T-hawk Plays Alpha Centauri

[Fluffball]

Thanks for the update and info, T-hawk. This continues to be both entertaining and information-rich.

Flying mind worms - ick! Have not encountered those yet. Can you attack them at all with non-flying units, or do you just have to defend as best you can? Artillery, maybe?

I was wondering about the amphibious pods; just what exactly do they allow a unit to do?

Did you give up on your playthough? SMAC loses a bit of its charm once you know how to absolutely obliterate the game on the highest level. It does gain a new charm in the "let's see how far we can break the game", but you'll never get back the original bliss of exploration. Unfortunately the same can be said of most games!

[T-hawk]

And I'll talk about buildings as Darrell asked for.

The critical point is not that buildings are bad per se, it's they always must be judged on the opportunity cost, competing against the other options of more formers and supply crawlers and colony pods.  And the formers and crawlers continue to win for a long time.  Early on, before boreholes and specialists, a decent base might be producing 8 energy of which 4 goes to labs so a network node adds 2.  A former can add that 2 energy with one solar collector... and do it over and over and over again.  An energy bank might add 2 energy, which buys one mineral... or a supply crawler can provide 2 minerals from forest.  And once the advanced terraforming options are enabled, formers to build boreholes and crawlers to work condensors still keep winning.  Or heck, probe units can even provide more; a single tech steal is hundreds of labs, and I stole more energy from the University than would have been yielded by spending those minerals on an energy bank.

Buildings are almost all multiplicative, so they don't do much unless there is some input to multiply.  Before that, you need the linear increase from formers and crawlers and more colony pods first.

The Children's Creche is usually the first building, the equivalent of the granary in all the Civs.  It doesn't have a huge effect if not qualifying for a pop-boom (20% growth is obviously less than Civ's 50% granaries.)  But a pop-boom is an incredible multiplier, changing the growth cost from 20+ food to a flat 2, effectively a multiplier of 10x or more.  The creche also gives an efficiency modifier for that base that does become pretty important; some of my outer bases are losing ~5 energy to inefficiency which would be twice that without the creche.

I'm not really sure about Recycling Tanks.  That's supposed to be your obvious first starter build.  But I always keep deciding that I'd rather have more formers making continuous improvements than the one shot of the tanks.  Despite the quote coming from Yang, the tanks are actually least useful to the Hive running Police State who can support the most formers.  Formers and pods always seem to be a better (faster payback) linear option and then other buildings are a better multiplicative option.  But I have to say you can't go really wrong by building it, particularly if you're a novice looking at a game horizon of 200+ turns rather than my expert's 100.

Good buildings are anything multiplicative once you've got a good baseline of 10+ input to multiply: network node, energy bank, research hospital, fusion lab, genejack factory.  That baseline comes after pop-booming to the early cap plus some boreholes.  Tree farms are good too although don't overrate the food if you can get that from nutrient bonuses and condensors and later orbital satellites.

OK buildings are situational: command center/naval yard for better military, biology lab/centauri preserve if you're building native life for conquest, one skunkworks if you're building a lot of prototypes instead of stealing them all like I am.

Some buildings are good in principle but bad for being overpriced.  This is the Hybrid Forest.  240 minerals is too much.  It feels good to get 3 food from every forest, but realistically that shouldn't help, by the time you finish that enormous building you can also build enough condensors for food instead.  A skunkworks is also often overpriced for what it does; building a prototype is just as easy by cashing in a supply crawler.

The hab complex is the most badly overrated building.  The size 7 limit does not mean "build a hab complex now", it means "do something else besides grow."  It adds zero productivity.  If you want to put 80 minerals into more population, build a colony pod and two formers instead.  What is that 8th population going to do that was worth 80 minerals to enable him?  The only correct answer is work as a specialist in your super science city with the labs-doubling SPs.  Never build a hab complex before you run out of room for new bases.  Even then, you require the food and drone control and means to pop-boom up to the new cap.  I will do that eventually here, but it requires all of orbital satellite food, Thinker specialists (+1 psych +3 labs) for the drones, and the Cloning Vats to boom since GA-booming will be impossible.  (All of this is possible by other means: hybrid forests for the food, psych slider for drone control, slow non-boom growth if your faction can't SE-boom; but all of that is worse efficiency than what I said first.)

Some buildings are bad because you just don't need them.  Don't build a Perimeter Defense, build or upgrade units to counterattack instead.  Pressure Dome is crap, if you've really got runaway sea levels and the council won't pass a solar shade, terraform-raise your land to save an imperiled base.  Punishment Sphere obviously isn't worth the negative of -50% research, aside from very occasionally in Free Market to home drone-causing units.

That leaves drone-control buildings.  Rec Commons are fine if you're not the Hive and can't do it all with police units.  Hologram Theaters are also OK if you don't have them by the Virtual World.  But the thing about these is you want to build them everywhere at the same time, so that the effect of the psych slider stays constant across the empire.  It doesn't do good to have rec commons in half your bases if the other half are still consuming 30% psych to stay afloat.

