T-hawk Plays Alpha Centauri

[Modo]

This particular game maximized the Stockpile bug more than I've ever seen before.

Does this bug apply to overflow production as well or just the natural one?

[yuris125]

And how come there's no HOMM reporting on your site?

This is a really good question, I would also like to know

[Kofiman]

Speaking of HoMM, am I crazy for thinking that HoMM 2 might be the best of the series?

[haphazard1]

Been a while since I played HoMM2, usually go with 3. More HoMM reporting would be a good thing. Maybe not here in T-hawk's SMAC thread (although we do seem to have difficulty staying on topic in SMAC threads ), but somewhere.

[T-hawk]

Well... I haven't been part of any community based around HOMM and eager for reports on it. I only ever played 2 and 3, and the last time was even before the Civ 3 Epics, aside from a couple playthroughs of the odd spinoff Clash of Heroes on the Nintendo DS.

And I definitely agree on liking HOMM 2 more than 3. 3 is surely a better game in terms of balance and AI, but I always found 2 quite a bit more fun - where the unbalance is the charm, like SMAC. The one aspect I never liked was how 3 scaled up the available troops per castle and their costs, but didn't scale up the income, so you were left recruiting a much smaller proportion of available troops and had to leave so much unused. That's not a problem per se, but I always found it sorely unsatisfying.

My other problem playing either HOMM was save-scumming. Like every single battle, I'd realize halfway through a better way to do it, and reload to start it over. Not luck abuse, the RNG in HOMM is pretty low impact, but just realizing how I could take fewer losses if I engaged and handled enemy stacks in a different order, ways that were deterministic enough to see at the start if I'd projected ahead that far. If I tried to stick to a no-reloading rule, what I'd end up doing was restart an entire scenario or even campaign for one flawed battle. I just couldn't ever break myself of the "I should have seen that" save-scumming habit; even now I do it quite a bit more in Civ/SMAC than the reports convey. (That's why I love FTL so much, no save-scumming, copying the file around out-of-game is cheaty enough that I don't.)

[yuris125]

I would love to see your report on a HoMM game (either 2 or 3) if you ever find interest in playing and reporting one

Back to SMAC, I'm sure you can win the election at this point, but if there are still interesting decisions on the way to transcendence, I would like to see that. Winning by election is boring

[T-hawk]

Chapter 14, in which the multiplicative math will crazy-go-nuts.  We begin the session producing 2600 labs/turn.

2213: Applied Relativity discovered.  This tech enables the Supercollider secret project, which doubles the labs production of its base.  This is a special doubling, not +100% additive with network nodes and such, but fully multiplicative with them.  I built it in one turn with an upgraded supply crawler, as always.

We complete several Orbital Power Transmitter energy satellites, which each add 1 energy per base.  (Orbital production isn't really shown anywhere on the base screen.  It's just included in the bars that show the total quantity, but with no other visible source.)  The satellites increased my labs production by about 300.  Plus more from the usual pop-booming, multiplier facilities (some newer bases are still completing research hospitals and fusion labs), and boreholes (my formers are finally getting to those in the northwest bases having completed the carpet of forest.)

Modo keeps asking how tech overflow and turn processing works.  I'm going to illustrate it here as clearly as I can with an example.



I currently have 1392 labs towards the 2160 cost of this tech.  During the turn processing step, each base in order as indicated in the lower-left window will add its production to that total.  When the total exceeds 2160, then right in the middle of upkeep processing, the popups occur to award the new tech and ask to select a new one.  Then the remaining bases in turn order continue to add their labs.  It can (and will) happen that the downstream bases continue to add up enough labs to get another tech, for which the popups will occur the same way all still on the same turn.

