Endgame advice

[Tim James]

What is the best way to approach the endgame in my first campaign?

I'm nearly done acquiring all the advanced research and fabricators I'd like. I've probably destroyed a few too many planets. It's my obsessive need to have paths cleared to all my bases of operation. Even if I try to be smart about it I know it's a bad idea based on the game design. I think the AI is down 20 planets from its original 79 and AI Progress is almost 400.

I haven't made much effort to destroy warp gates on my borders and I haven't kept up with knowledge farming. I have 12000 sitting around but none of the Mk2/Mk3/starship upgrades unlocked. (I mean the special ones I got from advanced research.) I assume I need some of those to have a chance in the final assault. Do I need to consolidate my borders to funnel warpgate raids to one or two places, then work on acquiring knowledge? Seems like a bit of a grind, especially since I left a few planets on alert most of the game.

The AI are split by 15 hops. Do I absolutely need two attack forces? I was thinking about queueing a lot of docks near the pathway down to one AI, then they can build as my other force is slowly destroyed.

[Kjara]

Honestly, I tend to hit one, followed by hitting the other right afterward with my entire force (having the reinforcements rebuilding your losses for the 2nd assaulte is a good idea yeah).  As long as you have the way mostly cleared to the 2nd homeworld (you won't want to take the planet next to it to keep it from being alerted, but having a planet 2 or 3 hops away is a good idea), you should be fine.

Yeah, you will generally want at least 2-3 or so of the ars units unlocked to mk3 level(and likely bombers/cruisers/fighters and the unit you started with), and I like having at least the fleet starships(spire line) unlocked for the endgame-- having a 2nd line on top of that(or even a partial line, say a few dreadnoughts or something) wouldn't hurt.

Knowledge raiding as you go is a good habit to get into as it means you have more ships to hit those intermediate targets, meaning lower losses and faster pushes.

You can probability get away without consolidating your borders if you can split your current force into a defensive force and a raiding force for your knowledge raiding, and have enough turrets on the critical planets for your assault(homeworld, adv factory/fabs, and staging points for the two assaults--if you do this, make sure you have safe mkIII plants in case you do lose a planet or two).

[x4000]

An AI Progress of 400 is not horrible at all, I tend to have 400-500 when going for the final AI planets.  Be aware: when you destroy the first AI home planet, you're going to skyrocket by 100 AIP.  So, if you can coordinate attacks that is nice, but often that is not possible -- definitely sounds out of the question in your case.  If you've taken that many planets and have that low of AIP, you've done a very good job of striking data centers, which is great.

The final assault is difficult, moreso than any one Mark IV planet, but here's the thing: if you keep the AI from noticing you, it can be very easy, depending.  It's all about how you alert the AI and when, and what the AIP is when you get there, and how well you are able to reinforce.

I don't think I've ever managed to take out an AI home planet with just one wave of ships, in solo or multiplayer, so be prepared to take multiple runs.  Since all of the ships on the planet will come after you when you are attacking the home planet command station, that means you'll want to make your first priority to destroy all the guard posts.  Then getting rid of any force fields and/or fortresses.  And lastly, taking down the command station itself.  Try to do that when there are as few AI ships remaining on the planet as possible (within reason), since those ships will come after your own planets when you destroy their home station -- I've literally lost, just when I thought I was near to winning, in that fashion.

Once you are certain you are safe from a counterattack (keep an eye on Threat), then move your ships and your production facilities over to that second planet, as you were thinking.  Just don't expect for any one strike group of yours to take out the entire enemy home planet at once unless you are really lucky and they are caught really off-guard.

That's the other thing: assuming that the home planets are not on alert, hopefully they only have a few hundred units there.  Even so, that will quickly grow to maybe a thousand in a short amount of time once you tip them off, if you are not constantly whittling them down there.  If you keep them on alert for any extended period of time, you make your job exponentially harder as they will quickly climb up toward 4000 or 5000 ships there.  

Anyway, so lightning-fast strikes against the guard posts, with new waves coming immediately after your first ones are destroyed to fill in the ranks, should do it.  It might turn out to be a really easy encounter for you if you've taken so many planets on the map and have so many ships built up, etc.  On the other hand, it might not.  Be prepared for the worst, and if you get lucky -- great.  But still be prepared for the worst when you then approach that second AI home planet, too.

