Author Topic: Recommended research prioities and safe AiP limit?  (Read 2508 times)

Offline KillerofGods

  • Newbie Mark III
  • *
  • Posts: 29
Recommended research prioities and safe AiP limit?
« on: April 21, 2015, 09:41:56 AM »
Hey! Just picked up the game, played the tutorials and started my first game. I'm about two hours in but I'm curious as to what you guys recommend for good research pickups (as well as what phase in the game you generally pick them up) and some to avoid/are more niche.

Also what is the recommended levels of AiP that you should try to avoid going over? Obviously you want to keep it as low as possible but it'd be a good benchmark to keep aware of... games can be really long so I just want to avoid stupid mistakes and get into the higher difficulties sooner, right now I'm just messing around in my low difficulty game to get me a good idea as to what to expect in the various phases of the game in a 80 planet map.

Thanks for any and all advice! Happy gaming! :D

Offline Pumpkin

  • Hero Member Mark III
  • *****
  • Posts: 1,201
  • Neinzul Gardener Enclave
Re: Recommended research prioities and safe AiP limit?
« Reply #1 on: April 21, 2015, 10:13:13 AM »
Remember this is opinions only. Other players may disagree with me.

Knowledge

For fleetships and starships, I recommend not unlocking MkIII before having MkIV via advanced facto/SSconstructors. Paying 4,000K for just MkIII is expensive, but 4,000K for MkIII and MkIV is fine.

I almost never unlock high Mk turrets, but I have a low AIP style of play (see below). I prefer focusing on fleetships or starships, depending of my mood. Support unlocks come in special cases, like fortresses for a turtle style of play (exowaves, Fallen Spire or aggressive AI types) or decloakers if the AI often sends cloaked waves. Adv transports, Zenith space-time manipulators, mobile docks, all that is fun or situational (of both). Also, unlocking at least one more Mk of Force Fields is mandatory if you tend to expand your territory.

Some special turrets can be useful, like the fearsome Heavy Beam Canon (HBC) or the useful spider turrets. Gravitational turrets and special mine fields are tricky to use but powerful once you got it. See Kahuna's guide for some hints.

AIP

The AI has a tech threshold. Below, it sends MkI waves, above, it sends MkII. Then it has higher threshold for MkIII and IV, if I remember correctly.
Consider 100 AIP before the first homeworld onslaught quite low. I Often end between 150~200 but I'm quite an "assassin" player, taking as few planets as I can. Maybe on diff 7/7 300 AIP is manageable with "just" a good defense and some K spent in turrets and energy.


Hope that helped. Take some other opinions; remember I'm just one player.
Please excuse my english: I'm not a native speaker. Don't hesitate to correct me.

Offline Mal

  • Full Member
  • ***
  • Posts: 110
  • Murder Time! Fun Time!
Re: Recommended research prioities and safe AiP limit?
« Reply #2 on: April 21, 2015, 04:00:59 PM »
Hey KoG,

Welcome to AI War! Pumpkin brought up some good points, I have a bit more to add.

Concerning Research Pickups

  • Counters- The most important thing to spend knowledge on is counters to what the AI is pumping out for their strategy. If you find yourself against a swarmer, then unlock more riot ships/zenith missile starship. Always find the most painful thing about the AI you are about to face and spend knowledge to counter if possible.


  • Defense- The second most important thing is to be able to hold the important systems like your Home and strategic assets like factories. Spend knowledge on the most important turrets, not all of them, just mostly to counter what the AI is using. 


  • Offense- When you have your intelligence and defense complete, then start working on winning the war. Build a task force that can crack each system. Never rely on the same setup for every system, as most systems are different. Anything you had from your last mission that you cannot use on your next move, scrap it unless it is prohibitively expensive to rebuild, this will get you moving faster.


  • Economy - Economy largely depends upon playstyle. If you favor fleets with some starship support, you can get away with surprisingly little economy. If you favor starships or turrets, you will need to pickup a lot of economy upgrades. As you take more planets, this becomes much less important to spend knowledge on. This is why it is the least important concept in the game.


