cIV-style expansion where bases are NOT self-sustaining at first. Hive is the closest thing to this early on because you don't get the base energy square. Can squares 'give' negative resources?

Hmmm, I really like the idea of squares giving negative resources. An imperfect solution, of course - but then, what isn't?

It's why I've always liked the idea of that SMAC mod for Civ IV. Never quite got into it, just because the atmosphere isn't quite the same, and while they added some things I liked, they also added some things I never really wanted (e.g. Civ IV-style resource system, which while is something I like in Civ IV, it's not something I've ever felt SMAC missed out on by not having).

 

[Average Manatee]

The problem is that encouraging bigger bases just leaves you with land covered 5% by bases and 95% by crawlers instead of 25% by bases and 75% by crawlers.

IMO, the best solution (probably best saved for a remake) would be to make ecodamage much more comprehensive. Make it so sharing eco damage between a 20 bases is no different than 1 base exploiting the same resources over the same region. Make it so that either only fungus is completely eco-friendly or trees are much weaker. Make it so planet attacking bases is actually suicidal past a point rather than being an infinite energy farm for your artillery.

This would make small, centralized buildups (commonly considered the "safe" way to play compared to conquering large tracts of land) actually much more unstable and prone to disruption than medium size builds. Which would lead to more competition for land, not because the land is strictly necessary to snowball, but because its necessary to snowball safely.

 

[MilesBeyond]

IMHO the best solution is to just remove crawlers entirely but I know there are many that would consider that heresy, combined with those ecosystem changes.

 

[Average Manatee]

Well since crawling resources also causes you to take eco damage, crawling would be quite limited if eco damage was properly revamped/.

It's just that right now its far too easy to split up eco damage across separate bases, and tree farms/hybrid forests completely eliminate terraforming eco damage for no apparent reason. Make it so food and energy extraction give half the eco damage of minerals, and make forests eco damaging (after all, it kills fungus!) and tree farms/hybrid forests only lower eco damage from forests.

 

[laclongquan]

Eco Damage is gold you know.

My Gaia and University run I set up one super-production city and it boom reliably. Each boom is worth as little as a year profit, and as much as ten, and they are easy to crush with the defender of that prod city. It's the early bonus money. The middle bonus money is using probe team to rob neighbours.

Morgan is harder to use that strategy because my defenders are scatter to the wind: defend against that hardass neighbour Lal, AND policing, exhaustedly, all cities. I dont even have enough to spare for a team of defender and formers (to repair the damage).

Although, a bit late but I figure out a cheap tactic to work: make a few worms and send them into the nearby fungus (to save cost). Then I will boom.
I got clean reactor but my production is running thin to make all other stuffs. Dont have spare cycles to make clean units at all.

I think it's this map. It make for Morgan getting a harder run at Transcend.

 

[Average Manatee]

Yeah, that's the other abusable part. If eco damage wasn't concentrated on a single city and worms started spawning everywhere in your territory fucking everything up, you'd be a lot more afraid. Or you'd just put Trance on every unit, not sure.

Consider this: The game takes all of the eco damage you would normally do (from worked tile improvements, minerals, and making energy/food cost some proportion of minerals) and then divides by the number of tiles you control on the map. This resulting value is then divided as eco damage between every base you control. Ecodamage reducers would only help in proportion to how many you build (i.e. if 80% of your bases have tree farms and hybrid forests, you'll get 80% of the effect).

Of course you'd need to test and play with a bunch of factors to get it balanced, but this seems like something that might stand a chance of being implemented by a very good .exe hacker.

 

[laclongquan]

As long as there are rules for fungal boom, we can find a way to get around and/or exploit it~

Say, if they are a wave instead of single point, we can just concentrate fast response teams in nodal city to react to cities in range. Two trance rover should be good enough. Or a few worms in fungal patches.

If you want to increase its strength, make it has a drone-increasing component: Making drones appear for 1 turn? Or influence your tile working result, like -1 to your nutrient/mineral/energy for one turn. That would fuck up your city's deployment.

 

[Average Manatee]

Remember worm spawns also screw tiles up, which is already better than just decreased output since you have to re-do the forming. Though its kind of silly since IIRC none of the actually expensive terraforming stuff gets destroyed, only the weaker stuff. Would probably be best to just do something like "any finished improvement the worm lands on is converted to an unimproved tile with the same improvement 5 turns away from finishing" so that regardless of where the worm spawned, you'd have to spend 5 former turns to fix it. Otherwise completely killing off boreholes and such would probably make them useless.