Any overflow from the base that discovers a tech is lost.  That's why I show this picture here, to show how I'm going to work around that.  I numbered those bases 00 through 02d to indicate their turn-processing order (the University founded those bases in-between mine, so they occupy intervening slots in the global base list, where they remain after I captured them.)  Right now, bases 00 through 02d will come to just about 100 labs short of the tech.  03 Leader's Horde is my best science base, and will actually produce twice that 260 this turn because it's about to complete the Supercollider.  What will happen is that Leader's Horde will apply 520 labs where I need only 100, and the remaining overflow will be lost.

To fix this, then I micromanaged those first several bases to make every specialist a Thinker for the labs.  Then the total would indeed be enough for 02d Monitoring Station to complete Silksteel Alloys, so that 03 Leader's Horde would then apply its full production towards a fresh tech.  From here on, I will refer to Leader's Horde as the SSC (super science center), that's the standard community term for the base with the labs doubler projects.

So at the end of this turn, my labs production is up by +600 from the prior turn, now at 3200.

2214: Research both Silksteel Alloys and Monopole Magnets.  I order up one 4-armor silksteel unit, to prototype it, so that I can use 4-armor for the next crawler upgrade.  

With yet more energy satellites and the Supercollider, labs now at 3800.

2215: Research Nanominiaturization, and fall just short of the next tech too.

Core cities are running out of useful things to build, now that I have a full complement of energy and food satellites (satellite yield is capped at the size of each base which is 16.)  They now go for biology labs and energy banks, small potatoes but may as well do them, also since each facility results in the Stockpile bug yet again.

Nanominiaturization enables the Nano Factory.  This secret project halves all upgrade costs, which will more than pay for itself in cheaper crawler upgrades for every other project.  This tech also enables the Hovertank (3-move) chassis, which I also prototype in order to build crawlers on it.



I also built this cool unit, a 4-movement probe, on a hovertank chassis for 3 base plus 1 for being elite.  (I forget exactly what factors increase probe team morale; it's a passive effect of several techs; in general late in the game they seem to end up elite without any particular effort.)

Gaia uses a probe to steal a unit of mine, an old rover that I left to watch the border.  I'm a bit worried that she's got 380 energy on hand, and Morgan has over 700, which could be a headache if they managed to mind-control a base.  I want to get to the Hunter-Seeker Algorithm (immunity to all enemy probe teams), but I keep spending my crawlers and energy on other projects instead.  In the meantime I do have that tank probe and one other on probe defense in my two bases bordering Gaia and Morgan.

Labs now at 4200.

2216: Research both Industrial Nanorobotics and Digital Sentience.

Industrial Nanorobotics enables the next mineral multiplier, the Robotic Assembly Plant.  But it's really not worth it; 180 minerals for one of those will not pay back before the game ends.  But Digital Sentience is where we really go nuts.



Digital Sentience enables Cybernetic Future Society.  First I'll talk about its negative: the -3 Police penalty is a headache, turning off four drone control in each base.  (I'd actually forgotten about that, because it's usually irrelevant on top of Free Market's -5 police.)  I have to deal with that by turning most of the specialists back into Thinkers for the psych.  I need the Thinker specialists anyway, to make the last few bases this turn and first few next turn add up to enough labs to complete the next tech, so that the SSC can apply all its production to the following tech without lost overflow.

But that police headache will last for only one turn.  Digital Sentience includes The Network Backbone secret project, which eliminates the negative of Cybernetic.  My SSC is just about to build that anyway, with a crawler upgraded for half price via Nano Factory.  This project gives another good shot of research: +1 for every network node in existence (around 40) and for every point of commerce at this base (also around 40), although that doesn't go through the labs multipliers.

The upsides of Cybernetic: +2 more Planet, which I don't really care about.  +2 Research, which of course joins the multiplicative party.  And +2 Efficiency is the key to everything here. A rating of +4 Efficiency or better zeroes the penalty for unbalanced sliders, so now I can raise labs from 90% all the way to 100%.  Everything multiplies together here: that last 10% slider step, the removed -8% unbalancing penalty, and the +2 Research itself all multiply with each other.



