Sid Meier's Alpha Centauri - Speed Transcendence


This report is rather old now and not up to my recent standards of writing quality and detail. It remains here as a somewhat well-known landmark in the SMAC community. There is a newer and better version and writeup here. Also see my SMAC home page for more writeups.

Here is a report of my fastest transcendence victory in Sid Meier's Alpha Centauri. It's done on the smallest possible map at 16x16 for the lowest tech costs. On Transcend difficulty; accept no less. Of course it will be with the University faction. I think the free tech was Centauri Ecology.

2101.gif - 107kb

I rolled quite a number of maps and starts, until finding this one that was almost all land, and provided some decent pod results at the start. My scout patrol moved south from University Base, popping a pod for a minerals resource, giving the city a nice 2-2-1 rainy minerals river tile.

The next pod popped an energy bonus, and I built the second base right on that tile. And it's in the jungle, for a 3-1-3 city square. That is important.

2103.gif - 106kb

Now here is the big huge problem with the 16x16 map. The bureaucracy limit is ONE BASE. Just owning TWO bases causes the first citizen in each to become an angry drone right away. My answer right now is a crazy 60% psych on the luxury slider. This would not be possible at all without that energy bonus (or a monolith or similar tile) at this base, and at least a river square at the HQ base.

2103b.gif - 76kb

Here's the research screen. Note that on this super-tiny map, techs cost FOUR BEAKERS to research. In my experience, this stays true for all of about the first five techs as the University. (The Uni gets some strange modifier where techs are half cost in the first few turns. This is true on other map sizes although not as noticeable.) I am "already" researching at one tech per turn.

2104.gif - 129kb

On turn four, my scout patrol meets the Hive. I sold him all three of my techs for money, taking advantage of a bug: you can sell techs for 25 credits each even if the AI doesn't *have* 25 credits. (My flimsy justification is that you can't actually tell when you invoke the bug, since you can't see how much cash the AI has unless you have infiltration.) I now used this money to rush the Formers at Uni Base, and to adopt Free Market -- on turn FOUR -- and lower the psych slider accordingly to 40%.

In 2105, researched Industrial Automation, and used the rest of Yang's money to adopt Wealth. 2106, my western scout patrol popped another pod for more cash, I think 75.

2107.gif - 140kb

In 2107, I make contact with Gaia, who would be the last AI on this planet. (These super-tiny maps mean that many of the AIs never spawn at all at game start. And when playing as University, the first AI to spawn is always Gaia, the second is Hive, and the third is Morgan.) Deirdre traded me her map, and also paid cash she probably didn't have for my techs -- note the 160 energy in the bottom right. I used that energy to rush a former at each base.

Gaia actually won the governor election for Planetary Council, by an amusing vote margin of 3 to 2 to 2. So I didn't get the infiltration and extra commerce now, a small loss. I would vote myself in as soon as possible, 20 turns later in 2127.

2109.gif - 182kb

Wow, there's a big break! The Hive just built a terraformer. I declared vendetta and killed it with my scout patrol. Now my unit is standing next to the EMPTY HIVE HEADQUARTERS. And it's even at SIZE TWO so it won't auto-raze!

2110.gif - 184kb

Uh oh, mind worms at Uni Base. My solution: bring the FORMER into the base to defend it. Mind worms attack with a -50% penalty before year 2115, so the former is enough to defend. (I really like this mechanic, and wish that domestic units like workers and missionaries had some small amount of defensive ability in the later Civ games.) Fortunately, it had already finished planting its forest on that minerals bonus.

The Unity Pod just north of The Hive would pop a river, which would be important. My scout patrol moved three tiles along that river...

2111.gif - 184kb

and found Yang's SECOND base ALSO at size two and undefended! I mercilessly stomped in and wiped Yang off the planet. Okay, four bases within 11 turns is a great start, don't you think?

You can also see another odd bug in that screenshot. New Arzamas got a Perimeter Defense out of nowhere. This sometimes happens when a base belonging to a faction that gets free facilities is captured. I would then sell all three Perimeter Defense facilities for another extra 75 credits total. (Civ 5 isn't the only game with magical free cash in the opening turns!)

