AI War state of the game

[keith.lamothe]

This conversation is... everywhere.  Good grief.

Indeed

However, that idea was forcibly discarded for this game so that 'early mark' ships aren't made completely obsolete during an upgrade.  It doesn't fit in any lore but was a base design decision.

I actually thought of a decent (imo) explanation recently: your ships use AI, but to avoid subversion by the enemy AI players you can only have so many copies of the same AI networked together (that common sci-fi malarkey about some kind of critical mass, etc).  Developing/finding sufficiently different versions/types of ships to avoid crossover is highly nontrivial.

Total BS, but hey

[LaughingThesaurus]

Once upon a time, I was telling my friend about the game. My friend pointed out how many people must have died in every campaign of AI War. I honestly didn't have an answer, because I didn't know that the ships were manned or not. I figured the scouts were always unmanned, but the fleet ships? I guess since they're made to die essentially, they probably aren't manned. The human race probably doesn't have that many people left alive.

[Wingflier]

If you nerf the econ guerilla attacks become unreasonable, you only blob because otherwise losses are even less recoverable.  You can't 'donate' units, you MUST fight with everything.  No Raid SS attacks on the flank.  No Eyebot assaults for an unreasonable guardian.  You blob, you bait, there is no fight, there is only overkill... or waiting 2+ hours to recover.  No more multiple assaults on MK IV worlds, they're brick walls.

Erm what?

Who used guerrilla attacks to begin with? 

You're preaching to save a dead child.  We've all pretty much agreed that blobbing is the best strategy.  Nerfing the econ isn't going to change that either way.

I agree with you that blobbing is a major issue with the game, but the way to fix it is not to LITERALLY OVERLOAD the player with so many resources he can't spend it all.

Sure, maybe the said player would be more likely to use guerrilla tactics, but he would be even more likely to just send his entire blob on said "raids", since guerrilla tactics aren't even necessary anymore.  I'm now playing a 9/9 game on Crosshatch against 1 technologist and 1 hard and 5/5 Advanced Hybrids, I own 5 planets now (including my HW), and my money is literally maxed out, I can't possibly use it all.  The AIP progress is 50...

I didn't take out those Progress Reducers with guerrilla-style attacks, I sent my entire armada and did it because f*ck efficiency.

End result is going back to the Econ stations and ignoring the Harvesters anyway.  If you're going to do an econ nerf, you need to go across the boards.

You don't get it.

Econ Stations are a choice, Harvesters are not a choice.  Harvesters have no downside aside from the knowledge you spend.  Econ Stations take the place of a much more defensible or practical Military/Logistics Station.  You may lose the game during a major attack because you had an Econ Station and not any of the other 3.

In other words, I think the price for Econ Stations is fair.  I see no problem with a player having max resources if he's paying the penalty in his defense.  This is why your argument on this is flawed.

Keith has stated the game design goal of the multipliers: To create counter-ship fights.  Most multipliers are in the 4-6x range, with a few hitting 8s (I'd have to look again).  Halving them again practically removes any reason to bother.  Is that the point?  That blob on blob instead of strike ship vs. particular foe is the point?

It's already blob vs. blob brony, and that's WITH all the multipliers.  The multipliers don't make it any less blob vs. blob.  I'm playing difficulty 9 with a Technologist, and if I want to win, I blob.  There's already no reason to bother.  You're preaching for a dead child again.

Wing's heavily advocating significant micro and this would actually defeat the purpose, not improve it.  Everything meet everything!  FIGHT!  Well, why not, they've only got fighters but who cares, nothing's really that much better.

My specific solution to this problem was not to change the multipliers, that was somebody else's.  My solution was to add a bonus ship omission file and add Super Guardians to the game:  http://www.arcengames.com/mantisbt/view.php?id=9450

Then do a complete hull type and armor rework, starting from the ground up. 

-----

Everything else you talked about wasn't really worth addressing.  The 10/10 argument I had was with Chemical_Art, and it was really unrelated to this discussion.  In terms of the "repititive nature of AI War", we all pretty much agree that could be addressed in a million different ways, so I'm not going to respond to the idea that its current state is fine just because it's a PvE game.  Once we've done all we can do, if it's still extremely repetitive, then we'll talk.  Just adding the bonus ship omission file and Super Guardians would make each game measurably more intense and dynamic than it is right now.

