General Category > AI War
Learning the Game
_K_:
Alright, Im a lil bit late for the first post, but here i go anyway.
When you play AI war for the very first time (assuming you have played other RTS games and have good idea of how they work, economy and military-wise), you set the 4/4 AIs to vanilla, or even lower. And just spend a whole bunch of hours getting the "feel of the game": what things do, how ships fly and shoot, what the map is like, etc etc. I think i have spent my first few hours in AI war just on pause, reading the tooltips and then the wiki.
If you feel bored by the AI doing nothing and want to fight stuff that shoots back, you can then switch to 7/7, or maybe 6/7, or 6/6 and start your first "serious" playthrough, preferably with 1-2 minor factions to study.
But thats just general approach i used. Now, considering your current situation:
First of all, there this "Objectives" tab in the "stats" menu. Very useful to tell you about stuff you could do.
Now, the point of the game is to kill the Home AI stations, right? Well, if you have no idea what to do, just start working towards that. I mean, you are playing against a 5/5! It is quite forgiving, so can you mess up here and there, be waaaaaay away of the optimal approach, it still should be fine.
So, first you scout as far as possible and try find the AI Homes and all core shield generators if you have em enabled. Scouts getting killed? They get spotted by tachyon guard posts at the wormholes, go kill those to scout further, it costsd you no AIP as long as you dont touch the comm center and the gate. All systems scouted? Good job, read how the core shield gens work ("Objectives" has a nice explanation). Once these are done, start working towards the AI home. I mean, to kill it, you need to get your fleet in there, and for that you probably want to conquer a system nearby. 2-3 jumps away from the AI home should be fine, as taking a system right next to it will set the Home system on alert and it will start getting MKV reinforcements. So you expand to the homes, maybe take other systems to maintain decent border line and get some bonuses.
Once both homes are killable, you, well, kill them. Its often a good idea not to kill the command station in the first AI home you attack until the second AI home is ready to die.
Also, here are some things to exploit:
1: Raid starships. Those guys can hit stuff protected by shields, have very nice damage and decent health. They 1-shot data centers and co-processors, and can kill guard posts, warp gates and command centers rather quickly.
2: Transports+cloaker starships to cloak em. An awesome way to get your fleet through an enemy system without taking any damage. Just dont forget to kill the tachyon guard posts. This is the best way to get your fleet in the AI home.
3: Dyson sphere neutral faction. Read its description. The ships it produces are incredibly good, just beware of the hybrids (AI plot), they can subvert it.
4: Harvester upgrades. Right now, the best way to boost your economy is to get MKIII harvesters. They will probably get a severe nerf in the near future, so abuse em while you can.
Cyborg:
Just read the after action reports and make sure that the difficulty is around what you play at.
apophispro:
Hi Dazio,
Thanks for your reply. So it sounds to me like you're saying there are a lot of dynamic variables, but there is at least one that is more fixed. Ultimately I'm going to need something to destroy the AI homeworlds. So what does that look like potentially?
Lets say I start the game with a ship cap of 300 total ships and 4 each of metal and crystal harvesters. What would I ultimately have once I was ready to attack the AI? Would I need at least 30 each of metal and crystal harvesters and a fleet of 3,000 ships? Would some of the special ships in the game reduce the fleet size I would need? You seem to be simplifying the game down to some primary variables, and that makes a lot of sense. How do those variables really interact?
I'll make a quick outline.
Start: One planet. Basic Resources. Two AIs of varying types.
Goal: Destroy both AIs by defeating their core planets.
Primary Variables:
Fleet Size
AI Progress
Resources
So if we were to really simplify this and boil this down. Forget all of the factors you can potentially add into the game like factions, golems, astro trains, etc. Just at the most basic level without too many intricacies, what are the primary variables and how do they affect each other?
For instance the end game requirements might be:
Resources:
Metal Income +400
Crystal Income +400
Knowledge - How much knowledge would I have gathered over the course of the game?
Energy
Fleet:
30,000 ships
AI Progress:
Below 200
Then over the course of the game if I take a system it drives my resources up or for an ARS my fleet size but it drives AIP up as well.
