Building infrastructure after Update 2.0

[Gufnork]

Why have early FM at all? Before your expansion phase is done the 2 growth/1 industry is almost as good as the 2 economy and if you need to sacrifice something for it, it's not worth it. I use FM for a VW Zak, for Lal (not Domai, he just builds CPs too fast) and ignore it in general for everyone else.

Miriam is the last person I'd use FM with, she needs to get a powerbase fast so that she can start kicking butt.

Edit: Ok Yang is the last person I'd use FM with, but still.

[Ogie Oglethorpe]

With proper maintenance costing the build of Planetary Energy Guild (PEG) becomes a tremendous boon. If midgame energy woes are giving you fits look by all means to build this gem of a SP as soon as you can divert your tech beelines.

[Curiosity]

USC,

Firstly, I admit that when I say early-FM I'm thinking 2120 or before. The earlier you start using FM, the easier it is to keep on using it, since you get crawlers et al. that much faster, and the drone problems are not significant for a while yet.
There's no particular trick to this, just research cent. eco, ind base and ind. eco and do it, creating doctors as needed. I think you'll be surprised.

Once you get crawlers early enough, building both WP and HGP in one turn in the 2150's will look fairly reasonable. Not with Miriam, obviously but with the good-to-middling tech factions. This is really the key; once you can research fast early on, you can solve your problems almost before they start.

The WP is very good, but IMO, mainly for a situation where you can get it built significantly before you get environmental economics. (I've had some interesting games skipping the whole Industrial line (heresy! ) and building the WP very early, then booming with condensor farms/forests. Works OK-ish on very small islands, but you kind of need an AA.)
Generally, if you have to choose then I'd take WP with low-tech factions, and HGP with the high-tech ones.

When I said planned with police, I meant going to planned and stopping to build a unit per base for police, not police/planned. Sorry, that wasn't very clear! In detail, I was thinking of a 2120 FM to boost early science, and switching to planned around 2150, when you have the essentials down. Police/planned wouldn't be to my taste, though it would be kind of viable with the Gaians or CyC... hmmmm...

Post some of your save games, btw! 2125, 2150 and 2175 from a game would be very illuminating.

[UnityScoutChopper]

My sincere apologies to all for not posting those saves! I've been too busy playing SMAC to get around to it! :-) Above all I've been testing Curiosity's proposal "early" 2120 FM. (I'd call it "hyper-early...). Incidentally my verdict is: occasionally extremely useful, often very harmful. Having now actually seen such extremely early FM in practice, I really have to say that, except in some special cases, the need to convert your second citizen (let alone your first citizen, as is necessary after the soon-surpassed base limit) to a doctor until an RC can be built does more harm than the extra early science can provide. Generally, during the very early game you can gain (admittedly undirected) science fairly easily, but drone control without police is a pain, whereas over time the reverse becomes true. Thus (again, having actually tried it out with several factions) I'd limit this strategy to cases where you need some directed science (e.g. an island start, a start demanding war using Impact Rovers, crawlers to build an SP, etc.) more than easy drone control. The beginning, to me, is all about expansion, so the real question is: will I expand better by having new tech of my choice, or by actually being able to build my way out of a paper, er, I mean, by having the benefits of not using extremely early FM?

An exception to this is cases where one has more money than one knows what to do with and can thus just rush-buy RC's everywhere, but such cases are rare.

On the other hand, I have learned that in many cases (i.e. barring an early war), Lal and Domai were almost born for running FM as soon as possible.

Likewise, extremely early FM really does represent salvation for those hideous island starts -- the slowed initial expansion is meaningless, since at first you can't go anywhere anyway, and thus you are to a certain degree getting something for nothing. Considering that that "something" is mainly directed science, and that this is precisely what you need to leave the island, the synergy is very good.

Now on to kick myself into finally posting those games!

USC

[Curiosity]

Ah, would you believe I normally expand faster in early-FM than in planned or simple?

The first reason is - I don't build scout patrols. I call this the Crazy Eddie expansion but unless you're playing high native life then it can work well.
I also apply my energy credits carefully and micro intensively, so that the base (hopefully) only spends one turn at size 2.

This works better in some situations than others. For example, it works pretty badly at 3a.m. when you're out of coffee. You need to really pay attention for those first 40 turns. It's also degraded if you get lots of nutrient resources.