There are a lot of other great suggestions in this thread as well to do things like that.

You may not want anything to change but that's just your personal preference.

[Kahuna]

Who used guerrilla attacks to begin with?

Anyone who wants to destroy a Data Center 5 hops away.
With MarkI and II RaidSSs you can completely neuter planets. Destroy Eyes, Ion Cannons, Tachyon Guardians or whatever.

Did you think everyone just unlocks all MarkII ships at the beginning and blobs around the galaxy till the end?

[Wingflier]

Seems to be working for me at least.  Then again I've got a Champion with a shield ability that blocks insane amounts of damage, Spire Corvettes that rip through armies of units like paper, and more resources than I know what to do with.

[Aeson]

If you nerf the econ guerilla attacks become unreasonable, you only blob because otherwise losses are even less recoverable.  You can't 'donate' units, you MUST fight with everything.  No Raid SS attacks on the flank.  No Eyebot assaults for an unreasonable guardian.  You blob, you bait, there is no fight, there is only overkill... or waiting 2+ hours to recover.  No more multiple assaults on MK IV worlds, they're brick walls.

Erm what?

Who used guerrilla attacks to begin with? 

You're preaching to save a dead child.  We've all pretty much agreed that blobbing is the best strategy.  Nerfing the econ isn't going to change that either way.

Just because you do not use guerrilla attacks does not mean that no one else does. Maybe I won't send a particularly small part of my fleet, but if all I want to do to a particular planet is kill one or two special structures, or remove a particularly annoying guard post, then I'm probably not going to send my entire fleet, because I have better things for it to be doing than fly off somewhere to go kill a couple things - I could be flying it off somewhere to kill a lot of things, or to kill some very big things, or to distract the special forces and keep them occupied while my smaller force goes off to kill something. Alternatively, I could be using my small force to bait the special forces out of an area so that my main fleet has a better chance of accomplishing whatever it is that I wanted the special forces put out of the way for. And if I'm going to bait the special forces out of the way, I may as well attempt to make the sacrifice of that portion of my fleet worthwhile by killing something that I'd like removed, anyways.

You say that the game doesn't have much strategic depth. I disagree with you, especially with regards to special forces - what more strategic depth do you want, beyond having a force which responds to attacks on important (to the AI) worlds that you can choose to (try to) distract by making a raid on some place while you send your main force off to kill something? Maybe I'm missing something, but generally speaking, the special forces are not something that I want to face in a fleet battle, unless I have superweapons available or something like that, or such a strong economy that I don't mind having to replace my entire fleet after getting into a major battle with a superior force (and yes, most of the time that I encounter the special forces, I would consider the special forces to be a superior force).

Econ Stations are a choice, Harvesters are not a choice.  Harvesters have no downside aside from the knowledge you spend.  Econ Stations take the place of a much more defensible or practical Military/Logistics Station.  You may lose the game during a major attack because you had an Econ Station and not any of the other 3.

Harvesters are a choice. They are a very tempting choice for the first knowledge unlocks you take. But that doesn't mean that you have to take harvesters, and I would not say that getting harvesters is so superior to taking some ship unlocks that I would always unlock mark three of both harvesters rather than taking a ship unlock. You say there are situations that you might lose the game because you took an Econ Station rather than a Military or Logistics or Warp Jammer Station - is this argument any different from saying that you can lose the game because you chose to unlock harvesters rather than opening up the next mark level of some ship type? If I recall correctly, I have to spend about 9000 knowledge to bring both harvesters to mark III. Certainly, this is a very tempting investment, but for about the same knowledge cost, I could have unlocked mark II caps of all three triangle ships and had enough left over in my starting knowledge pool to do the same for whatever my bonus ship was. Is unlocking both harvesters immediately so much of a boon that you'd rather have them than an extra cap of ships?