Not much happened for the next few turns. The bases built formers, aggressively cash rushing them as soon as they crested 10 minerals to avoid the double cost penalty. These formers put down a forest or two to start expanding, then their primary task was to build roads to new base sites. I researched to and adopted Democracy.

Happiness was managed thanks to another bug/loophole. Even when you are far past the bureaucracy limit (with four bases, I'm now three times past it), the first citizen in a base is never a double-drone, only a regular drone. The psych slider even at 30% or 40% was enough to handle that. If a base grew to size 2, it would have a double-drone, and not even psych could keep up with that, so it would have to hire a doctor specialist. Cities did not stay at size 2 very long though, getting that colony pod rushed ASAP.

2117.gif - 161kb

(I found an artifact but a worm eated it. :( )

The next important step came on turn 17, when I adopted Police State/Planned. Yes, this is the crippling -4 Efficiency combination. That is -100% energy produced at all bases except the HQ, grinding my research to a standstill. But I have all the techs I need for the moment: all the prerequisites up to Ind Automation and the critical Intellectual Integrity. Nothing poops out colony pods like Police/Planned, and a simple 8-mineral Police Infantry unit keeps each base content at size 2 no matter how far past the bureaucracy drones limit.

Getting precisely to Intellectual Integrity right at this point had required some fairly intricate manipulation of the tech order, to work around the mechanic where 1/3 of the techs are not available for research each time you choose a new tech. I'll skip those details here because I don't remember them. Part of the path was to skip Biogenetics, which is skippable; Recycling Tanks have too slow a payback period; Colony Pods are cheaper and produce more.

2125.gif - 221kb

Eight turns later, with more tech-selling to Gaia for cash to rush colony pods, and I'm out to seven bases total by turn 25. I beat that Gaia colony pod by one turn to Lab Three. Uni Base is starting to put out supply crawlers to work forest.

With the new bases available to support units (I made sure to re-home formers ASAP), I switched out of Police State back to FRONTIER (default) politics. That gets me off the -4 Efficiency paralysis. But I can't take Democracy. Staying at 0 support is critical for the new bases to get their free 10 minerals and produce a Police Infantry right away to keep themselves happy. Frontier is the right move; those 10 minerals are more important than the +2 efficiency and growth. Jungle cities grow fast enough anyway, with Gene Splicing now arriving to get the full 3 food out of jungle tiles.

2135.gif - 235kb

10 more turns later and I'm up to 12 bases, still in Frontier/Planned/Wealth. There's also a boat out there collecting pods on the water, which turned up an artifact and some cash.

This shows the one point in this game where I really felt delayed as compared to an ideal possible game. I was never able to wipe out Gaia early with scouts and pod-popped units. Rather, I felt the need to get up to Nonlinear Mathematics and build the famous 4-attack Impact Rovers. I prebuilt the Impact Rovers by way of building supply crawlers to cash back into the prototype builds. Tip: you can cash multiple crawlers for multiple copies of the same prototyped item, if you do it all on the same turn while the item still shows prototype status before you actually complete one.

Deirdre actually declared war on me first, and probe mind controlled a base from me (Blackroot Palace, which she renamed.) But, SMAC predates the Civ rules change that attacking units can't use an enemy's roads. So impact rovers can move 6 squares and eviscerate an entire civilization almost instantly. I also added one Impact Battery artillery bombardment unit, giving me the advantage for a 4-1-2 unit to beat the 1-2-1 behind the +100% perimeter defense in Thorny Vineyard.

2139.gif - 249kb

Three rovers performed exactly as expected, exactly as they have since the beginning of time. Four turns later, Gaia was conquered down to one base and submissively enslaved. Deirdre paid me 25 credits for every one of my technologies again, bringing me back up to 200 credits. I shortly used those credits to upgrade crawlers for my first secret project, the Weather Paradigm.

So I'm at 15 bases on turn 39. And look at what they're building: Children's Creches. We know where this is going.

In the year 2144, Advanced Military Algorithms arrived, and now it's time to break the game. Deirdre and I repealed the UN charter, allowing me to nerve staple every single base. Now I don't need a single happy building or police unit EVER AGAIN. I would re-staple every 10 years like clockwork. Continuous nerve stapling is time-limited: you can always staple each base at least twice and usually a third time, but then they develop a resistance to it. Stapling is only going to last 30 turns... so let's win within those 30 turns!

