Looking for advice/tactics

[Salamander]

Hello folks,

I’ve been playing AI War since 2010 but never on very difficult settings and quite sporadically despite how much I enjoy the game and shamelessly plug it to people often. I feel I am competent but have a lot of room for improvement so I’d like to request some advice from the community on certain things. I’ve been reading a lot of the articles in this section but I’d like to touch on where I feel I need the most improvement.

I was playing a 5/5 game, yeah I know not exactly difficult and I’ve beaten it on higher difficulties before but I’m trying to get back into the groove after not playing much lately, and was doing alright. I was playing an ‘X’ galaxy with my champion, 4/10 resistance fighters, medium golems/botnets, medium spirecraft, Fallen Spire, spire civilians, all ships enabled (barring the CSGs), traders, and my bonus ship was the Autocannon. I had located both the homeworlds but had stayed back to keep the nasty chains to them from being alerted while re-fleeting and preparing to try to burn through without alerting them. I was still in the middle of rebuilding my fleet after dealing with a 5.8k wave (AI progress had just hit III) when 15k youngling commandos were announced and they came in/with several carriers and various starships in my 2nd of 2 exposed and reinforced planets. I tore them up badly but not enough as they killed the forcefield and command station and while they would have ripped through the planet behind them, despite fairly heavy defenses there, they would have then slammed into my main hub/defense world at the very center of the galaxy and died there, the damage to my fleet was done and the time it would take to refleet and rebuild would have kept me in a permanent descending loop so retired from that game and started a new one.

If I’m ever to beat the game at higher difficulties I obviously need to improve my game. In asking for advice I will focus on where I feel weakest but those that are doing well on the higher difficulties can also see where I can improve and any advice is appreciated.

-I need to improve at ‘gate raiding’ and colonizing non-contiguous systems: I’ve avoided the isolated colonies since it’s difficult to reinforce them or react to Counter attack posts (not that I’ve seen them in these lower difficulty games), but I waste time and AIP when I close off areas via conquering every single planet in a chain. I know that if you eliminate the ability for the AI to gate in ships adjacent to a planet you won’t get direct waves but you can still get cross planet waves unless you deal with all the systems in that isolated chain so not sure how best to approach this one. You can use your own gates to directly port in reinforcements from elsewhere but some ships, especially spirecraft which I enjoy employing, lack those options.

-Using varied command stations: This is where I really need to stop staring at my income of metal and energy like they are the be-all and end-all of existence and use stations beyond just the basic economic one. My new 5/5 X game I started at the central hub and unlocked Military 2 already to hold the more distant areas, will likely unlock 3 later but want a warp jammer to deal with the terribly nasty NE branch that has 4 IVs in it but no homeworld. But also has some things I want to grab without taking the whole branch as I point out in my previous point.

-Research: This one is one of the fundamental aspects of the game. When to research what is one of the most critical decisions we can make in game and drives us to conquer/neuter/hack worlds for the research we need. I rarely research hack, but it reminds me of the old days of research raiding. I tend to focus on my offensive and mobile forces over everything else; but I do focus early on Scout II, Engineer II, and my harvesters but I was reading that many do not get the harvesters up fast?  Each game is different but some things are universal and scouting and engineers are always vital. I like having harvesters upgraded quickly but perhaps more focus on defenses early instead of the harvesters and offenses?

-Transports: Yeah I admit it I don’t use these enough. The regular transports are limited in their ability to attack by design and I rarely spend the research to unlock the Assault transports, but I have when I had too, like when assaulting Eyes, but I need to use these to lessen my losses and of course take isolated worlds per my earlier point.

-This one is mostly about exo-waves and specifically the thread on the Botnet Golem at: http://www.arcengames.com/forums/index.php/topic,16046.0.html  . I take it the Botnet Golem is more a super-weapon than other golems based on this thread, and this in turn makes nastier responses from the AI, but what’s the threshold? I rather like the Botnets, despite not being able to control the zombies. In that game I mentioned that I resigned from my one Botnet Golem had killed almost 5k enemy ships but if it’s not worth the trouble I can turn it off. I do like to use a lot of different ships of all kinds but as I still need to improve anyway it may be good to turn it off, at least for now.