Granted, I will usually take one, and occasionally both, types of harvester up to mark III at the beginning of the game, but I also sometimes take neither. Doing this will also encourage you to use more guerrilla-style tactics, as you don't really have the economy to quickly replace a large part of a Mark I and a Mark II fleet that early in the game, if you don't unlock mark 3 versions of at least one harvester type.

Sure, maybe the said player would be more likely to use guerrilla tactics, but he would be even more likely to just send his entire blob on said "raids", since guerrilla tactics aren't even necessary anymore.  I'm now playing a 9/9 game on Crosshatch against 1 technologist and 1 hard and 5/5 Advanced Hybrids, I own 5 planets now (including my HW), and my money is literally maxed out, I can't possibly use it all.  The AIP progress is 50...

I have played the game since at least the 3.0 period, even though my user account is very recent. It has never been my experience that the player HAD to make use of guerrilla tactics. Granted, I play on lower difficulties than do most players on this forum (usually difficulties 5 to 7, depending on how I feel; occasionally on difficulty 1 if I just want to screw around, although the AI acts like it's dead at that difficulty, and I think I tried difficulty 8 a while ago), so your experience might be a bit different, but the way I've seen things in game is that there have always been advantages to sending a small force off to do something, but it has never been necessary.

I think the problem that you have is more along the lines of that you've found what you consider to be the best way to play the game, and you aren't really trying to change how you play, rather than that the game itself isn't offering you meaningful strategic choices. There is a great deal of strategic choice involved in deciding whether to have more military power, or having the ability to replace the ships you do have faster. There is also a great deal of strategy involved in choosing how you take worlds - do you want to have a fully-contiguous empire, do you want to have islands under your control scattered across the map, a mix of both? For that matter, if you aren't going to try to conquer the galaxy, which planets are you going to take? A world with an ARS might be nice, but so is a world with four of each harvester, and so is an advanced factory or a fabricator.

Here is a suggestion - why don't you try a game where you do not unlock any harvester upgrades at the start, and instead unlock as many fleet ships as possible. Then play through and see whether or not you consider taking your whole fleet over to another world to be just as reasonable as sending only enough to get the job done. If you want to say that there is no real choice other than taking both harvesters to mark III, then why don't you try a game where you don't take both harvesters to mark III, and then see 1) if you consider that playstyle to be viable, and 2) if your tactics change any.

[Martyn van Buren]

I'm sympathetic to some of your concerns, but have you tried playing a game and not upgrading harvesters?  Do you like it more?  As someone said above, I do think it's true that AI War is kind of a toolkit for making the game you want; if you take every advantage you can squeeze out of it it's not as much fun as letting it throw some surprises at you.  On that note, I feel like it might be worth trying starting off with the bonus ships you don't like and not hacking ARS's, too --- do you not think it might be more fun for you to have to cope with some ships you don't want?

[Wingflier]

You say that the game doesn't have much strategic depth. I disagree with you, especially with regards to special forces - what more strategic depth do you want, beyond having a force which responds to attacks on important (to the AI) worlds that you can choose to (try to) distract by making a raid on some place while you send your main force off to kill something? Maybe I'm missing something, but generally speaking, the special forces are not something that I want to face in a fleet battle, unless I have superweapons available or something like that, or such a strong economy that I don't mind having to replace my entire fleet after getting into a major battle with a superior force (and yes, most of the time that I encounter the special forces, I would consider the special forces to be a superior force).

Not sure what you're talking about.

I'm playing 9/9 with Advanced Hybrids and a Special Forces Captain AI.  This means that there are about 3x as many Special Guard Forces Posts on the map.

I fly around the entire galaxy with my entire fleet DARING them to stop me.  They don't even pose a threat, they can't even put a dent in my resources.

Special Forces are a joke, they don't discourage blobbing.  I don't know why they were even nerfed in the first place, once you kill the first spawn they never reach critical mass again, even on a high difficulty with an AI that is spamming SF Guard Posts.

