Poll: Should AI homeworlds be buffed? If so, How?

[Martyn van Buren]

I am really against anything that would make the homeworlds harder --- the truth is, I find them pretty hard already.  I could always go down in difficulty if they were buffed, but I'd rather not; I'm pretty happy with the difficulty of the midgame on the difficulties I'm playing now (7.6 if I want to win, 8 if I don't).  I have never really done baiting, which could help, but I agree that that oughtn't to be more mandatory than it is now.

If we're concerned mainly about baiting, would it be possible for the homeworlds to have a "defenders floor", such that if they have less than a certain number of units freed ships aren't allowed to leave the system?  If it were set at the right level (I guess rising with AIP) you'd still get a nasty threat response when attacking a heavily-reenforced world, but it wouldn't be possible to clear too many of their ships without fighting them on the homeworld itself.

I can also see it being a good thing to make it more difficult to finesse the 100 AIP jump by basically neutering one world and then leaving the command station alive til you've killed the other.  I don't have a good enough sense of the game to be sure what's the best way to do this, but that seems like something that would trouble me less than things that would make killing an individual homeworld harder than it is now.

[Hearteater]

I want to add to MvB's comments on the Defender Floor, that I think the low number of AI units on the Homeworlds in a low AIP game is probably the biggest contributor to them being easy.  If you get to a homeworld and it has 200-300 units, it isn't going to be very difficult.  Maybe there should just be a flat floor that causes the homeworld to get reinforced when it is below that value.  I don't see any reason it shouldn't be at least 400-500 at all times on 7/7, and possibly quite a bit higher on 10/10.  This would have the effect that low AIP strats are made tougher, but high AIP go nuts games don't really get any harder.

[TechSY730]

I am really against anything that would make the homeworlds harder --- the truth is, I find them pretty hard already.  I could always go down in difficulty if they were buffed, but I'd rather not; I'm pretty happy with the difficulty of the midgame on the difficulties I'm playing now (7.6 if I want to win, 8 if I don't).  I have never really done baiting, which could help, but I agree that that oughtn't to be more mandatory than it is now.

If we're concerned mainly about baiting, would it be possible for the homeworlds to have a "defenders floor", such that if they have less than a certain number of units freed ships aren't allowed to leave the system?  If it were set at the right level (I guess rising with AIP) you'd still get a nasty threat response when attacking a heavily-reenforced world, but it wouldn't be possible to clear too many of their ships without fighting them on the homeworld itself.

I can also see it being a good thing to make it more difficult to finesse the 100 AIP jump by basically neutering one world and then leaving the command station alive til you've killed the other.  I don't have a good enough sense of the game to be sure what's the best way to do this, but that seems like something that would trouble me less than things that would make killing an individual homeworld harder than it is now.

This sounds very similar to the perma-defender idea I reposted on Mantis some time ago.

This, combined with a higher initial game seed rate and a higher chance of reinforcements once ship count (factoring in caps) falls below N ships, would be one possible way to implement a "defense floor".

[Martyn van Buren]

It certainly is! Upvoted

[Faulty Logic]

I don't think that the HW defenders should ever be freed. So they would all have the "perma-guard" tag.

[Kahuna]

Other: (No need to add to the poll anymore since it's so old)

AI Sensor Array (Defensive Raid Engines)
This could be a new AI type, AI plot, minor faction(?) or just a new building that is seeded when the map is generated. In my opinion AI Plot would be the best option.
AI Sensor covers the planet it's built on and all adjacent planets. It would also have planetary tachyon coverage on the planet it's built on. Player ships entering a planet covered by the sensor would cause all planets covered by that sensor to instantly spawn double reinforcements, to have the amount of reinforcements increased by 100% (Temporary (only when alerted) Intergalactic AI Troop Accelerator) and to go on alert. In addition to spawning normal reinforcement ships it would also spawn some Starships. All ships spawned by the Sensor would be normal guards and would defend the planet they were spawned on. Sensor reinforcements would be separated from the normal reinforcements and would have a cooldown similar to Raid Engines OR normal reinforcements. So entering, exiting, entering, exiting an AI planet covered by a Sensor would not make the Sensor spawn reinforcements every time. Like a defensive and less aggressive version of Raid Engine. If planet A has Sensor X and adjacent planets B, C and D. Player entering any of those planets (A, B, C or D) would cause all of them to get the reinforcements, to go on alert and get double reinforcements. Planet E with Sensor Z and adjacent planets F, G and H would not be alerted. Sensors could be placed on Core planets too.
Destroying a Sensor increases the AIP by 2. Sensors don't have weapons. They have Structural hull type, 6 million health, 2000 armor.