You can normally prevent worms calling by not moving through fungus squares, and only popping pods once you've built a base next to them (this ensures the result of a pod is always positive.) If they do appear then a terraformer can usually be stationed in the base before it attacks - a terraformer in a base has no combat penatly vs natives, for some bizarre reason.

If there's a lot of fungus around then I may comprimise by building a scout every 3-4 bases, and stationing it so that it can reach any of them in one turn. And of course, once I've got two or three trance scout rovers, well my empire's safe as houses!

[Gufnork]

You do that in MP as well? Because if I saw that your empire would be gone before you could say "Scout Rover". Plus, when I do that worms always find their way to my bases and rip 'n tear. I do it with Morgan, Domai and Lal in SP. I'd probably concentrate on defending my borders in MP with Morgan, but build defense with Lal 'n Domai.

[Chaos Theory]

You MUST defend as little as possible in MP until given incentive not to, or you lose turn advantage. Your scout rover would be spotted by my sensors or terraformers before it could take a base, ideally, and even if it grabbed one, I could immediately construct defense in my vulnerable bases after that.

The exception is when someone hits you so early that you don't have formers and sensors, or expendable bases, as Curiosity demonstrated. He nailed my capital in 2115, coming from the direction I hadn't yet explored. That's just a risk you have to take.

As for worms, build sensors, and be careful when popping pods. If you lose some dinky base, oh well, make another. All your important bases should be shielded by new expansion.

Incidentally, I'm a big fan of hyper-early FM, excepting Zak (gets to Wealth by the time he has 40 energy), Dee, Yang, and sometimes Domai (+3 industry ROCKS, and you need the +2 growth to keep up).

[Gufnork]

I'll take half your bases in one sweeping motion. After that you're screwed. Of course, careful terraforming would help. And keeping resources to speedbuild defense. Heck, if that's the way you run things I could just run a rover up to your front door, shout boo and you'd be screwed out of the game while desperatly trying to defend. You don't think I'm dumb enough to concentrate my forces at one spot early so you'll have an easy time defending?

Of course, I need to spot your weakness early. If you want to risk losing the game before 2150 to get turn advantage, by all means. I don't see how it's worth the risk. Worms and nasty people (like me) can screw you over too much. And I never rely on luck, it always fails on me.

[CEO Aaron]

If your entire strategy is founded on blundering into my undefended bases with 2-3 rovers, I don't see how you're relying on luck more than the guy who waits for attackers before building defenders. I've got sensors to see you from afar, and can cancel what I'm building in my border bases to rush-build some synth garrisons. If your offensive turns out to be too large to withstand, I'll sell off my upgrades and cut the roads leading to my other bases, and rely on the fact that I've still got 3-6 more bases built to keep you from consolidating your gains.

On the other hand, if you don't find me, you've blown all that early turn-advantage for nothing, and with IA safely under my belt, I'll be long-dogging it toward rushed wonders, tree farms and clean reactors.

It seems to me that the truly conservative approach is to expand naturally and explore aggressively. Once you've found your rivals and probe them, you can find out which faction is unprepared for your assault and build sufficent attack units to make the outcome of the engagement beyond doubt.

[Chaos Theory]

You won't take my bases so swiftly because I have sensors at the periphery. 1-2-1s or 1-3-1s in bases on sensors will slow you down greatly, and they can be rushed cheaply. Terraformers in bases with sensors hose native worms.

If you force me to build garrisons, you merely force me to return the turn advantage I had gained by skipping them. However, I'm returning it much later, so I'm coming out ahead anyway.

Worst case, you force me to struggle for survival. If there are more than two players in the game, everyone else pulls ahead while we pour resources into war. You lose. I lose.

If you find me far too late, I've pulled far ahead of you. You lose. I can win.

If you find me at an intermediate stage, you might convince me to help you out in exchange for not fighting you and falling behind. We both can win. This window is kind of narrow.

You can only win if you scout me out without tipping me off to your existence, and swarm me with rovers and probe teams. Better hope I don't stumble on you before you're ready.

[Gufnork]

Worms can come out of nowhere and smack your base.

Compare the cost of 3 scout rovers to rushing synthmetal in each base and you find out who comes out on top. Add any minerals you lose on changing production to defense. And if I can get my hands on even one base, I've gained from it.