Those are things I know I need to work on and I’m sure there is a lot more. We could make volumes on tactics, strategy, and the like for this game but other things I’m pondering include using Mercenaries; I love them once my economy is rolling but they were giving me issues in that game I mentioned trying to make good my losses especially the merc parasites with massive metal bottlenecks. I also tend to be light on the warheads although I will use them when I feel I need to. Any other advice would be appreciated and thanks for reading!

[Toranth]

(AI progress had just hit III)

There's yer problem...

On difficult 5, the AI won't hit tech level III until 750 AIP.  If you are finding and destroying all the Data Centers and CoProcessors (which should be high-priority, if you aren't already doing it), that means you accumulated roughly 1000 AIP total.  That's 50 worlds captured.

If you take some screen shots or saved games and post them here, we can get a better idea of what you are doing.  But just from that one quoted phrase alone, I think the biggest step you can take to improve your gameplay is to reduce the number of systems you capture.  You should only be taking those systems that are necessary - aka, they contain a Core Shield Generator you've yet to destroy, a Factory IV or Advanced Starship Constructor, and ARS, a Fabricator, are related to a minor faction, or have some major strategic significance.

Whenever you are thinking about capturing a system, ask yourself "What goal do I intend to accomplish by capturing this system?"  Then ask yourself if there is another way to achieve that goal.  For example, gate raiding an AI system prevents that system from being a source of waves.  It also prevents it from being a source of Exowaves.  IIRC, it also prevents the CPAs from spawning in that system.  And gate raiding costs 5 AIP as opposed to the 20 AIP that a capture requires.
If your goal is to capture a unique (A Golem, Fabricator, etc), ask yourself - Do I need that resource to win the game?  Also ask, do I need it right now?  Putting off AIP increases until late game means more time fighting against a weaker AI, always a desirable situation.

Again, if you have any saved games that you are willing to post, especially from earlier on in your game, put them up, and there will be people willing to look at them.
Otherwise, I suggest you read some of the After Action Report threads in the AAR forum.  There are some very good players that post detailed games there, explaining both the 'what' and the 'why' of their wins.

[keith.lamothe]

Yea, I agree with Toranth:

1) Good grief that's a lot of AIP   Various changes since 7.0 have made moderate-AIP (say, end-game in the 400-500 range on a game where you haven't gone far into FS) more viable, rather than requiring low-AIP, but hitting 750 AIP in midgame is still really asking for it.

2) For us to give good specific advice, a savegame (and a screenshot of the current galaxy map, at least, for those not downloading the save) is very helpful.  You can attach a file directly to a post under Additional Options when making/editing a post.

Anyway, glad you're finding the game enjoyable; I hope you will find it more so as you learn more

[Salamander]

Thanks guys. I already overwrote the save with my new game since I just tweaked a few settings and started over with a new seed but AI progress is crucial to success as you point out.

There's no question I'm taking too many systems, you're definitely correct there. I know not to take them all, I always tell those I'm teaching how to play not to take everything, but I am certainly taking too much. I turned auto-progress off for my new game, I normally leave it on, and usually win with the AI progress at II and not III. It was 750 or 770 for that game for III and I had just hit it on a recent attack.

The game was probably not the best anyway. I had 3/4 of the galaxy scouted, only lacked 100% from not being able to get through certain chokepoints, and found zero advanced factories which hurt a lot. I had taken all 'my' SW branch and the NE branch, the homeworlds were 6 and 7 jumps from the central hub in the NW and SE branches and I was going to clear my way as fast as I could, but I had gotten the progress too far to proceed despite earlier successes and lots of reclaimed ships and was stuck in that attrition war as I posted previously.

My biggest issue I think is when you find a dead end of systems, let's say it's an X map and it's a branch with 9 systems sprouting off of it and only 3 are worth taking, I tend to take them all to simplify my defense situation. In that game I only had 2 egress points and both were pretty heavily defended but in one the game's greatest challenges, if you push the AI too far, you can't defend so well even then. It's a conundrum I need to resolve so that's my biggest issues I think. But if you don't conquer/kill the AI command, you can't get the research either (barring hacking which has its own price) so it's quite the poser. It sounds like the best idea in this case would be to grab the 3, defend the heck out of them now with the system defense caps instead of the old global ones and neuter the surrounding AI via gate raiding and then hack when it was viable.