Note that these Sensors wouldn't be only on Core planets. They could be seeded anywhere in the galaxy.. on Core planets too.

Updated http://www.arcengames.com/mantisbt/view.php?id=9114

[Wanderer]

Yeah, I know, I'm late to the party.

I want to do a mechanics review of what can and can't make howeworlds trivial (comparitively).  Bits and pieces of this were mentioned everywhere, but in the end the mechanics will drive the experience.

Homeworlds as they were (and apparently, still are, just a bit beefier for guardians and posts):

Rarely reinforced due to not being on alert by careful management, leaning towards smaller #'s of defensive units (maximum around 500 on normal unless significant mistakes were made)
A series of smaller guardposts with only critical combinations forcing non-standard blob assault tactics.
A final assault on the command center requiring a Bomber Assault team because of the double ff + fortress III.

So, really, it's just another planet with fatter guardposts and bad luck with fortress+forcefield and a nasty AIP increase on removal.  Alright.

How do homeworld assaults actually threaten you?

Boredom: You get jammed up in an alert-reinforcement-bait-refleet cycle. You make some headway eventually but it can take forever.

Tandem Raid Generators:  Usually your fleet by homeworld assault time will be able to handle a single raid engine, even a MK IV one, unless you've gotten way out there with AIP.  The threat is in tandem generators.  The new CPAs might be hairy too but at least you get to use your prepared defenses in that counter.

Stagnation: The typical death knell of an AI War game... no traction.  One too many mistakes, one too many planets, loss of a critical MK IV factory... any of these events can cause a game stagnation.  It's not a win/loss scenario, it's a judgement call on the part of the player that they can no longer move forward with any of their current tactics.

Only one of those methods actually KILLS you.  Overran fleet and 3xRaid Engine (even at 50 AIP) can lay waste to your defenses, cause blackout cascades due to threat separation, and any other number of things.

Most of this discussion is about making it even MORE stagnant.  Oh gods, please no.  I'll get to that later though.  I'd like to stay on mechanics for the moment.

So, what tools do we have that can counter the stagnation:
Cloaked Nukes: This is uneffective, as everything at MK V ignores nukes.  Coreworld + are non-nukable.  It DOES deflate tandem Raid Engines for four minutes though if you keep your troops out of the way.

Rest of the missile sets: EMPs: Ignored by MK V.  Tachyons: Errr, why?  Lightnings: Now we're talkin'.  Bring ten to twenty of these bad boys along for a modest AIP increase and you can lay waste to the entry port + any significant combination threat (FF + Beam outpost will cause me significant problems PRE-Buff, particularly if an AI Eye is rolled in the brutal phase). 

What else do we have:
Penetrators: One shot every 30 minute ships that can pretty much level any one thing at choice.  Breaks the tedium on a problem child.  Non-standard game.

Artillery golems: Mixed usage, a lot ignores them.  Non-standard game.

Jumpship dropoff: Limited quantity of ships, best used with hard hitting, low fleet cap ships.  SSBs, Spire/Zenith SSs, etc.  Non-standard game.

Spire flagships: This is no longer about the homeworlds, this is exo-wave vs. supership, the HW's are candy.

Suicide Champs: New and unknown effectiveness for balance for me.

Transport assaults: Mixxed bag.  50/50 success rate.

Bait and clear: This will heavily depend on the brutal rolls for the AI if it's even an option.  Bait and clear against a triple Raid Engine is not viable until you've cleared the engines.

So that pretty much covers all the non-grind options.