[darrelljs]

I just had to download this from GOG thanks to your thread. I'll probably only play a game or two (Morgan first for sure).

Did you not build any Recycling Tanks?

Darrell

[Kofiman]

Great read, T-hawk!

Superdrones, (Or, as I called them, double-drones) are basically straight from Civs 1 and 2, modulo operator and all.

[TheHumanHydra]

Some buildings are bad because you just don't need them.

'Some vices miss what is right because they are deficient, others because they are excessive, in feelings or in actions ...'

Looking forward to the war with the University.

[T-hawk]

Seriously, play Morgan only if you intentionally want to handicap yourself and drive yourself crazy with micromanagement for ICS and pop-booming.  He is not a good faction, as LKendter learned, and not for a novice.  Anyone else running Free Market gives the same experience without the support and hab headaches, and can boom with Planned.  Do Morgan only if you really have your heart set on the Wealth+Green configuration.

I built two Recycling Tanks: one during the first round of pop-booming at the base where I complained about Lal's unit threatening terraformers, and one at my big science base once it built all the available multipliers.  These were the two cases where a linear increase couldn't be done with formers or crawlers, because the land was inaccessible or already full.  Every other case where I considered a tank just felt better to build a former or crawler instead.  Even colony pods are better: 1-1-1 for 40 from a tank is worse than 1-3-2 for 54 from a colony pod.  (That's the base square plus a worked forest for 30 minerals + 20 food + 4 turns of former support to build the forest.  Any tile bonus makes the colony pod way better.)

Yeah, I know Civs 1 and 2 have the same double-drones.  IIRC, Civ 1 has a bug where the luxury slider converts a double-unhappy citizen directly to happy with only 2 luxury units, so double-drones are actually good.

[haphazard1]

Thanks again for discussing and explaining all this stuff, T-hawk.

I have not given up on my play through, just been very busy. I am in the middle of preparing to move to a new city, plus have SG turns to play. But I will get back to it, hopefully in the next couple days. Of course, I have been building all the things T-hawk suggests never building. I am so used to the various limits on expansion from later Civ games that the ICS continuous expansion mindset is very hard to get back into properly. Still having a lot of fun with the game, even so.

[Modo]

T-hawk, I've read in many of your reports that the most fun part is managing different threads and seeing them come together at the right time.
In SMAC it seems it's always worth going linear (which usually equates horizontal) first then multiplivative (which usually equates vertical) and I haven't seen an example of the opposite in any other game; maybe aside from some isolated variant start or something.
In your experience what's the kind of game that strikes the best balance between the two? And I mean for a regular playthrough otherwise your Civ5 spaceship fast runs take the cake for sure

[darrelljs]

I play Morgan mostly for flavor, I just find the guy funny. I don't typically play Transcend unless University but I'm not a novice...

Darrell

[T-hawk]

Just to be clear, I wasn't saying not to build recycling tanks.  They're always at least like 80% of optimal payback which is fine for most purposes and you can't go wrong with it.  It's only my expertise that demands 100% perfect build orders and discards anything less.

Yeah, SMAC needs some kind of variant rule to do anything other than horizontal then vertical.  (And as I was saying to Fluffball, define your rules and properly call a variant a variant.)  The horizontal tools are available right from the start (colony pods, basic terraforming, one police unit) while the vertical tools need many techs to come together (more drone control, crawlers, booming, advanced terraforming).  The SE policies cause the sharp distinction, since your current configuration can be either for expansion or vertical growth but not both.  That distinction doesn't happen in Civ 4 where the civics aren't quite so impactful (nothing like pop-booming) or Civ 5 where civics are cumulative rather than switchable.  It does somewhat in Civ 6 which has impactful civics but permits frequent free switching.  SMAC also doesn't need any super cities for wonders/projects because of supply crawlers.  The Hive is actually the faction that leans most to vertical over horizontal, since the b-drones cut off Golden Age pop-booming.  I paused at 18 bases here and any more can't boom, but anyone else would have been happy with 30+.

The key to balancing linear vs multiplicative is a short time horizon.  That creates the necessary tension during which a linear increase may give more results before the multiplicative engine overtakes it.  That's what went right with the Civ 5 speed runs, that the game was so short that cities founded later than turn 50 wouldn't amount to positive value before the end of the game.

The kind of game that best exhibits this is Euro-style board games.  I play a lot of those, with a few different meetup groups in real life.  The nature of a board game needs to fit within a few hours and the rules within manual execution.  That naturally results in that short time horizon of a few tens of turns rather than the hundreds of computer games.  Good Eurogames give that weaving-threads-together feeling, where all your resources come together to buy some important building one turn ahead of someone else, and that one turn represents a significant differential in the game's short time horizon.  I was basically playing Civ 5 like a Euro board game.