Also, maybe make it so that worms can move and attack during the turn they spawned? That could cause a lot of former/civilian attrition.

 

[Nevill]

Anyway, once I turned that off I did this:

Spoiler

Forgive me for not going turbo-autist 2x2 cities, it's so soul draining to micro and I didn't have the heart for it a second time, so I went 3x3. This is me on the cusp of spamming satellites. I think 3x3 is actually a bit faster to get to this point since you run Hab Complexes earlier and get more use out of building multipliers even if you have slightly less base income, but you don't get the insane advantage that 2x2 does from satellite cheese, which upgrades them to hab complexes and the insanity that that pop density allows. Of course satellites are still useful for growing new cities faster and you can switch to 2x2 afterwards and stuff, but meh. Fuck the tedium.

At the moment I have about 12 bases producing all over 170 gross econ/labs apiece, for a total output of over 2k energy. I have 201 council votes, which is still beaten by Lal who landed in a monsoon forest

Alright, so I did a bit of digging and found one of my old PBEMs from 2006. The game was played on a huge custom map, with three big flat arid continents with literally nothing else in them but fungus. All terraforming was done from the ground up. No pods to help us with it, either. And we chose three 'sucky' factions - Believers, Spartans and Cultists (SMAX) - because we were mazohists like that.

The game was played with custom rules - no crawlers, and I was limited to 40 bases to prevent ICSing (Santiago and Cha Dawn were limited to 50 as they were deemed more 'sucky'). An ideal setup to test the builder strategy in harsh starting conditions.

Here are some screenshots from 2240, which is not too far off from your 2225.

I had to popboom through Golden Ages since I did not dare to get off FM - I was too dependent on the income and research. I had a core of very developed bases and my outskirts were in the process of booming or preparing to boom. I was a bit late with my boom timings, but that's probably due to the map.

Here are my central bases (sorry for Pirate graphics - it's a SMAX bug):

Spoiler

Churning the Locusts of Chiron once per turn. I had about 18 of such bases with every or almost every facility built. Locust was my choice because it does not create pacifist drones on FM, has unlimited flying range, costs no support on fungus, and because it can easily swarm enemy bases from a direction they might not expect. Also, I was researching Will To Power and was preparing for the Dream Twister SP:

Spoiler

Miriam researches at a 0.8 normal rate, while Zakharov researches at 1.2, that is, 50% faster. Still, I wonder what was the research output in your game.

I was earning 2.5K in cash each turn - which went to hurrying facilities in low-pop bases so that they could popboom, too.

And since it was a FFA for 3 highly experienced players, the Secret Projects were split evenly, so no Ascetic Virtues or Cloning Vats for me.

Spoiler

Can you post your numbers or just attach the save so I could look at it?

 

[Average Manatee]

That's quite impressive. My tech investment per turn is about 1200, even with the turn difference I doubt I could substantially beat yours enough for food harvesting/specialists to be considered OP next to pure trees/boreholes, if I could beat it at all.

Don't knock Miriam though. +support is almost certainly the most powerful bonus in the early game. I'd gladly take university with +2 support instead of +2 research.

One interesting thing I noticed is that you make substantial energy from trade for free, which specialists don't benefit from. Up to +17 energy per city vs. +0-4 is a huge deal. Though I also wonder exactly how much having other builder players in MP supercharged your trade. The AI would probably give much less. Of course you also have to deal with fighting players and apparently worse terraforming conditions, so its hard to tell who had the advantage.

 

[laclongquan]

That locus tip is useful~ I will use it when the time come~

 

[Johannes]

What's the rationale for Needlejets/Choppers creating pacifism drones regardless of position, anyway?

 

[Nevill]

What's the rationale for Needlejets/Choppers creating pacifism drones regardless of position, anyway?

It is a mystery.

However, Ais Superiority units (interceptors) act normally (i.e. only create drones if they are away), and Locust has the Air Superiority special ability.

 

[Grotesque]

"Reynolds, whose credits include Age of Empires 3 and Sid Meier's Civilization II, talked about a couple of tricks Farmville uses to get players to part with their cash."

 

[Johannes]

"Reynolds, whose credits include Age of Empires 3 and Sid Meier's Civilization II, talked about a couple of tricks Farmville uses to get players to part with their cash."