Why do the raid engines make such a difference compared to the rest of the brutal outposts?  AI Eyes are actually much worse in the long run, and those are instant defenders, and they can't easily be removed, they're rather well guarded!  Because you lack discreet control.  Pissing off a Raid Engine immediately provides x defenders to a system 1 minute after arrival.  In most circumstances you won't even reach the Raid Engine on trigger before the defenders have arrived, unless you've done a transport run which wouldn't survive unless you've already gone in and done a bait and clear on the majority of defenders.  Throw in a second engine and you have no serious chance to stop those defenders with just fleet.

And in four minutes, more are coming...  and if you didn't stop the first series, that second series is going to wipe the floor with you.

Compare this to the AI Eye.  Bait and Clear: you fire up what, 20 zombies to piss off half a planet to fight elsewhere?  Once that's done, your riots and raids can do cleanup in the system (preferably under cloak so they can wax the Mass Drivers first) with only a few zombies spawning in retaliation... which are easily cleaned up.  Low Fleet Cap ships are the direct counter to the old Eyes.  I'm ashamed to say I haven't messed with the new ones much.

All this does is slow you down to not being able to fleet ball.  It makes the final bomber assault on the command center a royal PITA because your bombers will get murdered repeatedly in base game, but that's about it.

So either the Homeworld has Raid Engines and is deadly as hell, CPA engines (with the new build and barracks) which MIGHT be deadly, or is just a frickin' grind.  Eventually, you neuter the thing, and start working on the other homeworld for a double pop, because there is no good reason to soak 100 AIP in the meanwhile.  None... well, except that ONE game where the AI HW had the SuperTerminal.  Put spire archives on HWs with no AIP loss on death and then maybe you might have something... but assaulting the mandatory brutal pick homeworlds with an extra 100 AIP can be a death sentence.

So, that's what the mechanics look like right now, at least to me.  I assume Kahuna and Faulty Logic who are playing hot potato for 10/10 god rights may have a different outlook, but I don't believe it's that much different from what I've read in their AARs.

Now, a lot of what I've read in this thread just makes it more grindy.  You're already 10-12 hours into the game, having 'grinded' your way through superior reinforcements and other things to have the power to assault the homeworlds.  A bait and clear tactic gives you a massive 'winner takes all!' kind of brawl, then it's cleanup.  I'm sure I don't want more grind.

So, grind basically means it's uninteresting repetitiveness.  Otherwise, it's not 'grind', it's fun!  So, some of what could be done (encapsulates a few different ideas in each one, and no, I wasn't completely thorough):

1) You want to remove cloaked cheese without player effort on the homeworlds?  Give every post its own tachyon emitter.  The AI has littered the entire frickin' galaxy with these damned things on EVERY wormhole.  It can obviously afford a few more.

2) The idea of a "calling all hands" response to a HW assault sounds interesting.  I definately could see that for SF forces and the surrounding core worlds.  However, this implementation needs some care taken.  It would be extremely easy to sheer off a single cap of fighters in a few transports, fire up one HW with a calling all hands, and then go hit the other one.  Now you get to deal with all those ships as threat with your defenses and it was a pointless exercise.  It also allows me to VERY easily neuter every coreworld with a little pause manipulation with some kiters on the homeworld.

3) Defense cap minimums.  These are VERY dangerous unless you WANT to constantly refleet.  Between the heavier reinforcement mechanics and now the lack of being able to fight the majority of those ships outside of a coreworld alert you're not making the fight on the homeworld more interesting, you're making it inanely grindy.  This mechanic will also very easily stagnate the homeworld assaults.  An example: One of my AARs has me trying to take out a single Raid Engine against a mostly cleared homeworld.  A 20 flock of transports with 10 bombers a piece only got off 2 shots before getting laid to waste by the leftover starships and slight reinforcements.  I did take it out with another full wave of missile frigates (which had the correct bonuses) but even they didn't survive long.  If the HW always had 300+ MK V ships, these base tactics would be nearly impossible, or would require nearly your entire fleet for a single base strike, with little expectation of survivors.

4) Extra brutal picks: Enforce variety here so you don't end up with inane combinations (ie: 4xRaid Engine) and I could live with that with the new ones coming into play.  It'd keep it interesting.  You'll still always end up targetting the Raid Engine first.

Everything else is generally a buff across the boards to the AI leading towards a stagnation grind.  The AIP split on even entering an AI HW makes some sense to me, as AIP is lorewise the concern of the AI.  That these humans are even in my house are enough of a reason to up my game. 