(The stapling had a ramification I didn't realize: Nerve stapling a captured base counts as an atrocity against the faction you captured it from, even if the charter is repealed! Deirdre broke out of the submissive pact and went vendetta on me for the entire game! This didn't cost me much - just a few pennies from selling tech - and posed no danger with the leftover impact rovers keeping her base killed empty of units.)

2150.gif - 260kb

Six turns later, almost every base is maxed out at size 7. For the uninitiated, if you adopt Democracy/Planned governments, and build a Children's Creche facility in a base, it grows one population EVERY TURN as long as it has +2 food surplus. This is called the Population Boom, and is well-known as the most broken mechanic in the game. The AIs cannot effectively pop-boom; most of them will never run Demo+Planned because of conflict with their governmental agendas, and the few that can won't manage the food and happiness to do so.

Managing the pop boom took serious micromanagement. I laboriously went through every city every turn, bringing it to exactly +2 food. You can see that my formers have been building nothing but forests. A jungle forest provides the all-important self-sustaining 2 food, plus minerals and energy.

Let me also talk a little bit about build density. SMAC is well-known for encouraging ICS strategy, but I'm not going maximum full-blown here. I actually think it's more efficient to give the bases a little room to breathe, about 8 tiles each. The reason for this is buildings, primarily the Children's Creche but also any multiplier facilities. Each building costs the same amount but is worth twice as much in a city twice as big. This network of bases is already working every available land tile at size 7. If I had more bases, they would both need more Creches and not all be able to boom to size 7, for the same total population and output.

Mostly the key for ICS is plunking bases down FAST; they don't have to be dense. I've done that by pushing hard with formers for roads, and also by conquering four established bases. Relaxed ICS also means I'm not adhering to a strict spacing grid, but instead planted each base at the best available short-term site, working around fungus and rocky tiles. Only Morgan, with his higher base square output (8 energy at +4 SE economy) and hab limits, really needs to ICS at true maximum density.

The other reason max ICS doesn't work quite so well is terraformer efficiency. There are exactly two types of tile improvements that blow everything else away in terms of productivity per former turn: forests and boreholes. Forests go to 2-2-2 (with either jungle or tree farms; mostly my cities outside the jungle built tree farms right away) for a tiny investment of less than 4 former turns per tile thanks to auto-spread. Boreholes at 0-6-6 are just huge once you can afford the former turns to build them -- and the minerals beget more formers for more boreholes. Mines and solar panels are a total waste of time. Condensor + farm + soil enricher is nice at 6 food from any tile (9 in the jungle), but takes so much former time at 18 turns, and on the whole really doesn't match up a borehole. So at this point, my formers mostly let the forests expand on their own and shifted over to borehole production.

Anyway, once the pop boom was finished, I bounced back to Free Market of course. This exploded my research rate from 97 beakers in 2145 to 336 beakers in 2150. On this tiny map, that's more than a tech per turn already! My tech path collected up to Environmental Economics, Fusion Power, and Advanced Eco Engineering, and every city started building some Super Formers.

2158.gif - 269kb

Now I started focusing on building two super science center cities, choosing Zvedny Gorodok and Lomonosov Park. These two cities got rushed Fusion Labs, Research Hospitals, Hab Complexes, and Energy Banks. (This is base SMAC, not the expansion, so no Planetary Energy Grid project.)

The other cities started building supply crawlers. I still just had only one secret project at this point (Weather Paradigm), but now it was time for the Ascetic Virtues. This timed nicely with a swap back from Free Market to Planned for the second round of nerve stapling (can't staple with -5 police) and more turns of pop-booming. I built the Ascetic Virtues in a single turn with upgraded supply crawlers; the mechanics of that are well-described elsewhere. Just assume I've done that for all projects.

The two SSCs in particular were working several boreholes, putting out lots of eco-damage. I welcomed that. Bigtime. Eco-damage is a wonderful ticket to yet more piles of money. A few empath rovers could kill any boils of mind worms that spawned, raking in enormous payoffs of 50 or 100 or 300 gold at a time. Eco-damage is GOOD if you know what you're doing. Sea level rise is a concern, but that takes something like 10 to 20 eco-damage pops to trigger, then 10 to 20 turns to actually take effect. I'm going to win before that.

