T-hawk Plays Alpha Centauri

[Kofiman]

I kinda think Yang's version of a Transcendence victory would involve killing the planet first, and then merging human minds together. Go eco-damage!

[Brian Shanahan]

If I knew what path would make the most heads asplode, I'd have chosen that. But I'm only an ugly bag of mostly water.

[Modo]

I've got no particular need to optimize for finish date here.

No particular need, get it?
And how come you didn't anticipate it, your particles are no different than ours right?

[T-hawk]

If you want to see the University war, then don't blink.

2184, before ending turn: set research to Advanced Military Algorithms.  Last time, I had to pick some tech after Planetary Economics, which was Centauri Empathy because everything else desirable fell into the missing-tech hole, but then the set of available options changed upon stealing Optical Computers.  Switching means giving up about 130 labs into Empathy (SMAC uses the pre-Civ 4 mechanic that progress accumulates towards one tech rather than separately on each), but that's alright, I don't need it.



2185: Leader's Horde causes another eco-damage fungal bloom.  It occured on the borehole, but that actually makes no difference - the borehole still yields the 0-6-6 productivity, overriding the fungus.  I have a rover to counterattack on land again, but I still can't hit a Locust on a sea square, so as before I clear out all my units adjacent to it so it will attack a base instead.  I expected it to attack Fecundity Tower, but it went for Hole of Aspiration instead which I then realized is the same distance away; anyway I won on defense same as before.  From here, I'll skip ordinary eco-damage reports unless something unusual happens.



Remember that loaded Spartan transport?  Here's the unit that was on it -- but I have a probe in place (the probe that returned after stealing Optical Computers) to bribe it!  206 credits is a lot, but it's totally worth it for early access to impact rovers without the tech!  And I can use the rover itself to kill the Spartan infantry that also invaded.



In other Sparta news, I again brought two probes with a transport to steal a tech.  As before, I landed the two of them next to a base that only had one unit on defense.  It counterattacked to kill one, but the other survived to steal, and got exactly the prize I wanted in High Energy Chemistry.  I had made sure this wouldn't screw up my beeline to Genejacks -- in fact I needed to acquire/steal/trade one tech somehow here, or else Bio-Engineering would fall into the missing-tech hole on my next step.



Finally I remembered the right way to use probes from boats.  Unlike later Civ games, unloading from a boat consumes only one movement point rather than all of it.  A 2-move probe can walk off the boat and steal right away.  My earlier probes had been on the 1-move infantry chassis, because that was cheaper to build and I thought the second move wouldn't be relevant coming off the boat.

This screenshot didn't actually occur as shown, I took it just to illustrate the principle, then reloaded so that my probe didn't actually land and no steal happened.  Because there was a defending probe in the base.  Even if mine won the probe combat, it would just get counterattacked and killed.



Back home, I've mustered units to capture that lingering Peacekeeper base.  Two 2-attack units will make sure to kill the one 2-armor defender, then my other rover can capture the base (in SMAC a victorious attacker does not automatically advance into the defender's square.)  Plus an artillery rover fired one shot from the tile south of my stack to damage the defender.



As I move in, Lal opens communications and I make the mistake of accepting.  He offers the submissive pact that I mentioned long ago.  But I can't take it.  I don't want his 2 techs (Doctrine Mobility and Applied Physics), since they will raise my own tech costs and also perturb the missing-tech holes that my current beeline is calculated to avoid.  Rejecting a submissive pact means the faction goes to permanent vendetta and will never talk again, but that's okay.  Destroying him will be a service to myself, not to humanity, but that's all I care about.



Morgan also opens communications.  First he wants to trade techs by offering Doctrine Mobility or Applied Physics.  I say no of course.  Then I repay my old loan in full (50 credits left) which allows him to offer a new one.  128 energy paid back 2/turn, even considering that as 78 net compared to the old loan... that's 39 turns before going below break-even and I already anticipate the game won't last that long.



Here's a happening that merits some note.  A volcano erupted from undersea.  It has no gameplay relevance to me; there's no faction near it and I don't expect to ever bother settling overseas.  But what's interesting is the backstory behind this event.  I can't find a citation but this is what I remember from threads on Apolyton back in the day.

