AI Homeworld Defense

[Gudamor]

So to restate the problem: due to their random placement, the Core Guard Posts are in a situation where they need to be balanced for "difficult enough to pose a challenge when sitting by itself" AND "not impossible when seeded close to another Post, or two, or five... and under a forcefield."

Diazo's plan would be to force the game into always being the latter situation, and nerfing the posts down to a level where that is balanced.

Some of the other plans would have the game forced into the former situation, where they are either spread out further from eachother on the Homeworld, spread onto Core Worlds so that the Homeworld guardpost density is less, or nerfed to have lower ranges so that overlapping occurs more rarely.

Of these, I think the lowered ranges would be the least Developer Time, but I think spreading them onto the Core Worlds would be the most fun.

[Vacuity]

Having followed the thread with great interest I feel a desire to chime in.

For whatever my opinion's worth, the idea I like best is Martyn van Buren's: not nerfing, but separating things up, so the core guard posts and brutal picks also get seeded on the core worlds as well as the homeworlds, with the core worlds being a bit easier than the homeworlds: brutal picks get placed on homeworlds first, and the "regular" guard posts on homeworlds are a bit tougher than the "regular" posts on the core worlds, but brutal picks would never get placed on the same world as another brutal pick unless there was nowhere else for it to go: i.e. three brutal picks in the RNG, one homeworld and one core world, so two brutal picks on the homeworld and one on the core world.  Very unlucky, yes, but not as impossible to deal with as it would be with the current system.

This may be more time-consuming than just lowering the ranges of everything, but I think would retain the desirable interest in dealing with the hard counters (blanket tachyon coverage, wrath lance, teuthida (I've never actually had to face a teuthida, the only games they got spawned I ended up losing before I got there anyway...)), but not allowing them to completely overwhelm all options.  The planet-wide hard counters are not such a bad thing after all, it's just overwhelming to meet them all on a single planet.

Now, this means that on maps where you're likely to have just one core world for each homeworld it will still be quite tough going, but that is partly a matter of choice on the map type.  If you choose a crosshatch map, you've got other problems to deal with, if you've got an X or spokes map, you've likely got fewer other problems to deal with, so your endgame is tougher.

I would personally prefer to retain the mechanic whereby the core guard posts have to be eliminated to attack the AI homeworld command station, so they cannot be bypassed completely, but as they are split up across worlds (at least two, but possibly more depending on the map setup), it's less of a chore trying to take out each one.

Allowing a bit more of the strategic reserve to be deployed in defence of core worlds might not be a bad thing either, particularly with a slightly lower rate of deployment.

The end result is that the core worlds are quite a bit tougher than present with a proportion of the core guardposts, a little more of the strategic reserve assigned to them and a possible single brutal pick, but the homeworlds are significantly easier than at present with much lower odds of multiple brutal picks, at most half the current number of core guardposts (though the "missing" ones are still getting replaced with very high level regular guardposts) and the option for the player to deplete the strategic reserve via the core worlds a bit more than is presently possible.

In game terms, the result is something like having one or more mini-bosses before the final boss, but the final boss itself is easier than it currently is.

Edit: Put another way, you retain the fun of working out a plan, executing it and either succeeding or failing, but massively reduce the !fun! of dealing with near impossible combinations.

[KDR_11k]

My thought on the matter is this:

Bosses are supposed to be final exams, a test of everything you have learned so far. However an epic open battle isn't what you learn in AIW and it isn't very fun in an RTS because it's your troops doing the fighting, not you.

What you learned is how to operate on a minimum of assets to accomplish your goal, strategically deciding when to strike and when to run. What to take out because every action you take makes the AI stronger.

I'd suggest removing planetary tachyon coverage entirely (no brutal pick with it either) from the HW and instead spawn loads of core sentinels. These are of course tachyon emitters and they come in varieties like static, mobile, perhaps a handful of cloaked ones, etc. Destroying a sentinel increases the AI's acute alertness (maybe the size of the strategic reserve or something) so again like taking planets you want to keep it to a minimum.

The idea is that you pretty much MUST stay cloaked as long as possible, avoiding the sentinels like they are guards in a stealth game. If that's too restrictive perhaps there should be some form of automatic cloaking on the AI HW or maybe some device you can capture that gives your units cloaking on the HW? Once you uncloak you have the strat reserve and all kinds of other nasties rushing for your troops.

Perhaps the strat reserve could be changed to appear from the edge of the gravity well and maybe some "reserve warp gate" structures that again can be destroyed at the cost of making the AI angry but getting rid of them makes the strat reserve take longer to arrive.