Continuous improvement is my goal and I like reading how people handle the difficulties the game can throw at us, part of why even with breaks I always come back to the game, a game I'm glad I helped with in the past (like the Armor Rotter!) and hopefully more in the future as well. 

[Toranth]

Just as a mental exercise, try this:
Start a new game with Complete Visibility.
Pause the game as soon as it loads.
Using the Galaxy Map, put Priorities and Notes on all the systems you intend to capture, detailing how important that system is and why.

If you feel like it, you can post that, or just use it for your own sake.  It's much faster than actually playing the game out to scout everything, and more useful for learning purposes.
A good, but much longer, second phase is to then actually play out the game.  See if your plans change, and take note when something unexpected becomes important, or something you thought was important turns out not to be.

[Kahuna]

You don't need to hold more than 10 planets.

Just as a mental exercise, try this:
Start a new game with Complete Visibility.
Pause the game as soon as it loads.
Using the Galaxy Map, put Priorities and Notes on all the systems you intend to capture, detailing how important that system is and why.

If you feel like it, you can post that, or just use it for your own sake.  It's much faster than actually playing the game out to scout everything, and more useful for learning purposes.
A good, but much longer, second phase is to then actually play out the game.  See if your plans change, and take note when something unexpected becomes important, or something you thought was important turns out not to be.

I actually do this every time I start a new game.
I do it because I play exclusively on 10/10 difficulty level and a crappy map can ruin the game or just make it annoying. And as you said for learning purposes/mental exercise. It helps me to become better at decision making and knowing what works in which situation. Nowadays my plans work basically every time though they may change a little bit during the game. But the basic idea works. Also that's how I climbed from 6/6 difficulty level to 10/10 difficulty level.

[Peter Ebbesen]

I don't see anything wrong either in principle or in practice with hitting AI progress III in either the mid- or end-game unless you are running a super-high difficulty setting like Kahuna and a few other veteran players, and from your OP, you aren't doing that.

While traditionalists might argue that you are running too high AIP, I have to disagree.

From your post, you are committing the cardinal sin of not matching your capabilities to your strategy: You are going on a conquest spree and investing in offences and mobile fleet assets over everything else rather than in defences, and that's really poor strategy in a world where the enemy's aggression response grows with your success at conquest and your static defences are expected to counter most of the enemy's aggression.

Since the AI doesn't rebuild destroyed infrastructure, it can be whittled away at your leisure so long as your defences hold, and the key on the defence is to be able to hold without requiring a major fleet presence for regular waves. (Your tempo affected by any automatic AIP/minute you've set up and civilian spire leaders, but unless you've set up a really nasty situation for yourself in this regard, you've effectively got all the time you need.)

So EITHER learn to run lower AIP while accomplishing your goals OR learn how to properly defend yourself - or better yet, learn both, and choose a style of your own that is a mix of the two and suits you best.

-----

With regards to learning how to play at lower AIP, Toranth gives excellent advice. How much is truly necessary to take? Very, very, little.

With regards to learning defences, I'd suggest you try a learning game where you start out by unlocking Mil2 and Mil3 stations, HBCI, Scout II, and then get the Neinzul and Leech starships with your first knowledge conquests, after which you only invest knowledge in your fleet in case your existing fleet proves incapable of dealing with problems you encounter on the offence, and invest all other knowledge in defences or the economy.

This should result in a game in which you get a much better feeling for how little offensive power is actually required to be successful, while at the same time giving you a better appreciation for just how strong it is possible to build static defences and allowing you to experiment with different combinations.

Once you've got that knowledge, you should be much better prepared to playing a normal game where you focus more on your offence and less on defence than in this learning game, but are able to tailor your level of investment in both and in the economy as well to match the threat you actually face.