There is a random event that can happen even if the option for random events is turned off.  It is a meteor strike on one of your bases, destroying it entirely.  I know this event happened regarding my base Leader's Horde, because this popup occurred during production upkeep, right after that base produced some item (and caused eco-damage again) but before zooming to the base screen.  There are a number of conditions on the meteor strike: it only happens if you are first in the dominance ranking, it does not happen if there are any sea tiles within a certain range of the selected base (because the event wants to leave a crater on land identical to Garland Crater), and I think there's also a flat random chance of the meteor not happening which I think is 80%.  If the meteor strike fails to occur from any of those conditions, the volcano occurs instead.  Bullet dodged, although it's a gun with many failure points and only one bullet.  There also exists a separate volcano event with identical results that is enabled by turning on the option for random events, and some reports conflate these two possibilities; the latter just happens as a volcano without any of the meteor strike threat.

2816: Advanced Military Algorithms discovered.

2187: Neural Grafting discovered.  That was two techs on consecutive turns, meaning the overall speed is under 2 turns per tech.

2188: Here goes the war with the University.  Don't blink.



The AI had been correctly amassing all its forces in Zoloto-Gold, as the tactical chokepoint facing my territory.  It had had 8 or 9 units for quite some time, enough that breaking through would take quite a number of impact rovers.  At the moment, I have three impact rovers staged where the defensive rover is (southeast of Zoloto), but that's not enough to drill through four 2-armor units behind a perimeter defense.

However, note that drone-riot icon on Zoloto.  A drone riot halves the cost for enemy probe mind control.  So does the University's inherent probe weakness.



Bam.  540 credits and it's mine.



It comes with all the units inside, and even the adjacent ones as well.  Including that 5-1-2 Gatling Rover.  Yet another stolen weapon prototype!  And the units come with full movement available to use this turn.

I immediately sent that gatling rover to attack University Base, two squares away and connected by road.  It knocked off the best defender there.  I then had my (4)-1-2 artillery rover shoot down the defending (2)-1-2 in an artillery duel.  Then I had two (2)-1-2 artillery rovers (one of my own, one just mind-controlled from Zoloto) fire at the last defender.  Then I started sending laser infantry units from Zoloto to attack that last defender.  They had a disadvantage to start with and a hasty penalty for traversing the one square of road, but I didn't care; these units were expendable meat hunks; I'd never use them for anything else, and SMAC has no war weariness penalty for losing them.



Two laser infantry died but then the third killed the last defender and my defensive rover from the staging area captured University Base.

Next, all still on this same turn:



I had one more probe in the area, one that had promoted to elite from several energy steals and had 3 movement.  It could reach the next University base.  This is the one that was building all those secret projects: Virtual World, Command Nexus, Merchant Exchange.  But I had conquered the University's HQ.  The cost to mind control is based on the distance from a faction's HQ, but if the faction has none (you don't get a free replacement Palace like in the later Civ games), it counts as a default large number for the distance.  That means a drastic drop in the cost to mind control, here now 385 credits, and that's without the drone-riot halving.

Mir Lab also came with a gatling rover.  But unfortunately it's an artillery model, which can only fire on the lone defender in the next base, not kill it.  If that had been a normal rover, then I would have conquered the entire University in one turn.



I finally brought up my own impact rovers, and the next turn did finish that off.  This didn't eliminate the University; they still have one more base far away north of Gaia.  Zakharov now contacts me offering a truce.  I take it because otherwise I expect he'll offer submission and give me the techs I don't want.

To afford that 385 credits for that second mind control, I had to scramble for money.  I sold Children's Creches in the closest bases to my HQ, since they weren't doing anything at the moment.  (No Creche will ever matter for growth again, since all future growth will be Cloning Vats booming.  But the efficiency modifier still matters for anything at any distance from my HQ, plus the morale against worms.)  I also scrapped the Tree Farm in Leader's Horde, as the food wasn't needed for the moment and the base could rebuild it quickly.  SMAC pays out 1 credit per 2 minerals on scrapping a facility, not the 1:4 of later Civs, so this raised the ~200 money that I needed.



I had also visited Deirdre to see if I could extract a loan from her; no go, but in the conversation while I was there she did trade me Nonlinear Mathematics!  Not that I need it for impact weapons now, but it's a prerequisite for later.

A year later, then Deirdre turned around and demanded I declare vendetta on my new trucemate Zakharov.  When I didn't, then she declared vendetta on me.  Never any peace around here.  (This is entirely normal AI behavior for SMAC.)



Wrapping up another thread, I built one impact rover in the northeast to take out Lal's last base there.  The "personal interrogation" isn't a game mechanic, you just get a video showing the victim screaming inside a punishment sphere.

2189: Back to domestic development.  Bio-Engineering discovered this turn.