Concerning AIP Levels

  • AIP determines New Reinforcements- Every so often, the AI gets reinforcements determined by a formula that involves Difficulty of AI, AIP. As AIP gets higher, it gets proportionately more reinforcements. When these reinforcements get too high, the AI releases waves, CPA's, and the like.

  • AIP determines Tech Levels- As certain thresholds of AIP are crossed, the tech-level of all reinforcements gets bumped up by a level. This can make fights you were winning a draw or a complete loss, this also forces you to spend knowledge just to keep up to the AI's level.

  • AIP means more at higher difficulties- The AIP reinforcement formula gets much more fierce at higher difficulty levels, especially at Diff 10. So just keep this mind if you play at higher difficulties that the same amount of AIP in a Diff 7 is much deadlier in a Diff 8, 9 , and 10 game.

  • High AIP does not mean you will lose the game - Don't be too paranoid about high AIP. In certain late game situations, once the AI releases a CPA, it can mean the loss of a game if you don't throw everything under the sun at it ( including nukes). But otherwise, just always be calculating that the AI is gonna get much more defended and act much quicker as the AIP rises.



Most of all, build everything you can.
« Last Edit: April 21, 2015, 04:25:49 PM by Mal »

Offline Red.Queen

  • Full Member Mark II
  • ***
  • Posts: 191
  • Mad Hacker
Re: Recommended research prioities and safe AiP limit?
« Reply #3 on: April 21, 2015, 07:45:50 PM »
+1 on the welcome to the forums and the game, Killer.  Mal and Pumpkin have covered a lot, but I'll throw in a few bytes as well from the perspective of a new-ish player (~80-90 hours played).  Be advised I have a taste for high auto-AIP games with lots of time pressure and consider warheads to be a perfectly acceptable standard response, rather than a pure emergency option -- this is generally not recommended, I just do it anyway.  :)

I got a bit absorbed writing this as I kept thinking of things that might be useful (and have still left out a billion other things), so brace for incoming brain core dump!


Thoughts on Research

Starships vs. Fleetships -- I would recommend deciding fairly early in a match which route you are going to favor and sticking with it, so you can gear the rest of your unlocks to complement it well.  From what I've seen so far, a "jack of all trades" fleet where you have a little bit of everything at Mk.I-II doesn't do nearly as well as a more focused core of Mk.III-IV ships of your choice, supported by expendable, auxiliary stuff at lower marks.  What to unlock varies hugely on what's available in fabs/ARS/design backups, but one constant is you will need some serious bomber power for the homeworld assaults at the end, and frankly it's pretty important in the midgame as well for cracking all those Mk.III and Mk.IV worlds.  Whether that comes in fleetship or starship form is up to you.  Starship vs. Fleetship style has an influence on your non-offense unlocks.

Economic Unlocks -- Heavily affected by your fleet style, and the composition of the map.  Starship-oriented fleets are ungodly expensive to run, and are even worse to rebuild if they wipe.  Economy should get some upgrades if you go this route, or if you go the Kahuna-style "God of Defense" option, as unbreakable turret emplacements also are expensive.  If you identify a finger of the map that can be sliced off and isolated behind a chokepoint, consider unlocking higher tier Economic Command Stations and building an Industrial Zone.

Harvester upgrades pair well with Military and Logistic stations, as they let you squeeze more metal out of any given planet, while letting you have stronger defenses on it.  A lot of players like the Military stations, especially as they have wide-area Tachyon coverage for enemy cloak breaking (if the AI ends up rolling a lot of cloaked ships, you are really going to want these) and boost the damage of defenders on the planet.  I don't have experience with Logistics stations, but there are some players who do amazing things with these due to their mobility boost and ability to block enemy teleporting.

Lastly, Warp Jammer stations are special -- they cost you resources instead of give you them, but any planet with one on it can't have waves sent to it (but can be targeted by Threat and CPAs, don't know if they are targetable by Exos yet).  It will also not put neighboring planets on alert.  Good for when you need to not wake up a very dangerous neighborhood (like a Core world or AI Homeworld), or want to drastically reduce the odds of a planet with something vitally valuable getting attacked (Spire Archive before it's finished extracting, rescued Rebel Colony, crucial factory/constructor/fab).