-----

Regarding Harvester research.... Do the math on Economic station upgrades vs. Harvester upgrades. You'll thank me for this. :p

Or rather, thank the veteran players who argued long and hard for the Economic station to be changed to make the upgrades actually useful. The result is that IF you have planets that are not on the frontline where you can afford to plop down Economic stations rather than Logistic/Military, they are a much better investment for Metal~Energy purposes than Harvesters until you've conquered a lot of territory.

And if you don't have any such non-frontline planets, then you probably have only a minimal territory and little knowledge to spare and need to seriously consider whether either upgrading Economic stations (for the automatic one at your Home) or Harvesters are really worth the knowledge cost in the first place, given that you've got salvage as an additional metal income source. There are situations where improving Harvesters is worth it at the moment, but as a general rule improving Harvesters before improving Economic stations would require a really weird map and expansion strategy to make sense.

[Kahuna]

Thanks to the new turrets (per planet caps) now every planet can be a whipping boy. And thanks to the new Carriers Laser Turret are is most important Turret at the moment because it counters Carriers (Heavy) and Bombers (Polycrystal). And if you unlock Spider Turrets you have 3 caps worth of AI Guardian (Medium) countering Turrets (Mark I Missile, Sniper and Spider Turrets). Spider Turrets work very well with Mark >= II Military Command Stations. Or you can unlock Mark II Missile or Sniper Turrets for 4 caps worth of Medium countering Turrets (A cap of Mark II Turrets is equivalent to 2 caps of Mark I Turrets). I've noticed caps of Mark I Missile, Sniper and Spider Turrets ~2 shot a Mark I Spire Starship.

(Once again.. the top 5 hull types to counter are: Polycrystal, Artillery, Medium, Heavy, Light/Structural)

Check out this .sav file. Over a 1000 Bombers are attacking Ascs 26-I. I've placed all non radar dampening Turrets under a Force Field to protect them from AI Zombards (Zenith Bombard) and in front of the Command Station to tank damage. You can disable the Force Field if there's less than ~30 Zombards and less than ~100 Structural countering AI ships in range of the Command Station. Radar dampening Turrets are not protected by the Force Field to lure Zombards to come in their range. When the AI Bombers get in range of the Command Station re-enable the Force Field to tank more damage. Now I know there are no Zombards in this wave but I've positioned the defenses like that because the AI has sent like 200 Zombards a couple of times and I thought I should explain why the defenses are the way they are.

Also select all Engineers and Remain Rebuilders and position half of them next to the Command Station but not inside the Force Field. Make sure to use attack move (Hold X and right click). You can position these next to the Force Field protecting the non-radar dampening Turrets and disable/enable the Force Field depending on situation.

Position the other half in the front line under the Force Fields. One quarter next to the other Tractor Turrets and one quarter next to the other Tractor Turrets.

And yes.. even little things like positioning Engineers and Rebuilders correctly can make the difference between victory and defeat. Actually.. positioning Engineers and Rebuilders correctly isn't a "little thing". You can even fully repair all Force Fields and the Command Station while the Plasma Siege Starship is reloading. So you can tank it indefinitely. Just order like 20 Mark II Engineers to repair all Force Fields and the Command Station by holding Shift and clicking all of them.. and as soon as they get a chance they will insta repair all of them in middle of the fight.

Visualization:


It was the Rebuilders and Engineers rebuilding and repairing the Tractors and Force Fields that kept me alive. I tried scrapping all Engineers and Rebuilders and I got overrun.



Also not that these defenses were not optimal and this wave was quite easy/small.

[Salamander]

Thanks all!

I started a new SP game and now am on a concentric galaxy with an amazing setup, I found an ARS two hops from home and an advanced factory right after that! I also found a chain of 3 systems in a line that were all poor prospects so am gate raiding them instead of taking them.

In that game in my first post my defenses were actually thick in the 2 systems, I had just beaten 2 waves of roughly 6k IIs and the 15k III wave was knocked down to less than 4k before they capped the station, but yeah I need to invest in the stations and am in fact doing just that in my single player game and a new 2 player game I started. I had several turrets unlocked to 3, mini-forts, zenith modular fort, the spider V controller, etc. but had the progress been less or I had better stations...