Just because you do not use guerrilla attacks does not mean that no one else does. Maybe I won't send a particularly small part of my fleet, but if all I want to do to a particular planet is kill one or two special structures, or remove a particularly annoying guard post, then I'm probably not going to send my entire fleet, because I have better things for it to be doing than fly off somewhere to go kill a couple things - I could be flying it off somewhere to kill a lot of things, or to kill some very big things, or to distract the special forces and keep them occupied while my smaller force goes off to kill something. Alternatively, I could be using my small force to bait the special forces out of an area so that my main fleet has a better chance of accomplishing whatever it is that I wanted the special forces put out of the way for. And if I'm going to bait the special forces out of the way, I may as well attempt to make the sacrifice of that portion of my fleet worthwhile by killing something that I'd like removed, anyways.

Maybe someone else does.  My point to the previous poster was that nerfing the economy by removing MKII and III Harvesters would make it MORE likely that you would use guerilla raids, not less likely.  Guerilla raids are efficient, they don't waste your entire army on the trip.  With a small amount of resources, you can accomplish a lot.

If you also had less resources, maybe Special Forces Guard Posts would be more of a problem (I doubt it), but still, both of the things you're talking about would be more of a factor if you weren't continually stuck at 999,999.

I think the problem that you have is more along the lines of that you've found what you consider to be the best way to play the game, and you aren't really trying to change how you play, rather than that the game itself isn't offering you meaningful strategic choices. There is a great deal of strategic choice involved in deciding whether to have more military power, or having the ability to replace the ships you do have faster. There is also a great deal of strategy involved in choosing how you take worlds - do you want to have a fully-contiguous empire, do you want to have islands under your control scattered across the map, a mix of both? For that matter, if you aren't going to try to conquer the galaxy, which planets are you going to take? A world with an ARS might be nice, but so is a world with four of each harvester, and so is an advanced factory or a fabricator.

Well, I mean if you consider unlocking MKIII Harvesters right at the start, and mindlessly blobbing my fleet around the map until I win a "strategy", then I suppose you're right.  I don't think it's much of a strategy though, you could make an AI to play AI Wars for me.

Here is a suggestion - why don't you try a game where you do not unlock any harvester upgrades at the start, and instead unlock as many fleet ships as possible. Then play through and see whether or not you consider taking your whole fleet over to another world to be just as reasonable as sending only enough to get the job done. If you want to say that there is no real choice other than taking both harvesters to mark III, then why don't you try a game where you don't take both harvesters to mark III, and then see 1) if you consider that playstyle to be viable, and 2) if your tactics change any.

As a general rule, I'm always going to try my best to win, so I typically don't avoid unlocking things on purpose just to make it harder.  But next game I'll try to stick with just MKI Harvesters and see how it goes.  I expect it to be a lot harder and more exciting than it is now.

On that note, I feel like it might be worth trying starting off with the bonus ships you don't like and not hacking ARS's, too --- do you not think it might be more fun for you to have to cope with some ships you don't want?

I suppose I could put a bunch of artificial restraints on myself like this, it just seems a bit silly.

edit:  Maybe I should stop popping Data Centers too.  I'm on a difficulty where each DC is only worth 10 points, and yet I'm already down to 37 AIP and I still have 5 planets.  There are still some DCs left and also If I pop the 4 Co-Processors that's a huge reduction as well.  It just never seems like the AIP gets high enough to challenge me and then my blob just runs amok uncontended.  I'm not sure what to do about that.

[Lancefighter]

I wonder. How early in the game do you needt o have a signifigant presence in a special forces defended planet to make them completely irrelevant? If you could, in theory, get enough units to take and hold a planet that the special forces deems important, and do so for the rest of the game.. Would that not make the entire special forces ai basically useless on the defense?

[Aeson]

I'm sympathetic to some of your concerns, but have you tried playing a game and not upgrading harvesters?  Do you like it more?  As someone said above, I do think it's true that AI War is kind of a toolkit for making the game you want; if you take every advantage you can squeeze out of it it's not as much fun as letting it throw some surprises at you.  On that note, I feel like it might be worth trying starting off with the bonus ships you don't like and not hacking ARS's, too --- do you not think it might be more fun for you to have to cope with some ships you don't want?