2165.gif - 278kb

Next came Orbital Spaceflight, and I poured some mind worm cash into bunging up three food satellites. I only needed a few, just enough to take all the cities from size 7 to 9. Growing beyond 9 was too expensive, needing the hab complexes. Also came the Cloning Vats, so now I could pop-boom while also staying in Free Market's economic power.

2165b.gif - 187kb

And here's a look at my optimized SSC. Two crawlers on 9-food condensor jungle squares, five boreholes, and the rest specialists. It's meticulously set up for +2 food to boom every turn and would continue doing so.

It is not really building the Virtual World. Instead, I am discovering Applied Relativity this turn, so that is a prebuild for the Supercollider secret project! Here's another look at the city five turns later. It's gone from 100 labs to 326 in just five turns! The other SSC would similarly add the Supercollider soon too. Also remember that's before the modifier of 40% for +4 SE research (University running Knowledge).

2170b.gif - 193kb

After acquiring everything to build the SSCs, the mid-game beeline target is Digital Sentience, enabling the Cybernetic Future Society which adds another 20% research across the board. That came in 2170, so let's see what my tech rate looks like now.

2170.gif - 193kb

Insane. Techs cost 750 labs and I'm producing 2000 per turn. Nearly three technologies per turn and still rising fast. Most of the bases never built a single facility besides the Children's Creche, and Tree Farms outside the jungle. They've just put out ever more formers and supply crawlers, except for the two SSCs. My cash and borehole availability have all gone into goosing the SSCs.

Why two separate SSCs? Don't you get more labs by multiplying together the Supercollider and Theory of Everything? Not on a super-tiny map. Truncation throws up a brick wall. There is no overflow on technology research in SMAC, so the best that one city can ever do is one tech per turn. Any labs beyond that are wasted. One city with just one of the doubler projects can reach one tech per turn here. The late game of a speed transcendence run becomes that of managing your way around the overflow limitations so as to lose minimal labs production.

I've picked those two cities in particular as SSCs just because they are the right distance apart in the city list order (you can see where I've put numbers in front of each city name to keep track of that.) With six cities in front of the first SSC, I can tweak the micromanagement every turn for those six to finish the current tech, then the first SSC (07 Zvedny Gorodok) plus the next six cities to finish another tech, then the second SSC (14 Lomonosov Park) applies its full value without truncation losses. I do this by shifting around specialist types, supply crawlers, and borehole tiles. University Base also became a halfhearted third SSC; there is no third labs doubler project, but it added the Network Backbone for straight labs production and most of the multiplier buildings.

After Digital Sentience, the next tech beeline is Secrets of Alpha Centauri for Transcend specialists, but otherwise doesn't really matter. Three techs per turn is just about terminal velocity, and there isn't anything else available to speed things up. Nanohospitals and Quantum Hospitals come too late, literally just two or three turns before the end. Mostly the cities just built supply crawlers for the late projects, including the Universal Translator for two free techs. I actually didn't need very many projects at all; here in this shot I pasted in the entire list. Nanofactory was worthwhile, the half upgrade cost representing a net gain for upgrading crawlers for the big final SPs.

My 30 turn window of nerve stapling ran out just before the end, so hence the Virtual World. That along with all the transcends producing psych was enough to keep happiness in check for the last couple turns.

2175.gif - 283kb

Here's the last turn before the end, even though I'm still two techs and two projects away. I have already filled cities #15 and #17 with 600 and 2000 minerals with crawler upgrades, selling all the Children's Creches to get the cash. (The Telepathic Matrix and Space Elevator are not real builds, but placeholders for Voice of Planet and Ascent to Transcendence.) SMAC has a ton of cracks in the interface to make this possible. I can discover the two techs, scroll-ahead to build the Voice of Planet in city #15, then city #16 is building something throwaway just to jump into the city interface again, and scroll-ahead to city #17 for the Ascent to Transcendence. Victory on turn 76.

victory.gif - 202kb

No longer mere earthbeings or planetbeings are we, but bright children of the stars!

You may download the saves to see for yourself.