The net result of all that: my labs production skyrocketed from 4200 to 6600 in this one turn!

2217: Research all of Unified Field Theory, Doctrine Initiative, and Homo Superior.

Homo Superior enables the Nanohospital facility, another 50% labs multiplier.  It's expensive at 240 minerals, but I order it up in most bases anyway, since they don't have anything else that matters to build now.  With those ordered up, I decided it was worthwhile to do the Living Refinery secret project (+2 Support rating, enabled back at Advanced Spaceflight), for the few extra minerals that might complete a few of those a turn sooner.

Homo Superior also enables the Universal Translator.  This is a secret project that grants two free technologies, equal to Darwin's Voyage in Civ 2 and Theory of Evolution in Civ 3.  Hive HQ orders up this project and builds it with an upgraded crawler too.  It would be slightly more optimal to let the Translator wait for later to cover more expensive techs, but I'm taking it sooner, both because 1) juggling 5+ techs in a single turn is already complicated enough and I'd rather get that out of the way earlier, and 2) reaching Transcend specialists sooner will make up much or all of the cost differential.

Unified Field Theory enables the other labs-doubler project, the Theory of Everything.  As I mentioned before, this and the Supercollider each double a base's labs, multiplicatively with each other and all facilities.  The SSC builds this project this turn with an upgraded crawler.



I said I was done showing crawler upgrades, but here's one more with the new toys that have come into this process.  The 3-move hovertank chassis makes it more expensive, and the Nano Factory halves its upgrade cost.  This is now a gain of (324 - 36) = 288 minerals over the crawler's base cost, for 190 credits, now a ratio of 0.66 credits per mineral bought.  The standard ratio for rushing a project is 4, so upgrading crawlers is now 4 / 0.66 = 600% as efficient as direct rushing.

After Cybernetic SE, my next usual endgame beeline is Secrets of Alpha Centauri to enable Transcend specialists, which produce 4 labs each.  The other possible path from here is Super Tensile Solids to enable hab domes for unlimited population cap, but I'll skip the hab domes because I'm tired of micromanaging for pop-booming.  Also because I found a direct beeline to Secrets of AC with no missing-tech holes, but the path to STS incurred at least two however I tried to arrange it.

Labs now at 6770.  (That's the value shown on the F2 screen before ending turn, but doesn't include the changes that will come from the next turn's production, including the ToE.)

2218: Research all of Centauri Meditation, Centauri Genetics (the two freebies from Universal Translator), Centauri Psi, and Sentient Econometrics.



The three Centauri techs each increase various production from fungus squares.  This had the effect of finally giving some life to those Spartan bases I conquered.  I had those super drop formers planting forest as fast as they could, but Sparta's island was still densely covered in fungus.  But now with those fungus tiles boosted to 2-1-1, the Spartan bases that had been suffering suddenly became considerably more viable.

I'd been trying to buy facilities to get the Spartan bases going, but couldn't come up with the energy, because I kept needing to reassign Engineer specialists to Thinkers for various reasons (the Cybernetic police penalty, and working around overflow limitations regarding the SSC), and spending the energy I did have (190/turn or so) on crawler upgrades for the projects.  Next turn with Transcend specialists producing economy, I did have a lot of energy (800) to rush each Spartan network node, which seemed like the best available energy:labs conversion.

Labs per turn now at 7800.  Most of the gain was Theory of Everything, which doubled the 600 in the base itself, which is before the 40% SE research modifier, so the overall total from the project comes to 840.

2219: Research Secrets of Alpha Centauri, Photon/Wave Mechanics (freebie from Secrets), Probability Mechanics, Self-Aware Machines.

Secrets of AC enables the Transcend specialists.  The game auto-converts all Empaths and Thinkers to the strictly superior Transcends (+4 labs, +2 economy, +2 psych), so on the previous turn I assigned all specialists to thinkers.  Here it so happened that Secrets of AC came early in the turn processing order, so that everything switched to Transcends early in the production cycle, giving the benefit of most of them this same turn.



