1) Build a vehicle factory. (I do this first thing.) 2) With the rest of your starting resources, build a constructor and a spinner. Then build a second spinner. (I do this because when you play against real opponents they bomb your starting fabs. 3) Use your first constructor to start getting metal. 4) Let your economy grow for a bit. This is really important. Building new fabs does nothing for you until you actually have resources to use. It just looks pretty and costs a lot. 5) After you get a good econ, build another constructor, and use it to build an air factory. 6) Build a scout and find the enemy. 7) Look at all the fastest routes from the enemy base to your base and build a double turret at each place. (Don't forget ur op walls.) 8) At this point, set your vehicle factory to produce mostly tanks, some spinners, heavy tanks, and constructors. I almost always leave my factories on continuous. 9) Your turrets should deflect all the opponents attacks. Spinners and towers can protect your metal from air. Sit back, build econ, and get to t2. 10) Get a lot of fabs to crank out a t2 constructor. Then start cranking out t2 tanks and shellers. 11) Steamroll the enemy. He'll usually have wasted all his resources on avengers and then send his commander to try to kill your t2 army. Blow him up. This worked for me.
If this works... which it's a weird opening build... It won't work for long. When the AI gets smarter tactics like this won't really work.
Maybe not. But it works great now. If I had to summarize the strategy... build towers and walls, and then sit behind them while you tech up to t2. It works great against hard AI and human players.
Factory first is a recipe for energy stalling. Do that first in a 1v1 and you're dead before any shots are fired.
I start with 1 power gen, then i build 1 bot factory thats set to build 1-2 basic bot fabbers straight away, i build 1-2 metal extractors with the commander, and i use the fabbers to turtle/ build turrets/walls. While fabbers are building defenses i use the commander to build 2-3 more p-gens and extractors. When these p-gens and extractors are up i start building anti-air bots and doxen.. next i build my main attack force, usually 8-12 basic air factories and an advanced bot factory (air is still OP, easiest win is with 100+ scouts and as many bombers as you can build as fast as you can (i try for 50+ bombers) . Move the scouts and the bombers in small steps slowly together keeping the scouts close to the bombers, they become an airborne meatsheild soaking up most of the anti-air attacks (this is why you need lots of scouts).. If you manage to get an advanced bot factory out before you start the bomber snipe then build lots of gil-e's to pick off as much enemy AA defenses as you can. This tactic only works for me on a single planet battlefeild,, with multi planet systems a whole new strategy is needed (which im still working on figuring out)
Hard AI really isn't that hard to beat. Your build in kinda slow though, I think it may not work against real people.
This is why I'm a little disappointed in economy modifiers for the AI. It is perfectly reasonable to lose a game, review it, and then attempt the same build. Players will get some really bad habits if they try to play like a cheating ai though.
Meh, you'll be able to tailor it eventually and see behind the curtain of what makes each level tick, if the SUPCOM2 interface is anything to go by.
this. but this isn't my only issue. How about taking some of the top players from the ladder and asking them their build and tips to code it into the AI like I suggested when real progress on the AI first started showing? I really would love the idea of facing an authentic AI. An AI that literally really is playing better than me.
That would need to be done over again every time the balance changes. The advantage to how Sorian does it is that the AI can learn to play again on its own after a balance change.
Oh, I wasn't aware the AI was so stupid it needed help. It seems like it could just be taught to micro like a fiend. But whatever. Well, thanks. I've reverted back to 2 metals and 1 energy for my start.
Now its very hard to expand faster at beginning than AI, but its easy to make few gunships&fighter groups and just kill its fabbers. Then you send bunch of yours, make 2-3 flaks + radar + 1 lolzor turret, go for next outpost and repeat. This way AI cannot expand and it feels almost like beating a kid. But usually i let AI expand to get more fun and legendary battles, nothing beats smell of burned metal in the morning.
I think the AI's main problem right now is not enough emphasis on the ground war. But it really really really loves nukes. A lot. If it spent as much metal on T2 vehicles as it spent on nukes, I think it would be a lot harder to stop.