[Tim James]

Hmm, I hadn't thought about having to reinforce my strike teams. Definitely sounds like I'll have to do one after the other, but the problem is after taking all these planets I still have 6 or 7 hops to get to each homeworld. Do people just build up a reinforcement group and use it to slowly make their way through the wormholes, destroying all the enemy reinforcements that get filled in behind? It's going to be a bit of a grind going through 15+ lightning turret groups with each wave of reinforcements.

This wasn't so much of a problem in the intermediate tutorial!

[Lancefighter]

Also keep in mind, if you can remove as many of the guard posts as possible from range (preferably at 20k with dreadnaughts..) then you are removing an area of reinforcements. I'm not surely sure if the reinforcements are merely redicted towards the other ones, but it DOES mean that you wont have to fight on so many sides.

[x4000]

Haha, yeah, the tutorial definitely was a bit easier in that regard!

I tend to like to set up a base on a planet two hops away from the AI Home planet.  Then I can just sit there and crank out more guys to send over.  If you pair that with transports, you can deliver a lot of your units almost undamaged to the AI home planet, and make lots of quick waves.  Using transports can keep the alert level low until the last second, too, which is another help.

[x4000]

Also keep in mind, if you can remove as many of the guard posts as possible from range (preferably at 20k with dreadnaughts..) then you are removing an area of reinforcements. I'm not surely sure if the reinforcements are merely redicted towards the other ones, but it DOES mean that you wont have to fight on so many sides.

Yes and no.  Having fewer guard posts does cause the others to reinforce faster, but it also does cause them to have a lower overall ship cap for the planet.  So it's a net gain in the end, but you may feel some pain in the interim -- as you say.

[Tim James]

I have to admit I gave up. I discovered I am up against an entrenched homeworlder and a turtle. That does not sound fun to do battle with in the final push toward their home planets.

In both my first and second campaigns I learned some things that would make future games more enjoyable and less frustrating. I think in my third game I'll put all that to use and have fun, but for now I need to table it before I get annoyed.

I also hope when I get back a few things are fixed and rebalanced, like stacked-up enemy turrets that prevent me from targetting certain things, fewer or weaker lightning turrets, and slower reinforcement/remains rebuilding. But mostly I need to apply what I've learned (take fewer planets, use more turrets, don't play turtles, get dreadnaughts for those lightning turrets, etc.)

Thanks for the game.

[x4000]

Well, glad you are enjoying it overall.  The entrenched homeworlder is kind of the basic AI that leans a little defensively, and of course the turtle is super defensive.  So, that would be a tough fight on the second one there, at least.

The lightning turret balance hasn't been changed in a while to my recollection, and is something that most players seem pretty happy with to my knowledge -- though, I will say that the turtle uses WAY more turrets in general than normal, and the entrenched homeworlder uses perhaps slightly above average.  So that may be what you were running into there; given that the turtle never attacks you, you gotta figure they have loads of extra stuff at their planets.  You might prefer playing against raiders, or mad bombers, or something of that sort -- attacking their planets is vastly easier, but you also get pummeled pretty regularly.  If you are good with your defenses, it can make for an exciting game.  If not, it can make for a shorter game.

For the turret stacking thing, that was fixed in 3.011 or 3.012, although it wouldn't retroactively fix older saves that had them stacked.  Though, a potential fix for that with at least some existing savegams just occurred to me, too, so I'll check that out.

The reinforcement rates, also, are based on two factors: how many planets you take, and the type of AI.  They are by far the highest with the turtle, and slightly above average with entrenched homeworlder.  Again, sounds like those fortified AIs are just not your bag.  The remains rebuilding... I think a delay might make sense for those, the more I think about it.  Would make things less hectic with the AI, anyway, which is always a good thing.  I'll make a note to look at that for the next beta version.

[Tim James]

I'm definitely enjoying it overall. Part of me wants to start another game but I have a backlog to get to. AI War will be up pretty quickly for replay though.

It strikes me that the game is a lot like a roguelike. You need to lose a few times in order to win. My first game I was stuck behind three Mark IV planets that had been on alert for hours until I figured out what was going on. In this game I was having a blast until I ran into a wall. The randomness means you occasionally encounter situations that an experienced player can navigate (and even enjoys doing so) but are too daunting for a new player, just like some status condition or difficult monster in a roguelike.