[LaughingThesaurus]

I feel like making the homeworld more of a cluster rather than a stealth mission would cater better to that though. A stealth mission's just going to annoy me and AI War isn't explicitly about stealth either. The idea is, you make strategic decisions to carefully gather the strength you need, then smash the AI homeworld with a decently sized hammer that isn't really nearly big enough to take it on in a straight-up fight. However, in the main game of AI War before the homeworlds, you end up circumventing AI chokepoints, picking apart the AI bit by bit, attacking where it's weakest and taking things that increase your power overall.
To that end, if the core/brutal guard posts were distributed among the core worlds, you have to pick apart the home and core worlds which will make everything a bit easier to fight overall. Imagine it this way; Each core world has a bunch of guard posts and all that, and each one contributes an extra weapon to every other core world and the homeworld. You find the place where the AI is weakest, and you will be able to steadily raid and pick apart the layers of enemy defenses, bit by bit.
If not that, then imagine it a bit of a different way. Each guard post acts a bit like the 'secondary' shield generators. Meaning, you might find Core Laser Guard Post B in some really impossible to attack places on some worlds. However, one is relatively in the open, and easy to assail, on a core world you have access to. You found the boss's weak point, blow it up, and all of those other Core Laser Guard Posts on that frequency will also die. These are limited to only the core worlds and the homeworld, of course, and they will allow you to actually pick apart the AI's defenses. Brutal posts might be represented by Prime posts on each core world. Destroy the prime post on each core world, and the homeworld's brutal post goes bye-bye. Once you've picked apart all the layers, the home command station is ready to be destroyed. What this also allows for, both of these ideas, is the idea to turn the idea of guard posts being in impossible positions into more of a game mechanic, rather than a frustration.

[Draco18s]

So to restate the problem: due to their random placement, the Core Guard Posts are in a situation where they need to be balanced for "difficult enough to pose a challenge when sitting by itself" AND "not impossible when seeded close to another Post, or two, or five... and under a forcefield."

Diazo's plan would be to force the game into always being the latter situation, and nerfing the posts down to a level where that is balanced.

Which is, of course, possible because it's the only expected outcome due to the fact that it's the only possible outcome.  You don't need to worry about "this post can't seed near posts X or Y" vs. "this post is superdangerous near post Z" logic / balancing factors because ALL posts will have overlapping ranges, so it'd be 100% guaranteed to be like a fleet blob fight: counts that counter the counter-counter all at once.

Anyway.

That was one idea.

[ZaneWolfe]

While this may not be relevant in the slightest I can say that I don't find most of the AI HWs I have faced to be too much to handle. I play with FS on and 2 HWs though, so my perspective might be more than a tad different from what everyone else has. I CAN say that there are some combinations that are just beyond brutal. I've not had a HW assault with the new OMD, but my previous ones had a Wrath Lance, a Gravity Post, (cant recall what it is called) a Superfort (I have never used Trains every again... HATE THEM SO MUCH) AND a Fortress. Of course the Lance and the Gravity Post were covered by the Superfort because the RGN had to kick me as hard as possible, just like the two forts were covered by the Lance. I lost a full 6-7 city spire fleet, complete with Dreadnaughts, THREE times, before I said "F@&K THIS S%$T" and just finished the FS victory condition. And even they might have had some trouble if I didn't back them up with my 4th rebuild on my spire fleet.

Now was that an extreme circumstance? Yes. And not one I have encountered before or since. I can see how with the new OMDs things could become even more crazy, but I hope those get some adjustment soon. My point being, yes the AI HW assaults can have situations where they are beyond ridiculous, but baring some rather extreme cases I've not had too many issues. Perhaps the issue is that the new HWs are more balanced as if you HAVE superweapons like the Spire Fleet to play with, and MUCH more difficult if you are playing a base game without any superweapons. Because from my understanding, most of the counter-cheese buffs the HWs got were all focused on using superweapons like Spire Rams/Perpetrators, and Hive/Arty Golems. Even the crazy powerful Wrath Lance was something that was added to try and give the Spire Fleet (the player controlled one) and/or the Shadow Champions and their minion fleet a decent threat. (From my understanding)

I cant speak about the Reserves, I am actually not sure how much of a threat they are to be honest. My Spire Fleet just tends to melt most fleet ships, Core or not. So yeah they show up when my fleet does, but unless its a metric ton of something like Bombards, I don't notice more than ships exploding as they get in front of a swarm of Heavy Beam Cannons and Photon Lances.

[TechSY730]

Maybe bump down the number of brutal picks by one on all difficulties, and some optional superweapon minor factions can add a moderate chance of adding one more. (Like if you have golems >=medium, it would add a 20% chance for the AI to get one more brutal pick)

EDIT: The various minor faction would stack by adding their odds to give the AI a new brutal pick. it would NOT allow the AI to get more than one extra, no matter how many "player favored" minor factions are on. At worst, it would assure that the AI would get ONE extra brutal pick. Remember, the base number of brutal picks has been reduced by one.

Also, a lot of these discussions are about "long term" fixes. What about "short term"/"stopgap" fixes that can be done in a patch or two, just to "ease the pain" until a more thorough solution can be "hammered out"?

I've posted my thoughts on a short term "fix" already. (Though I guess you can add the first paragraph to that list)

[chemical_art]

Tech I dont find the sunnation of your ideas simple so they dont seem so appealing. By not simple I dont mean mechanics, but in balance. They simply seem hard to balance and that is also time consuming.