Defense Unlocks -- How much K you can sink into this depends on how much you're spending on your fleet.  So not only do you need to balance Starship vs. Fleetship, but Fleet vs. Fixed Defenses.  I second the recommendation to check out "Kahuna's Guide to AI War" here on the Strategy board for an in-depth discussion on this, it's pretty much the unofficial user manual to the defensive side of AI War.  Even if you are a more fleet-oriented player and thus can't sink as much K and resources into defense as he does, his approach scales down very well.  You can also use turrets offensively to do something called "beachheading" -- I've never done this, but Kahuna talks about it in his posts, and it can work very, very well.

The main general tip I can add is that the Core Turret Controllers are very, very powerful.  If you are an extremely fleet-oriented player, I would strongly recommend making capturing or hacking these a high priority, as they dramatically boost your defenses without chewing up K that can then be spent elsewhere.  They are also very efficient -- turrets all cost the same amount of energy to run, but do tremendously more damage as you go up the Mark tiers.

You'll start getting a sense of just how much you need to defend at a given time/difficulty/AIP level as you play.  There isn't a hard and fast formula, so as you're learning the game it can be a good idea to periodically test your defenses.  Let a wave come, and hold your fleet back -- note what its total strength point value is.  Watch how far the enemy gets.  If it dies fast and does little damage, you've got plenty of headroom in defensive capability, probably don't need to allocate more resources to it for a while.  If you notice certain problem units getting dangerously close to causing trouble, examine what types of defenses counter it, and improve your defensive positioning.  You may need to unlock something new, or upgrade something you already have.  If the wave as a whole pushes too close, it's time for a larger-scale upgrade of your turret unlocks, and may be time for another tier of FF too.

Just a warning, but certain ships can bypass FFs in whole or in part, and require special attention.  Raid Starships, Eyebots, and Infiltrators can fly right through them, for example, while Zenith Devastators can shoot through them without any loss of damage, and Plasma Siege starships' shots do "splash" damage through them.  The latter also just do enormous amounts of damage in general, so that splash damage is not trivial at higher marks.  They also knock down FFs fast if they get within range, so never, ever let them.  High marks can actually one-shot some command stations!

Another special consideration are long-range artillery type ships, like Zenith Bombards and Zenith Siege Engines.  These things can sit back outside of range of many standard turret configurations and pound on your fixed defenses with impunity -- and Zombards at least are Sniper Immune (don't remember for the ZSE).  They also do a crapload of damage, but reload slowly.  You need to either bring in your fleet to deal with these, or force them to move forward -- their AI has a kiting component, and they will try to move away from anything that can damage them.

You can exploit this even with turrets -- throw a few short range turrets on the side of the wormhole *opposite* from your command station. The Zombards will enter the system, say "holy sh$%, there's an enemy right behind me!  Better move out of range" and move forward -- towards your main emplacement.  If you are fast, you can nail them with your other turrets (manual targeting) before they realize they just flew into range of even more dangerous stuff.  You may find this tedious, and it is difficult to set up if you have a very long distance between the wormhole and your main turret emplacement, so you will probably want to Design Corruption hack these away from the AI.

Last point on defense unlocks -- sometimes they can end up feeding your offensive unlocks.  Certain ships are modular, like the Spire Corvette.  Certain turret unlocks do double-duty of letting you install more powerful modules on those ships.  FYI, only researched unlocks affect this, not captured turret controllers.  No Mk.V modules for free.  :)


Thoughts on AIP

Pumpkin and Mal covered the majority of this, I'm just going to add a few thoughts.

AIP and Time -- In many ways, AIP is time, and time is AIP.  This is very literal if you play with auto-AIP enabled, but still is true to a lesser extent due to waves, CPAs, and especially the escalating Exos that start once you attack a homeworld.  As AIP climbs, your projected lifespan falls.  You have quite a bit of control over when significant trades of time for AIP happen, and I find that timing these trades can be just as important as the AIP total itself.