1-2-1s or 1-3-1s in bases on sensors will slow you down greatly

No, it will turn me around. Once you get them my objective is achieved, you've lost all the money you for some weird reason hadn't spent already. Mission accomplished, now they head to scare another sucker.

But if you can afford to rushbuild all that military, you don't have turn advantage. Spending your money gives you turn advantage.

But do you really think I rely on this in a game? This is just something to give me an edge. And if I happen to find a pair of unity rovers from pods, it doesn't even cost me anything. And it's not one of my most common strategies, since I rarely manage to find someone I feel I can screw over with this.

[CEO Aaron]

I'd much rather sacrifice a turn's energy income on turn 40 than a 3-5 turn's mineral output on turn 20. Especially if it turns out you don't find me, which, on a huge map, is pretty likely. Yes, eventually I will have to build defenders, and I'll usually have them built in most of my bases by the time I've completely finished expanding.

[Chaos Theory]

Sensors see worms before they smack your bases. Forests clear fungus. Formers clear fungus. Formers are EVERYWHERE. Any worms that appear will kill a former first. After that, I simply move all the vulnerable formers out of the way, and into vulnerable bases. No, native life doesn't worry me (Exception: hordes of locusts produced by pops in certain conditions).

Gufnork, I think you underestimate my typical cashflow in the scenario you're proposing. I'm not sitting on a pile of cash. I might have 50 or so ecs unless I'm saving up for a project. But I'm making 20-40 on up a turn.

Your 3 scout rovers cost you ~60 minerals directly, plus the support or reduced former count. If they're unity rovers, obviously, this cost is negated, but then I have unity rovers, too.

When they reach me, if they're just scout rovers, they'll be greeted with scout patrols. Your rovers will have small areas that they threaten, given the amount of forest I like to plant. I can simply pop out scout patrols in the threatened bases at minimal cost. Worst case, in a fresh base I'll have 3 mins/turn, so a scout patrol would cost me 6/9 * 44 ecs = 30. Whoop-de-doo.

If you get your hands on a base, you probably obliterate it, because it's probably size 1. In doing so, you cut several percent off my size, but it's nothing irreplaceable. If you grab a size 2, it'll be a size 1 kept out of drone riots by the very unit that captured it. Size 1 bases in hostile territory aren't very productive, and leave you vulnerable to probe raping. In fact, I might even let you take one for that purpose.

Actually, another factor in this is whether the game is PBEM or TCP/IP. The sequence of builds and attacks varies.

TCP/IP: (Each player finishes construction), then (each player can attack)
PBEM (AFAIK): Each player (finishes construction then attacks)
with parens indicating grouping

[Mongoose]

CT, you overstate the cost to rush build a scout patrol.

- The first 10 minerals of a facility cost 4ec/mineral.

- Construction may be freely switched, even from one category to another, up to 10 minerals.

Rushing the scout in the circumstances you describe above will only cost you 24 ec, if done correctly...set production to a facility...spend 24 ec to rush 6 minerals...change construction to the scout. The six minerals are not reduced by the switch. The scout completes the next turn, unless one of your forests is occupied or destroyed.

The only time this approach becomes problematic is in your HQ if you have already built all facilities for which you have the tech. All other bases can use 'Relocate HQ' as the facility to 'build'.

The move sequence you describe in TCP/IP depends on whether simultaneous moves are being used or not. If not, the only difference from PBEM in the sequence of play is that all the AI take their turns at the end of the game turn, at the same time as the native life.

[Chaos Theory]

Mongoose: you're correct about the scout patrol, of course. I'm not used to being able to profitably hurry one thing and switch to another.

I'm pretty sure that non-simult TCP/IP games still have all players complete construction before anyone moves. You can twiddle your cities before your turn comes up, and you can't change what they produce before your turn comes up, after the start of the turn. However, it has been a while since I played in these conditions. In simult TCP/IP, you complete construction even before comps/natives move, so if a mind worm or something appears next to your base, you can rush a defender. There's a nasty cost for this advantage, however:

PBEM: you complete construction, ecodamage causes a pop with worms, your units move and attack, remaining worms move and attack
TCP/IP: you complete construction, ecodamage causes a pop (not sure about the order of these first two), worms move and attack, your units move and attack

This has the important consquence that an empath scout rover is NOT adequate defense against huge pops in TCP/IP, whereas it is in PBEM.