Building infrastructure after Update 2.0

[Gufnork]

You must terraform much faster than me, if you're able to have sensors and forest everywhere that early. But if you feel that comfortable with your strategy, by all means continue. We're not likely to meet on the battlefield anyway.

[Chaos Theory]

The first thing every base of mine does is make a terraformer, preferably rushed. The second thing many bases do is make a terraformer. Multiply that by many cities and you get a horde of terraformers. The number of cities in my empire is proportional to its radius squared, but its circumference is proportional to its radius, so I end up with a great many formers at my periphery.

Of course, exactly what year are we talking about possible conflict? On the early extreme, 2115, I'm vulnerable, but even if I had build garrisons they'd be out popping pods. On the late extreme, 2150 or so, you can't make a dent with most forces.

Also, TCP/IP games start slightly accelerated: your HQ has a rec tanks, rec commons, and 2-4 population (2 for Zak and Morgan, 4 for Dee, I think 3 for all others). You also get a free extra scout rover and terraformer. This makes it all the easier to defend against an early attacker.

[Curiosity]

Hmm, lots of interesting discussion.

Bottom line, if someone hits you with no warning early in the game then it's a very bad situation. Scout patrols, even synthmetal garrisons, don't make much difference. Impact weapons aren't hard to get, and lasers will eat up synthmetal with a little attrition. Defenders are pretty damn useless in SMAC, you need forces to kill his rovers on the way in and counterattack with.

However, someone shouldn't - generally - be able to do that. If you start reasonably close to someone then it's obvious by the borders being distorted. Also, the map generator is semi-predictable about how it places the players, and you can scout the most likely occupied spots on your continent. Once you've located someone, you ideally either pact with them or infiltrate, so there's not much chance of them sneaking up on you. Not unless you fall asleep.

[UnityScoutChopper]

Curiosity, what do you mean when you write that the map generator is semi-predictable about how it places the players? Other than suspiciously often placing my Morgan starts next to a homocidal maniac? :-)

USC

[Gufnork]

Curiosity, I'm not talking full assault. I'm talking sending your scouts on a raid. Even a scout patrol has a good chance of defending against a scout rover and you can't afford attrition, so the wisest thing is to give up if they have any defense. Or pop their formers.

[Curiosity]

Gufnork,

Raiding is pretty unattractive in MP. The sequence would tend to be, you take 1-2 of my bases, I load up some defenders/formers to stop you, you give up, then I send as many impact rovers as I can build to wipe you out. You need to put the other guy out of the game almost as soon as you attack.

There is certainly a segment of the game where what you're talking about is possible, in that occasional case where you get a couple of rovers from pods and we start very close - as CT said, I've done it to him recently. However, even if he had built some defenders, I would still have come in all guns blazing in another 10 turns, as the strategic situation meant he had to go. If it hadn't, then I wouldn't have attacked him - I would have tried to pact/trade tech etc. because I'm looking for the greatest overall advantage, not just to kill one of the three opponents.

So, IMO, early defenders will very rarely make much difference. I'm more worried about the other 19 games out of 20, where they'll slow me down to the point of not being able to compete.

USC,

You can see where the main landmasses and continents lie on turn one by the distortion in the map grid of black squares, and the start positions are chosen according to landmarks and continents. E.g., In my game vs CT I saw a particular landmark, knew the generator was likely (~50%) to place another faction to the north of that, and shot over there, turning up CT about where you'd expect.

Open up SMAC and generate about 20 games, and look at each start position in the scenario editor. It differs between SMAC 4 and X I think - I find it much more predictable in SMAC - but you'll get a feel for it. It's an intuitive thing, as much as logical.