Trade in a big chunk of your lifespan too early, and you may have doomed yourself -- you essentially said "I bet you that I can win the game by X deadline" when you bumped the AIP up a significant step, and the game will hold you to that gamble.  It's safest to make these big trades towards the end, where you need to survive for only 30-90 minutes, vs. weathering escalating onslaughts for 3+ hours.  Especially as AIP increases tend to breed AIP increases, either by blowing up more gates to control waves, taking more time fighting fires when you have auto-AIP on, or using more warheads more frequently.  You can see where this goes.

However, as Mal said, don't be overly-AIP shy, especially on non-insane difficulty levels.  AIP boosts also tend to come with boosts to your own power, in terms of captured planets, more K, more metal/energy, etc.  If you're too weak, you won't stand a chance against the homeworlds at the end, and will eventually get ground down by CPAs and waves.  Also, the homeworld defenses/strategic reserve assume a floor of 200AIP, so coming in under that at the end means you are probably leaving fleet power on the table, unclaimed.  Knowing exactly when to spike the AIP late game as you go all-out on offense is something that you'll learn with experience -- and exactly when and how is something that varies from player to player depending on their style.

Right, But I Still Want To Know How Much Is Too Much! -- Well, Pumpkin chimed in with their perspective on AIP around 7/7 difficulty as a low AIP style player.  I've seen other players say that 300-400 at that difficulty is reasonable, for a medium AIP perspective.  I'll add my thoughts from the higher AIP style end of the spectrum (higher, not extreme -- that's ShruggingKhan and his team).  I was able to successfully win a 7/7 game closing at 519 AIP, and it wasn't too rough.  I am currently posting up a completed 8/8 game **SPOILER ALERT DO NOT LOOK IF YOU DON'T WANT TO KNOW HOW IT ENDS**
Spoiler for Hidden:
that I was able to win that ended on a staggering 936 AIP.  Don't reveal until I finish posting it please, some of my readers want the suspense.  :)

Final Thoughts

The forum's loaded with useful tips to help improve your play.  I personally found reading AARs to be especially helpful, as people tend to document what they unlocked, when, what the AIP was, etc, and you get to see how a game flows from start to finish at different difficulty levels and minor faction/plot combos.  Some people have high-detail writing styles and will describe their tactics and long-term strategy.

The current regular denizens of the "should be impossible" 10/10 world are Kahuna (extreme defense) and Faulty Logic (extreme offense) -- if I'm forgetting anyone, I don't mean to slight you, chalk it up to a memory glitch from working on this post for far too long.  I believe RockyBst has a recent 10/10 or two as well -- he also has some hilariously unorthodox games posted where he plays with the weirdest combo of settings, which is a fun way to see just what some of those things actually do.  Wanderer and Diazo also used to play at or around that level, but haven't posted in a while -- still well worth reading even if some of the fine numerical/mechanical details no longer apply.  Peter Ebbesen also did a very entertaining 9/9 AAR showing the Fallen Spire campaign mode.  Mind you, 10/10 is its own world, and requires extreme AIP management to a level that is overkill at low/moderate difficulty, so don't read those and think "oh man I am at 120 AIP, I can't win, better restart" when playing at a sane setting.   :)  Besides, sticking out games that have started to go sideways and finding a way to win is an excellent way of learning a lot *fast*.  Just keep plenty of save games at different points in your play so you can roll back and try different things if you want to -- no sense in replaying the very early game if you have it down and are focusing on trying to study the mid or end game, unless you are intentionally playing an Iron Man run.

There are also plenty of AARs in more normal 7-9 ranges out there.  Pumpkin, Mal, _K_, Alex Heartnet all have recent 7-9 games up, and, shameless self promotion, so do I. (I'm documenting my attempt to climb the difficulty scale, starting with my first serious game at 7/7 and going from there.)

Happy hunting!
« Last Edit: April 21, 2015, 07:50:21 PM by Red.Queen »
Infiltrating hostile AI networks to rewrite reality.

[[Hacks available from this unit found on the AI War Modding subforum.]]