This tech allows Clean Reactors, a special ability that eliminates mineral support for that unit.  All the guides tout this, but it's only moderately helpful in real terms.  It costs "2" in the unit workshop, which is not 2 mineral rows, it is +50% of the unit's cost before special abilities, rounded up.  A 30-mineral impact rover costs 50 with a clean reactor.  Don't pay 20 minerals to save 1/turn, spend the difference on a supply crawler instead for 30:2.  Particularly if the lifespan of the rover comes to less than 20 turns, then the minerals spent on the clean reactor were a waste by any reckoning.

What is worth a clean reactor is a unit that is cheap enough to cost only 2 mineral rows so the cleanliness only adds 1 row, and has a permanent expected lifespan rather than death in combat.  This means police and terraformers.  Building either of these clean costs 10 minerals to save 1/turn, which is most often worth doing.  And upgrading a police unit to clean costs 20 credits, which is also a 10-turn payback horizon.  Upgrading a former costs 30 plus the former's turn, so that's less worthwhile and I often don't bother.  Plus once the police are clean, most of the formers occupy free-support slots in their home bases.



Unfortunately, we're not rich enough to upgrade all the police the easy way.  More micromanagement: the right way to do this is scroll through all the bases and choose as priority for upgrades police units in bases paying the most support and where the saved mineral can make a turn's difference on the current build order.



The University may have crumpled quickly, but not so my next enemy, which is Planet.  Here come the sea levels and the raging native life.  But I'll be ready.



Here's how to harvest worms for maximum credits, which I really haven't seen described in any guides.  This exploits the confluence of several mechanics:
- A fungal bloom spawns a stack of worms on the affected tile.  The size of this stack increases with more incidents.
- A fungal bloom also spawns one mind worm on each adjacent unoccupied fungus tile.
- Killing a mind worm increases the lifecycle level (morale) of all adjacent worms, a "splash buff".
- Killing one mind worm in a stack kills the entire stack.
- The credits earned from attacking mind worms increases with morale, 10 per level.
- The 3:2 psi attacker advantage is usually enough to kill even high-morale worms.

Add these up and this is the right strategy: First nibble around the edges killing the single worms, which will increase the value of the main stack.  Then kill the main stack.  In this example, I jacked up the value of the main stack from 20 credits per worm to 40 before killing it.  I'll show more examples later on, when this gets really excessive.

2190: Retroviral Engineering discovered.  Next research is Synthetic Fossil Fuels, on the path to Biomachinery and the Cloning Vats.  454 labs/turn, still running at just better than two turns per tech.

[T-hawk]

Big huge overview, click to expand again. This was only six turns since the last one. Time accelerates precipitously in this game, you end up doing a LOT in a compressed time scale.

There's the Genejacks beginning construction. Each base will do one when it finishes its current build. I timed it in a few bases by slipping in a cheap build (another former or crawler) for one turn just before they would start on the Genejack. After that, it's hab complexes and tree farms for food, which should be timed right about along with the Cloning Vats to boom up to maximum size.

[Modo]

The "war" was really dirty, you should clean yourself after that. The rationale behind selling the chreches was slick tho, in a normal game what and how much do you actually sell by the end?
From what I'm seeing SMAC doesn't have a stat screen like Civ4 where you can see global numbers like units / buildings built, destroyed, current etc is that correct?

[TheHumanHydra]

Yeah, that was a pretty sweet turnset. SMAC war (and probe actions) never ceases to frighten and amaze me. Also, really interesting info on the mind worms.

[T-hawk]

I want to be clear that this is a normal game.  I'm merely using powerful mechanics without holding back.  Most often you scrap bad buildings like a perimeter defense when one comes in conquest.  Drone-control buildings can all go with the Telepathic Matrix.  A loophole: you can scrap a Hab Complex once you've reached maximum size since its lack only prevents growing not remaining at that size.  (Civ 3 fixed this, can't sell a building that lifts a hard cap when you're over it.)  Stuff like Command Centers and Skunkworks once you know you'll never use them again.  Besides that, I think all buildings remain useful enough to keep.

SMAC has a few stat screens here and there.  F7 shows you what units you currently own and have lost.  F3 lists your current buildings.  It doesn't have any stats relating to enemy units since the units aren't standardized to fit in a list - it would have to cover every slot in every enemy's workshop.

[T-hawk]

Short update again (five turns in two nights) before I'm away for the weekend.



A leftover detail from the University and Peacekeeper wars: Morgan was in those vendettas on my side and he left a bunch of units hanging around.  I'm trying to be at least slightly careful with defensive units to not lose a base to a sudden backstab.  One of Morgan's units was a probe team that stole a tech from me.  The game offers me an option to cancel our Pact in response, but I don't see any need.