Right now the homeworlds seem to me to be the only place we actually use our guerilla tactics.  It doesn't makes sense in most other cases.  Those guerilla tactics however are what appear to be unbalanced because of attempts to make them more attractive.  The penetrator and jumpships are primary examples of this.  One shot kabooms are also valid for homeworlds because they are the end-game.  You're not going to be suffering the consequences of AIP increase from warhead use for long.

I'm not convinced the homeworlds need to be harder.  I'm really not.  Most 10/10 games will die long before encountering the homeworld other than with a scout to see if it's even playable (for me anyways).  A 9/9 game is not supposed to be a rediculous slugfest there either.  It's the icing on the cake.

I want to see a throwdown, a massive winner take all kind of fight when I get there.  I've grinded my arse off to build my fleet, kill your shields, blow up your support, now I want to KILL you dammit... not spend the next 3 hours trying to punch you in the nose while you hold me back by the forehead like some gradeschooler.

So... no... in the end I don't think they need to be buffed.  Tweaked for difficulty consistency?  Yes.  I should expect a difficulty skewed brawl when I get there, and maybe have to try completely different tactics and guerilla myself a hole first.  One homeworld shouldn't be a walk in the park while the other one is Judge Dredd's court.

Dissect, abuse, etc, as needed. 

[TechSY730]

(omitted for length)

And here is the classic issue. How can we make the AI more compitent at defense, but not make the game more grindy?
Competent, dedicated defense almost implies grindiness.

Two things you mentioned.

One, I would like to see the AI be smarter overall (both in a macro sense and in a fleet tactics sense), thus making guerilla tactics more necessary for the rest of the game. Not so much that it makes the whole game a micro grind fest, but enough so such that the homeworld fight doesn't feel quite as "out of place".
EDIT: Also, fighting an enemy that is harder because it is smarter is generally more fun than fighting an enemy that is harder because it has more.
Two, Keith already mentioned what you alluded to, the penalty for a failed homeworld assualt, barring raid engines, is not all that large. I hope this means that Keith is going to be doing something about this.


I would like to say though that if we humans are able to stack ludicrous numbers of defenses on our homeworlds, why shouldn't the AI?

And for your off handed comment about tachyon guard posts, 9078: Reduce initial game seeding rate of tachyon guardians, which would necessitate the AI being smarter about how to deal with cloaked threats.

[keith.lamothe]

Competent, dedicated defense almost implies grindiness.

Unless the sorts of things you have to do to defeat that defense are fun

[Faulty Logic]

So, really, it's just another planet with fatter guardposts and bad luck with fortress+forcefield and a nasty AIP increase on removal.  Alright.

Agreed. That was what I wanted to change.

Cloaked Nukes: This is uneffective, as everything at MK V ignores nukes.

I disagree: nukes kill whatever mkIV fleet ships are there (usually one or two hundred) kill the eye if present,  and knock out the fortress III by putting it out of supply.

Suicide Champs: New and unknown effectiveness for balance for me.

Technically possible, but now that champions are immune to transport, unutterably boring.

a different outlook

Yes, on page three are my arguments for buffing the AI HWs. I think the HWs are too easy, even against a 10 turtle.

I certainly agree that grinds, or worse, stagnation are undesirable. I just don't see these buffs leading to stagnation, only a climactic fight.

Finally, welcome back.

[Kahuna]

I assume Kahuna and Faulty Logic who are playing hot potato for 10/10 god rights may have a different outlook, but I don't believe it's that much different from what I've read in their AARs.

I'm not. As I said in another thread I'm taking a break from playing AI War.

Good thing Guild Wars 2 was released when it was. I didn't think it would be possible.. but I'm getting a bit bored of AI War. According to Steam I've played it 1143 hours. Some of that is afking ofcourse. But I've played the last ~700 hours of AI War without playing any other game every day when I'm not studying. Imma buy Guild Wars 2 tomorrow.. and maybe FTL: Faster Than Light (Available: 14 September 2012. This game will unlock in approximately 1 day and 3 hours). I'll take a break and play Guild Wars 2 for a while.