Offline Tolc

  • Full Member Mark II
  • ***
  • Posts: 161
Re: Recommended research prioities and safe AiP limit?
« Reply #4 on: April 21, 2015, 08:10:59 PM »
Just a warning, but certain ships can bypass FFs in whole or in part, and require special attention.  Raid Starships, Eyebots, and Infiltrators can fly right through them, for example, while Zenith Devastators can shoot through them without any loss of damage, and Plasma Siege starships' shots do "splash" damage through them.  The latter also just do enormous amounts of damage in general, so that splash damage is not trivial at higher marks.  They also knock down FFs fast if they get within range, so never, ever let them.  High marks can actually one-shot some command stations!

Oh, god, yes. I have that in my current (and first) game (started last september, on hiatus since october, need to get back to it!). It really is a sinking feeling when you are cruising with your fleet somewhere, watching the number of attackers on a planet of yours dwindle rather rapidly... Till it's stuck at a count of 1 or 2 and you start to think "That's weird...I'd better have a look..." only to find one or two plasma siege starships shooting your FF and the command station benaeth it to pieces...

I'm not the OP, but as a noob myself: Thanks for all that info!

Offline KillerofGods

  • Newbie Mark III
  • *
  • Posts: 29
Re: Recommended research prioities and safe AiP limit?
« Reply #5 on: April 21, 2015, 10:38:24 PM »
Thanks for all the comments everyone!!!  :D

I got exactly the info I was looking for!

Final Thoughts

The forum's loaded with useful tips to help improve your play.  I personally found reading AARs to be especially helpful, as people tend to document what they unlocked, when, what the AIP was, etc, and you get to see how a game flows from start to finish at different difficulty levels and minor faction/plot combos.  Some people have high-detail writing styles and will describe their tactics and long-term strategy.
Just discovered that forum late last night and was having lots of fun reading your threads! I'm gonna be checking out other threads from that forum later. If other people go as in detail as you do (even if they don't do the narrative which is pretty nice) then it should be a joy to read.  :)

PS: On a side note when neutering a planet do you destroy the wormhole guard posts as well? Do they allow the AI to stack up an additional 250 ships per post?

Offline Captain Jack

  • Hero Member Mark II
  • *****
  • Posts: 808
  • Just lucky
Re: Recommended research prioities and safe AiP limit?
« Reply #6 on: April 21, 2015, 11:00:42 PM »
They really aren't worth devoting attention to, they've way too much health and don't do significant damage. Either you take the planet and destroy them, or take scratch damage as your fleet passes through to its actual destination (or hide your fleet under cloakers and take no damage).

Offline KillerofGods

  • Newbie Mark III
  • *
  • Posts: 29
Re: Recommended research prioities and safe AiP limit?
« Reply #7 on: April 22, 2015, 12:37:03 AM »
I remember reading on the wiki that the ai has a cap to the amount of ships it can use to reinforce a planet, I read that this correlates to the amount of guard posts (plus the command post) on the planet. I have two M3 planets in my cornered off portion of the galaxy, I destroyed the wormhole guard posts as well because I don't know if they contributed to the cap.

I'd prefer to not have to destroy these in the future as they just have so much health and it'd be a waste of time if they don't contribute to the ship cap. Potentially 500 ships at my backdoor I can deal with but potentially 1750 (250 for each CP and each guard post) is another story. So I figured better safe then sorry.

Edit: Oh, how do you get the game to tell you the AIs strength on a planet while you have a scout there? Also is there a way to see what a ARS will give you before you take over the planet?
« Last Edit: April 22, 2015, 12:55:31 AM by KillerofGods »

Offline Red.Queen

  • Full Member Mark II
  • ***
  • Posts: 191
  • Mad Hacker
Re: Recommended research prioities and safe AiP limit?
« Reply #8 on: April 22, 2015, 01:07:53 AM »
You're welcome KillerofGods (and Tolc, I laughed at your "what's that one ship -- ooooh crap" scenario), and I'm with Watashiwa on the wormhole guardposts.  I just ignore 'em, and every so often roll the fleet through to kill the nominal reinforcements they produce.  I call it "mowing the lawn."