You might consider either playing up the fact that you have to try a few games (like Dwarf Fortress does with "losing is fun") or including an option to tilt the random number generator toward a balanced game for newbies.

The last parallel with roguelikes: I think the game is more fun without save scumming. There were some tricky parts that I tried to force my way through with quicksaving. But having an entire planet or army wiped out means you can come back smarter next time instead of just pounding on it over and over.

[x4000]

I'm definitely enjoying it overall. Part of me wants to start another game but I have a backlog to get to. AI War will be up pretty quickly for replay though.

Oh sure, fair enough.  I understand about lists like that!  I have on a mile long myself.

It strikes me that the game is a lot like a roguelike. You need to lose a few times in order to win. My first game I was stuck behind three Mark IV planets that had been on alert for hours until I figured out what was going on. In this game I was having a blast until I ran into a wall. The randomness means you occasionally encounter situations that an experienced player can navigate (and even enjoys doing so) but are too daunting for a new player, just like some status condition or difficult monster in a roguelike.

You might consider either playing up the fact that you have to try a few games (like Dwarf Fortress does with "losing is fun") or including an option to tilt the random number generator toward a balanced game for newbies.

That's very interesting!  I had never thought of that sort of comparison before, between AI War and roguelikes, but you are absolutely right.  There is already a way to make the generated maps easier, though -- the difficulty of the AI themselves causes fewer or more upgraded planets, that sort of thing.  A lot of people like to start out at difficulty 7 and lose before winning, whereas others start out at diff 5 and seem to have a much, much easier time of it.  The in-game recommendation for starting difficulty is:

" - If you are an experienced RTS player, but are new to AI War, you might want to start with difficulty level 5 or 6 (and many people don't like anything below 7, there is a big jump in intelligence there).

 - Once you are familiar with the mechanics and general strategy of AI War, difficulty level 7 is considered the standard level (again, a lot of genre vets find the AI pretty lame below this level, and awesome here and up)."

But, via the forums (and probably the wiki, I suppose), there's a lot of pressure to try 7 because that's where the "real game" is.  Which is true, but that pushes it more in the roguelike direction.

The last parallel with roguelikes: I think the game is more fun without save scumming. There were some tricky parts that I tried to force my way through with quicksaving. But having an entire planet or army wiped out means you can come back smarter next time instead of just pounding on it over and over.

Oh yeah, I'm the same way.  Unless I'm testing, I generally just accept whatever consequences and move with it.

[RCIX]

The last parallel with roguelikes: I think the game is more fun without save scumming. There were some tricky parts that I tried to force my way through with quicksaving. But having an entire planet or army wiped out means you can come back smarter next time instead of just pounding on it over and over.

Oh yeah, I'm the same way.  Unless I'm testing, I generally just accept whatever consequences and move with it.

Never! Save Scumming FTW!

[Black]

I save scum when I am learning a new tactic. After my first mistake, I don't reload if I repeat it.

To answer the question: The easiest endgame is the fastest endgame. I tend to stop expanding after 9 planets, and then start hunting for an advanced factory. Once I have that, I locate the core worlds and plan a route to them that meets the least resistance. I always try to kill the core planet before it can reinforce, because it receives truly murderous reinforcements.

Attrition is the warrior's greatest enemy. Siege is the commander's folly.

[Tim James]

To answer the question: The easiest endgame is the fastest endgame. I tend to stop expanding after 9 planets, and then start hunting for an advanced factory. Once I have that, I locate the core worlds and plan a route to them that meets the least resistance.

Do people move their force from one wormhole to the next without touching enemy planets, or do they wipe out planets in a path up until a hop or two away from the core planets, then rebuild the force for the final attack? In my game I wasn't sure about this tradeoff between AI Progress and being stuck on an island.

[Kjara]

It really depends on the person I'm sure .

I've been known to strike an ai world from 4-5 or so planets away if a game is going well.  In a closer game I'll generally jump to a planet 2 away, take it and rebuild there.  Most of my planets are usually unconnected though so an extra one or two doesn't bother me.