This actually strikes me as being a lot like philosophy. When you deal with philosophy, you don't introduce them to everything all at once. You just kind of get them started. For instance if anyone who didn't know formal logic got all of the intricacies dumped in their lap in no particular order, they would have no idea what to do. When you learn formal logic you kind of learn that there are genuses and species and definitions to start. You learn that there are some basic rules for those, and then you learn the myriads of problems with those rules later. You couldn't just jump up to the highest level of epistemology right off the bat. You have to learn how it's simple first even though it's really not. Then you can learn how it's complex.
So I can tell that AI War is pretty similar. Everything really has an alternate way of going about it. Chances are there's even some way to defeat the AI with a much smaller fleet size than most people would use. All of the variables affect each other in nearly incalculable ways, and so on. However is there a way to simplify the variables into an endgame goal and a set of rules for how they affect each other that is relatively accurate despite the fact that later on you will learn all kinds of ways to toy with those rules? I'm sure with certain types of AIs or other options those rules would already be broken.
What I'm getting at here is it seems like you're saying that there are no rules because the rules that there are can be broken in an infinite number of ways. However if you learn about definitions in formal logic, there are a set of rules that we can assign to being the standard for the concept of definitions as a whole. Then we talk about all the ways that they can be broken. Is there a set of rules like that for AI War that kind of define the gameplay as a whole?
apophispro:
Hi K,
I completely forgot about the objectives tab. I remember reading about that somewhere. I'll have to take a look at that.
So two points in your post confuse me. The first one is what is it that ends up taking over 10 hours? The objective seems so simple. I think that's a big part of the reason that I got to that point and felt so confused on what to do. I'd heard from so much of what I'd read that I couldn't just go straight for the core worlds, but I didn't know what else I was supposed to be doing. Every other goal other than going for the core worlds is flexible? It's just the difficulty of accomplishing that which makes the game last so long?
The second point is about capturing planets. I know this is sort of guerrilla warfare, but even in guerrilla war you wouldn't leave yourself in defenseless islands cut off from all sides. That seems to be a primary part of this game though. For instance starting out in the game I spotted a planet two jumps away from my home planet which had a large amount of resources. My immediate question was do I take that planet while leaving the planet between me and it? If these AIs are so strong and cunning especially at the higher levels (which they obviously are), how can I ever hope to defend all of these island planets? Is building defenses for each of them really going to cut it? Also I have a limited amount of ships, and do I need to be dividing them up to defend the planets I have? Sorry, the proper word seems to be systems here, and I keep using planets.
I scout out away from me in the beginning of the game. Now do I take the neighboring systems to my home system as a defense even if they suck for resources? Is that necessary or optional? Then it seems like every other system I'm going to take is going to be largely out on its own. The ARStations won't likely be right next to my home system either. It just seems like no matter what strategy I was going to use that by the end of the game I'm going to end up looking like a series of very small islands in a sea of AI. Of course that goes against every RTS instinct I have where the goal is always to build as strong of a base as possible. In almost any other RTS I've played going around and building little mini-bases around things that I liked would be suicide.
Thanks for your post!
Diazo:
Alright.
I'll give you a quick run down on how I start a game. Note that I am one of the more defensively minded players on the board, but that's not a bad thing as that gives you more depth to recover from any mistakes.
It also leads to a longer game as consequence however.
First, conquer every system adjacent to your homeworld, I want that defensive depth.
Second, scout out as many hops as you can to get the lay of the land. An easy trick is to stick the scouts in transports, that usually gets them a hop or two further then they do on their own.
Third depends on what you find. I play with Core Sheild Generators enabled so I have to deal with those before I go after the home worlds.
When I don't have an obvious target, I look at the map and guess where the AI homeworlds are, then I pick a system 2 or 3 hops out, depending on the map, and conquer it, the repeat the scouting process from my new base.
It all depends on how you value the different structures in your own strategy you will develop.
Note that the AI (generally) does not make unannounced attacks if all the minor factions are off so you have warning to prepare your defences when a wave or cross-planet attack comes in.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version