The Transcends bring my labs production to over 10,000.



This is my favorite moment in all of Alpha Centauri, the dawn of the Transcends.  This is the culmination of SMAC's multiplicative splendor; nothing in any other Civ game ever attains this level of power and payoff for the effort to get there.

Their productivity is so enormous that you want to make just about every citizen possible into them.  There is no land productivity now besides a borehole that is worth working over the specialists.

Notice how there are just four laborers left on the land.  Two condensors plus orbital food feeds an entire size-16 base.  (And those condensors could even be crawled instead.)  I don't care about the -2 food shortage, the game will end before the food box runs out.  That's another advantage of pop-booming; growing that way doesn't even spend the food, it stays in the box and can be depleted at shortage afterwards.

This is what I meant that cities never need all their workable radius.  This is why I never built hybrid forests or even the soil enricher terraforming improvement.  Food goes obsolete.  Forests go obsolete.  Tiles go obsolete.

I love how SMAC's mechanics here deliver what's going on thematically.  We have robots and satellites doing all the material work of reality for us.  All we now do is think our awesome thoughts until we emerge into the next form of existence.

[Modo]

Well, this already feels like more than a Civ5 spaceship speedrun endgame "and now we fill the 8 turn beaker window and use our GP's to complete the tech tree and...."
I don't even want to count how many times you've used "multiplicative" in this one post, it multiplicatively multiplied as the labs kept exploding.

So I'll answer my own question, SMAC definitly takes a player from 75% to 100% and the endgame seems more fulfilling as you still have areas to improve and ways to translate one resource into another (would still like to know if the stockpile bug applies to overflow production as well but this can be easily tested).

Now for the questions

1) Is it interesting or viable to do an OCC in SMAC?

2) If microing the cities to not waste labs on completing a tech doesn't work (assuming there are such cases) are there other alternatives to avoid the waste?

[T-hawk]

I want to stress that this endgame is normal for SMAC.  My Civ 5 stuff involved all sorts of rigged map setups and particular faction and religious abilities and low-difficulty exploits.  Everything I did in this playthrough is entirely possible on any typical map and with any faction on any difficulty.  Even the supply crawlers never really did anything I couldn't have done otherwise just a bit slower.

The Stockpile bug doesn't care about overflow, it just credits minerals equal to your current production.  (Now I see where you're confused: overflow isn't part of the per-turn production number as it is in Civ 4; overflow up to 10 is just placed into the production box.)

OCC is very much both viable and interesting, and SMAC is in fact where the idea started.  I did it two or three times long ago.

Lost labs overflow is never more than one base's worth.  (But that one base's worth can be immense if it's the SCC.)  If you're going to lose enough to care about because that one base overshoots, the right thing to do is switch some labs specialists back to engineers or land laborers instead, so that the would-be-lost labs never happen and you get other stuff instead.

[Modo]

That makes perfect sense (on both overflow topics).
So....OCC SMAC?
I know you've had your fill of SMAC, maybe book it as a future project; after all, how does a normal OCC playthrough looks like? And I mean a playthrough of yours.

On the Civ5 stuff, you actually ended up using all sorts of rigged map setups and particular faction and religious abilities and low-difficulty exploits but in the begining I think it was closer to "average" setups. Come to think of it, as you improved the execution you've showcased the different threads of the strategy in different attempts so one would have to read them all to grasp the overarching strategy....have you thought about an article focused on the strategy of it alone with links pointing to which attempt achieved best on each component? A sort of synthesis if you will. You have a couple of them on the city count and social policies but the info is spread and not cross refenced or indexed and does not attaing the academic level; besides, it's always fun to quote and reference yourself.

And to expand on that, do you ever go back to Civ1, 2, 3, 4, or 5 or are you prepping for 6?