Another leftover: I still had all those rovers in University Base.  And I noticed they were three road squares away from the Gaian base Velvetgrass Point.  So I easily captured it.  I hadn't intended to go aggressive against Gaia, but this base came at no cost at all, so why not.  On Gaia's turn, then she called me up to request peace, and I decided to just go with it, signing both truce and treaty.  I even then asked her to call off her fight with the University which she did, so we're all friends now.



Now that I'm at peace with everyone except Sparta, I make this move: call the Council for a Global Trade Pact.  It passes 3-2 without me having to ask for any votes.  Nice, that does indeed give me more income than conquering Gaia would have.

And now that I was done spending money on clean police upgrades, I could start rushing Genejacks to completion.  This is obviously excellent payoff by tempo advantage: often buying 10 minerals to complete the genejack a turn sooner also yields another 10 by it operating a turn sooner.



There's the first of the Genejacks.  Holy crap that's a lot of minerals -- but holy shit is that a lot of eco-damage!  So actually, most other bases order up a rover for worm duty before they start their genejack.  It's no real loss to do the rover before the genejack, since the genejack will produce so many minerals that it will end up overshooting the rover's cost anyway and waste overflow.



But that's how to turn eco-damage into riches that would make Morgan jealous.  Killing one worm kills the whole stack.  So as long as you have a rover in range, it doesn't matter how many worms stack up, more just means more profit.



But this one is a mess.  The fungus actually popped offshore, spawning two Locusts on the highlighted tile plus the one Isle of the Deep in addition to the two land worms.  I couldn't counterattack the offshore worms, so could only wait for them to attack the base, and lost a terraformer along the way.

These two shots demonstrate how to control eco-damage properly versus how it can run away from you.  The dangerous part isn't one big stack of worms, it's when separate worms spawn across many adjacent fungus tiles and you won't have enough counterattackers.  To keep this in hand, promptly remove the fungus each time it appears, so that each bloom is confined to one stack of worms.  Actually, the important part is merely occupying the tile with a former; that alone blocks worms from spawning and it's OK if the fungus removal goes slowly.

The other runaway part of eco-damage is rising sea levels.  I got a second warning (missed taking the screenshot) that sea levels would rise 266 meters in the next 20 years.  (I'm not completely sure but I think this is on top of the previous warning; each warning indicates a separate sea level rise with its own scale of time and magnitude.)  I'll cover more of how to deal with this soon.



2192: Synthetic Fossil Fuels is researched.  But for the first time I encounter this problem: the tech I want next (Doctrine Air Power) fell into the missing-tech hole at this moment.  With all the previous beelines, I was able to avoid that by arranging the sequence or timing a trade/steal at the right time, but this time it was impossible on a strictly linear sequence of prerequisites with no wiggle room.  I have to pick something else for the moment, which is Cyberethics, since I want the Knowledge civic and it's a prerequisite for the next goal after Biomachinery which will be Fusion Power.

And I also decide to take care of some secret projects that I hadn't yet: the Neural Amplifier (50% psi defense), Longevity Vaccine (drone control), and Planetary Transit System.  That last is worthwhile now.  It says all new bases start at size 3, but it also brings up all existing bases to size 3 as well.  I had held off slightly because bases at size 3 would risk starvation -- producing 6 food to sustain 3 laborers can't be done on just 1-food tiles; you need one additional food from one 2-food tile or a supply crawler or recycling tank.  But now I had at least one crawler at every base.

And now I get to show off doing secret projects by way of crawler upgrades.  This didn't happen for the earlier SPs (simply built with non-upgraded crawlers) both because I didn't have energy to spend and because I didn't have access to enough pieces to design a sufficiently expensive crawler model.  But now I do.



They start with this as the shell, a supply module on a rover chassis.  Then they upgrade to this:



A supply rover with 2 armor and the expensive Hypnotic Trance and Deep Radar special abilities just to increase the cost.  That's a price tag of 126 minerals, meaning 18 mineral rows.  The upgrade cost is 190: the formula is 10 per mineral row of the new unit plus 10 per difference in weapon and armor value.  That last is why I used 2-armor plus special abilities instead of 3-armor.  That 190 credits buys a difference of 84 minerals without SE flipping and 156 with it.



And now that I have access to the Power SE choice, I can flip my Industry SE rating all the way down to -1.  That means each row is 11 minerals.  Now that 18-row crawler is worth 198.  When I flip the industry back on, each of the 300-mineral projects costs 210.  So that one crawler plus each base's normal production will finish that secret project right away.  BTW, it's possible to upgrade a crawler then add to a project the same turn (even though the upgrade consumes its movement): do the upgrade within the base, then you can select it (even with no movement) and press O for the convoy order which brings up the crawler-arrived popup.  So in this way, I completed each of the Planetary Transit System, Neural Amplifier, and Longevity Vaccine in one turn at the cost of one crawler and 190 credits.