I'm trying to teach my friend that never played above 4 how to gate raid and such in a new cluster game we started. Already have eco 2 and military 3 stations unlocked there. I'm also using the P0-9 markers to lay out places for us; P9 is something we need to skip and gate raid, P8 has been gate raided, P0 has stuff we really want etc. I am trying to get him to beef up defenses, slapping down a total of 20 turrets and 5 minefields is not going to work! He's only played with me on lower difficulties and I showed him my border post with every single turret I can put there plus mines and the 2 miniforts, plus some spire ships.

I do leave extra engies and rebuilders in any system that is exposed. Even a couple make an enormous difference.

I appreciate all the advice guys. I've played a little over 300 hours but I tend to fall in ruts a bit too easily. I will keep reading, testing, and improving!

[Salamander]

Finally finished the game I mentioned here with just under 20 hours invested into it and got a few new achievements to boot (capture 35, lose 30,000 ships (  ) , etc.). I used the ratings ala Toranth's suggestions after scouting out everything. I love the Mk IV Scout for its ability to get intel pretty much anywhere. I did leave Fog of War on for this, I had the Mk IV go on a walkabout to find out most of this intel.



P0s were planets I wanted to grab or had to to use their resources (i.e. Golems). I had taken most of them at this point but decided to get in there and finish the AIs off in what one considers a hurry for AI War.

P1s are all gone but they were somewhat desirable, not as much as P0s.

P3s are nebula I have not done yet. I was working on them but stopped to push to the end. I used raiders to take out Sentinel Posts and then would take my Champion (in low power mode) with Cloaker Starships to the planets I wasn't going to take to finish those scenarios off and then came back the same way.

P5s are either skippable or hackable planets. I did a lot of hacking this round.

P9s are planets I didn't want to grab for whatever reason, and related to that, P8s were planets I didn't want but had gate raided. In retrospect the one P8 that is otherwise completely surrounded I should have just taken out but was testing AI progress as well so let it sit.

P7s are the homeworlds. The one P6 was just a planet I left fallow after getting it's research to give me a buffer zone from the first AI homeworld.

The P7 at the bottom with ships still in it was of course the Bouncers Homeworld, I had already taken out the Shield Ninny's by this time. I used a big EMP on the IV to it's left and blazed past the defenses to hit the homeworld, it was a mean planet and not worth taking. I did save before the attack just in case that move flubbed but it went pretty well.

I had already purchased Harvester upgrades before it being mentioned here but went ahead and grabbed the military and economic stations and the warp jammers. I didn't get the logistic ones. They have some nice advantages but then prior to this game I've never spent research on the stations. I love the military 3s, put them on high pressure border systems, and the eco 3 ones are lovely for resources but I'm not too happy with the Warp Jammers. I knew they didn't provide a 'seal' on their borders but when I left not only a couple of alerted but previously raided systems open as well as two non-raided borders to influence the AI attacks but they still kept hitting warp jammers, I was a bit confused. I knew that if you left them with no gates nearby they could attack anywhere, which is why you try to influence where the waves arrive via selective gate raiding so why they would hit the WJ systems instead of the open ones was perplexing to me and forced me to layer on more defenses there than I had originally planned. Luckily after Counter-Attack Posts were added (and I remember when they had a 1-minute or so warning only!), I put at least some defenses everywhere, but having to add so many to the WJ systems kind of defeated the purpose of using them.

I only used 3 EMPs the whole time. I used to use more warheads but was trying to minimize the AI progress until I was in the end game phase. I was able to hit all the co-processors, capture all but one of the Spire Civilians, although eye bots killed one of them under a FF, hit all the data centers, and even used the superterminal briefly. I had to kill that last one fairly quickly since it was in the same system as a Spire archive and even with it under a FF I didn't want to risk losing it.

The AIP would have been exactly double what it shows here (1064), I was quite happy with my reduction this time as I did take a lot of systems, even with paring them down and pushing quickly at the end there. Not quite to IV, but pretty close had I not taken the spires and every single AIP reducer.

My bonus ship was the Vorticular Cutlass but I ended up with more bonus ships than I ever had before, partly from lots of hacking and selective conquering but also from playing on an 80 size galaxy where I usually play 60s.