Regarding ARS-peeking, this is doable and very useful -- if it lets you find out about a certain ship that's vital to your style, it can have a major impact on how a match goes.  All you have to do is get a human-built Science station onto the planet (moving a captured ARS from one planet to the new one won't count, it will still show you its own ARS tab).  Mk.I or Mk.II works.  Don't worry about it surviving -- you only have to get it onto the planet, not right up to the ARS, so just move it through the wormhole and pause the instant it shows up in the unit list on the righthand side of the screen.  Then click on the Science station and take a look at the new ARS tab it shows.  Now you can make an informed decision about whether you want to do the (kind of expensive) Advanced Research Redirector hack.

For seeing AI strength on a scout-picketed planet, there's a filter in the Galaxy map.  Switch the first filter to Mine and Allied Ships and the second to All Mobile Strength, and you'll be able to see how much grunt its (and your) fleet has on a planet.  I assume there's some way to get it to show the value of fixed defenses/guardposts but for some reason All Units/All Strength doesn't seem to do what I expect.  Someone else probably knows what's up with that.
Infiltrating hostile AI networks to rewrite reality.

[[Hacks available from this unit found on the AI War Modding subforum.]]

Offline TheVampire100

  • Master Member
  • *****
  • Posts: 1,382
  • Ordinary Vampire
Re: Recommended research prioities and safe AiP limit?
« Reply #9 on: April 22, 2015, 01:18:01 AM »
I may be one of the few that relys more on economy as other people. I prefer to unlock engineers very early in the game just because I want to rush my productions. I also heavily defend all my planets even the ones that aren't so important. I was very happy when the galaxy wide cap of turrets was changed with a planet wide cap instead. Made my playstyle easier.
If I would be an AI I would be Turtle: Just burrow in and wait for the attacks.
I'm also not offended by bigger AIP sizes, I can deal pretty good with it.

Offline Mal

  • Full Member
  • ***
  • Posts: 110
  • Murder Time! Fun Time!
Re: Recommended research prioities and safe AiP limit?
« Reply #10 on: April 22, 2015, 03:58:31 PM »
I remember reading on the wiki that the ai has a cap to the amount of ships it can use to reinforce a planet, I read that this correlates to the amount of guard posts (plus the command post) on the planet. I have two M3 planets in my cornered off portion of the galaxy, I destroyed the wormhole guard posts as well because I don't know if they contributed to the cap.

I'd prefer to not have to destroy these in the future as they just have so much health and it'd be a waste of time if they don't contribute to the ship cap. Potentially 500 ships at my backdoor I can deal with but potentially 1750 (250 for each CP and each guard post) is another story. So I figured better safe then sorry.

Edit: Oh, how do you get the game to tell you the AIs strength on a planet while you have a scout there? Also is there a way to see what a ARS will give you before you take over the planet?

Wormhole posts do count for reinforcements. There is a cap to reinforcement totals but then the AI just begins to upgrade their guards to higher marks. So they still keep getting reinforcements even at cap. Time is the enemy.

Offline Pumpkin

  • Hero Member Mark III
  • *****
  • Posts: 1,201
  • Neinzul Gardener Enclave
Re: Recommended research prioities and safe AiP limit?
« Reply #11 on: April 23, 2015, 02:38:40 AM »
I remember reading on the wiki that the ai has a cap to the amount of ships it can use to reinforce a planet, I read that this correlates to the amount of guard posts (plus the command post) on the planet. I have two M3 planets in my cornered off portion of the galaxy, I destroyed the wormhole guard posts as well because I don't know if they contributed to the cap.

I'd prefer to not have to destroy these in the future as they just have so much health and it'd be a waste of time if they don't contribute to the ship cap. Potentially 500 ships at my backdoor I can deal with but potentially 1750 (250 for each CP and each guard post) is another story. So I figured better safe then sorry.

Edit: Oh, how do you get the game to tell you the AIs strength on a planet while you have a scout there? Also is there a way to see what a ARS will give you before you take over the planet?

Wormhole posts do count for reinforcements. There is a cap to reinforcement totals but then the AI just begins to upgrade their guards to higher marks. So they still keep getting reinforcements even at cap. Time is the enemy.

I sometimes test my Fallen Spire's fleet firepower on them.
But if you are worried about reinforcement on a neutered planet, consider putting a defensive sniper beachhead on it to spawn-kill this reinforcement.
Please excuse my english: I'm not a native speaker. Don't hesitate to correct me.