2194: I have 860 energy on hand, after killing several big stacks of eco-damage worms last turn and this.  Here's my plan now.



In every one of the newer bases, I changed the build order to a Genejack if I hadn't already, and rushed it.  It cost about 70 energy per base (the Genejack costs 70 minerals which is 140 energy, but the cities already had some accumulated plus the current turn's production.)  I did all this right now because this was the last turn I would spend in Wealth's industry discount.  Cyberethics finished at the start of year 2195, and I adopted Knowledge right away.  Wealth's industry had been great, and I know I'm giving up a multiplicative interaction between that and the Genejacks; but now the labs from Knowledge would have more effect to speed up what I'm doing.

During the wars, I hadn't really thought about the development of the new wave of founded and conquered bases.  I'd been sending formers and police units their way all along, but hadn't thought about it beyond that.  I set a Creche as each of their first build orders out of habit, but then rethought that after I mentioned how the Creches are irrelevant for growth because the Cloning Vats are coming.  I think the plan will go Genejack - Tree Farm - Hab Complex - and then the labs multipliers.

The terraforming plan for these needs to be the classic all-forest.  Just because that's the most economical plan for former time.  A borehole costs 5x the effort of a forest but doesn't give 5x the productivity.  Boreholes are what you want when former turns are plentiful and population is scarce, so you need to increase productivity per land/population.  But pop-booming to work five forest tiles is the same former investment as a borehole for more yield.  Genejacks will provide enough minerals to build the Tree Farms in good time, so the forests will also provide enough food rather than needing condensors.  This is the classic forest-and-forget approach to base development that everyone and their guides advocate.  It is effective not in terms of absolute or per-population productivity; it is efficient in terms of former-turns.  Which is what drives the situation for average players that didn't build enough formers (I'm now over 100 but that's still spread thin among 32 bases), and now for me as well since the game pace is accelerating faster than I even could build more formers.

I also started sending most of my supply crawlers to rehome to the new bases.  The established bases didn't need them now.  A constant 2 minerals/turn of course has more snowball effect in a fresh city than a mature one already making 30.  So I laboriously trekked every crawler over to the west.  Every turn of every crawler's progression, I micromanaged it to land on a forest or condensor whenever possible to convoy for that one turn, then reactivate and move it along next turn, until it got to a new base where it would rehome and stay.  Removing the crawlers from the original bases hurt my labs production slightly but noticeably, since the squares vacated by the crawlers were replaced by the librarian specialists going back to work.

I'm actually not sure how best to do this, how to balance bootstrapping new bases.  How much should mature bases build formers/crawlers to send to them as compared to their own facilities?  Should the new bases build more formers/crawlers before facilities too?  Is it correct to skip facilities altogether and keep expanding perpetually with more colony pods and formers/crawlers?  (With the Raise Terrain terraforming order, I can make more land however I want.  I've already done so at a few of the coastal bases.)  My speed-transcend game worked around that question by using a map tiny enough to literally fill the whole world with bases.  I don't really have to care about the optimal path in a game where I'm not benchmarking the finish date for anything, but I'm still curious as to where the calculus would end up if I were.  In general I seem to be building another one or two formers at each base after the genejack, which will plant a good amount of forests by right around when the tree farm completes, and then the hab complexes will also finish in line with the Cloning Vats.

On the formers themselves: I now have a few different models designed in the workshop: the base model, clean reactor, fungicidal, and both of those.  I select each build order depending on the mineral production of the base.  If it's got enough to finish a clean or fungicidal (or both) model in the same turn, I do that one, or else build a cheaper one if that's all the base can afford.  This is also applying to anti-worm rovers; I have different models there also with combinations of the Trained, Trance, and Clean abilities, which I select according to whatever that base can finish in one turn without wasting overflow.  It doesn't seem worth paying to upgrade formers to clean or fungicidal (30 credits each); what I'll do there is wait until the Super Formers special ability before upgrading.

Don't need an overview image; nothing changed since the last except the one captured Gaian base and everything's up to size 3 with more forest.

[Modo]

Saves? Pretty please?
Moving and using the crawlers along the way is such a fitting idea for their name and game theme, they don't make games quite like that (well, we do have Factorio).
I wonder why they thought about that worm stack rule, what's the point of the stack if you're able to kill it entirely by killing one of them? Maybe if you don't kill them they spread?