Thanks for all the advice, I just started a new 6/6 game with the Space Tank as my bonus ship and the random AIs are Bouncer (again) and Tag-Teamer but I just started and so far I have grabbed Military 2 and 3 stations, Scout 2, and Engineer 2, not the harvesters like I used to always do. I've beaten the game before but I knew I was close to the edge of the abyss when starting to move the difficulty up, even 4-5 is a big difference I've found. Need to play tighter and more focused as the AI becomes less forgiving.

My defenses are similar to Kahuna's although I don't use the FFs on the turrets as much as I should it appears. I do use minefields to a lesser degree and have always found them very handy. My question to you Kahuna if you read this thread still is: I'm curious as to why you specifically indicate some engies  outside the FF, but I am unsure as why this is specified? I normally hide all the engies and rebuilders under the domes but you obviously have a reason for that placement and it intrigues me.

On a side note, I'm really enjoying the hacking system.. I hacked 3 or 4 systems for research, several design backups (some I DLed, some I corrupted), fabricators, etc. I burned through most of my Hacking Progress as the game went on. It adds another wrinkle in the way it works backwards from our AIP so it requires a balancing act, which I love for games like this. Every decision should have weight and this makes you ponder what to take vs. what to hack. 

[Kahuna]

Finally finished the game I mentioned here

Con grats

grabbed the military and economic stations

Both are very good unlocks. If you want you can choose Munitions Boosters as your Bonus Ship for the attack damage boost and then only unlock the Economic Stations. You will get attack damage boost and economic boost and you will save Knowledge. You wont have the Military Stations' Railguns and Engine Damage though.
Just a suggestion/idea. I really like it works for me. Just a few Munitions Boosters will boost dozens of Turrets/Ships.

My question to you Kahuna if you read this thread still is: I'm curious as to why you specifically indicate some engies  outside the FF, but I am unsure as why this is specified?

It's because of Plasma Siege Starships. If the Engineers are under the Force Field the splash damage will destroy the Engineers in a couple of shots. If the Engineers are outside the Force Field they're more likely to survive longer even if the AI has some long range ships. The AI ships usually prioritize ships and turrets over "civilian" targets like Engineers.
Of course if the AI doesn't send Plasma Siege Starships it's better to keep the Engineers under the Force Field.
EDIT: The Engineers in the front line are always under the Force Field because.. well.. they're in the front line. But the ones in the back are relatively safe outside the Force Field.

[Salamander]

It's because of Plasma Siege Starships. If the Engineers are under the Force Field the splash damage will destroy the Engineers in a couple of shots. If the Engineers are outside the Force Field they're more likely to survive longer even if the AI has some long range ships. The AI ships usually prioritize ships and turrets over "civilian" targets like Engineers.
Of course if the AI doesn't send Plasma Siege Starships it's better to keep the Engineers under the Force Field.
EDIT: The Engineers in the front line are always under the Force Field because.. well.. they're in the front line. But the ones in the back are relatively safe outside the Force Field.

That definitely makes sense and as I slowly turning the difficulty back up I see exactly why you set them up that way. I started that new 6/6 campaign and remember this was the highest difficulty I beat the game several patches ago. Starships appear in normal waves (including of course the aforementioned plasma siege) and the Eyes are showing up. I have already unlocked the eco and military stations to 3, scouts to 3 (and just got an advanced factory), and FF2s, DLed the Gravity Ripper and my first ARS gave me teleport battle stations. I remember when eyes were not so big and tied into the grid like they are now, used to send in Spire Penetrators to blow their doors off but that got patched pretty quick!

The system with the advanced factory was a IV with an Eye and the last station was not only a Spire shield station but it had an FF and a shield guardian parked there so after some raiding on the other 4 posts (cloaker starships and low powered raiders and bombers are so handy!), I sent in an EMP and 2 loaded transports (1 was just a bit too small) and unloaded several ships close to the thing to take it quickly, then with the eye gone I was able to bring the rest of my fleet.

I'm noticing more IIIs and IVs since I last really played. Not used to seeing so many and it's another factor I have to account for that I'm playing more of the game again. All the advice here and elsewhere is greatly appreciated as always.