[TechSY730]

Alright then, how about this for simple to balance (this suggestion does not include my previous suggestions)

-Make superfort a brutal pick and zenith trader exclusive (details in the other thread, minus the Mk. IV fortress proposal) EDIT: Just to clarify, the superfortresses' stats would be left as is, even though it is now becoming a brutal pick.
-Reduce the number of brutal picks the AI gets by 1 for all difficulties (notably, this means that difficulty 7 can never have more than one short of buying one from the zenith trader)


I'm not even sure removing planetary tachyon from the homeworld would even be immediately needed with these changes.

The only thing I am unsure about is "variable" brutal pick counts. Like should difficulty 7 go to 1 exactly or 0 to 1? (currently, diff 7 gets 1 to 2) EDIT2: I am leaning towards making all brutal pick counts constant, instead of some difficulties getting sometimes n sometimes n+1. Also, this change does indeed imply that all difficulties <6 get no brutal picks. That's fine.

[chemical_art]

That certainly is simplier, and good. The super fort is in my book a brutal pick. Like other brutal picks it is fine alone, but due to current AI HW mechanics its pain has to be tightly regulated.

I think it would be good actually tjat brutal posts have very defined values. Either you get one or you dont. Makes it easier to balance.

[Toranth]

Wait, how did this turn into ADDING an additional nasty thing to the Homeworlds?  Right now, the Superfortress isn't a pick.  Period.  Only the Trains or a specific AI type (Fortress Baron, IIRC) can cause it to appear, and there is wide agreement that Trains need fixing.  Why bring the SuperFortress into this at all?

What I see as the general complaint is that sometimes the Core Guardposts (especially the Brutal ones) can cover each other, like the Grav Reactor and Wrath Lance teaming up to violate the hull integrity of any suddenly-immobile invading fleet.  This causes them to become much nastier in combination than they would be individually.
I agree, this can be a problem.  So far, the suggested ideas seem to be:
1)  Split the posts out over the Core worlds as well as the HW,
2)  Gather all the posts around the AI Home CS and balance them to support each other,
3)  Nerf the picks to be not-so-brutal, so even in combination they don't become super-difficult,
4)  Capture something on a Core world that timer-destroys the Brutal posts (AIP rise included?),
5)  Remove the Global tachyon coverage of the AI Home CS.

I object to #3, because brutals are boss-posts.  They need to be nasty and mean, or they're "just another guardpost".
#1 and #4 try to bring the Core worlds more into play.  This is certainly a way to make the Homeworld assault more interesting - multistage it, so to speak.  I dislike the idea of being capturing a Core world for something, though - it overlaps with CSGs, and capturing a Core world is already part of Beachheading the HW.  Splitting out the Core posts could be interesting, but very difficult to balance.
#2 also revolves around balance.  Either all the guardposts need to become completely interchangable or a bad RNG roll can being back the impossibility.  This plan also has the unfortunate side effect of making all Homeworld assaults nearly identical, since system topography no longer matters.
#5 is risky, because it re-opens to door to all the old cloaking abuse.

Yes, there were other ideas (about the Reserves, for example, which I support) but while I think some of them are interesting, they are drifting away from the primary issue:  Brutal Guardposts becoming super-Brutal when in conjunction.
And I think that should be the first item of concern, if there's enough support to do something.

[LaughingThesaurus]

What if instead of global tachyon coverage, there was just a set of patrol routes and a constant spawn of AI decloakers? Avoid them, and you're fine and maybe the cloaking cheese won't be nearly as silly while you're rewarded for observation and clever movements by still being able to cloak.

[Diazo]

Actually, I just had a thought. (Only seconds ago so this is still rough.)

What about instead of making the guard posts all spawn on map-gen, make then sequential.

By this I mean there is only one guard post on map gen, but when you destroy it the next guard post spawns in it's place, until the last one when the AI home command spawns. Would also require slashing the number of core guard posts, maybe only 2 or 3 on diff 7 before the AI Home Command spawns.

This ensures that a guard posts will always spawn alone and so they can be balanced that way, this would probably require buffing the non-brutal guard posts to almost-brutal in stats.

It also had the disadvantage of being totally different from the rest of the game in terms of how it works. There would be variety based on the stats of the guard post that spawns but there's no real options, it just influences which units you bring along.

I'm assuming the mechanic that drops remains when a turret is destroyed could be applied to do this so it would not be a massive investment of developer time.

D.

[chemical_art]

Wait, how did this turn into ADDING an additional nasty thing to the Homeworlds?  Right now, the Superfortress isn't a pick.  Period.  Only the Trains or a specific AI type (Fortress Baron, IIRC) can cause it to appear, and there is wide agreement that Trains need fixing.  Why bring the SuperFortress into this at all?

It doesn't add them inherently, it puts them into the category of brutal post in terms of spawning. It means in non niche games you could see it more often, yes. However if it does it is in place and not in addition to a brutal post.  So it takes the place of for example a wrath post, rather then the potential of next to it. It also prevents the above situation of it appearing next to a poly-crystal crashing post.

So if anything, it brings super forts more streamlined: You see it in more situations, but prevent the situations where it wrecks everything.

[Cyborg]

I'm only joining the thread because somebody mentioned CSGs.


I'm watching you.