Yes, I have tried games where I haven't upgraded harvesters, and no, I don't like it more, though I also don't like it much less. That doesn't mean that it isn't a strategic choice - in choosing between harvesters and fleet ships, I'm making a choice between having the ability to field one strong fleet which I can slowly replace, or a weaker fleet that I can replace more quickly. There are advantages and disadvantages to both, and how much fun one is as compared to the other depends very much on how well you can handle downtime if you just lost a big chunk of your fleet. Having a fleet with three times the firepower from the start is a very big advantage over upgrading your harvesters, and if you take advantage of it you can probably gain as much benefit as you would have from having upgraded your harvesters - for one thing, you now have a whole bunch of Mark I ships that are relatively disposable, and whose loss will not cripple your fleet. Hence, an excellent supply of guerrilla ships, and, at least until the AIP and knowledge gains catch up with your choice of early military power over early economic strength, you have a relatively easier time than you would if you chose harvester upgrades. Certainly, the early game becomes slower due to the lackluster economy, but this is at least partially made up by not needing to build turrets for defense - if you have caps of Mark I and Mark II ships, and you're used to using a fully Mark I fleet at this point in the game, then you could:
    1) Leave the Mark IIs at home, and have a very strong defense force
    2) Leave the Mark Is at home, and have an offensive fleet that's roughly twice as strong as what you usually have
    3) Leave an equivalent to the Mark I cap at home, and send out two offensive groups each about as strong as your normal Mark I fleet
    4) Etc.

My preferred early unlocks are one harvester to mark 3, and then one or two fleet ships to mark II, and decide what to do with the leftovers later. I also go for both harvesters to mark 3, and neither harvester to mark 3. All of these choices work (at least for me), and each has their own advantages and disadvantages. I'd say that the biggest disadvantage that not unlocking any harvester upgrades has is that it makes the early game slower. Later in the game, it doesn't matter as much as you're going to have a fairly high metal and crystal income, anyways, so at that point the higher income is relatively less important than in the early game. The biggest disadvantage of the economic choice is that I have fewer disposable forces early on, and it's early in the game that I'm most inclined to make raids with small forces to draw the special forces away, or to hit some valuable target. Thus, since I like to have a strong economy, and I also like to have a military force that can afford to treat some units as disposable, I usually compromise between a full-economic opening and a full-military opening. Certainly, by making such a compromise I have more difficulty replacing losses than I would if I'd gone for a full-economic opening, but I also have more ships available, and I have some ships that I can detach from my main fleet without significantly affecting my main fleet's fighting strength. Compared to a full-military opening, I don't have as many relatively disposable ships, but I am also able to replace losses more quickly, and I'll attain my full fleet strength faster (though my full fleet strength is less than what it would be if I'd gone for a full-military opening). These are very much strategic-level decisions, which have a direct impact on both the tactical level of the game and on my strategy for the near future. Full military? I'd best make use of that, and try to capture some planets quickly to make up for the lacking aspects of my economy. I might also be better able to engage the special forces, or I could sacrifice my Mark I ships as a distraction while my Mark II ships go liberate something special, or I could split groups off to go raiding while the main body stays home while I slowly build turrets and prepare for an assault on a target. Full economy? Now it's the other end of the early game spectrum - I can quickly replace ships, but I don't have the military strength to directly challenge strong targets and I want to have many turrets up quickly so I don't need to keep my Mark I fleet at home.

Choosing military strength is also, at least to a point, advantageous if you want to have a discontiguous empire, because with more and stronger warships, you can be in more places at once and still have roughly the same strength present as you would if you were to concentrate the majority of your force in one spot in a game where you chose economic over military strength.

I also rarely bother hacking ARS's. I don't find it generally worth the bother of dealing with the spawns, and it also can be fun to try out new ship types. It can be fun to make use of a tried and true ship, but it can also be very boring if I have the same ships game after game.

I wonder. How early in the game do you needt o have a signifigant presence in a special forces defended planet to make them completely irrelevant? If you could, in theory, get enough units to take and hold a planet that the special forces deems important, and do so for the rest of the game.. Would that not make the entire special forces ai basically useless on the defense?

No idea. My usual method of dealing with the special forces early on is to try to get them to go hunt a raiding group, and then send my main group after wherever it was I wanted to attack and needed the special forces out of the way in order to do so.

[Wingflier]

I wonder. How early in the game do you needt o have a signifigant presence in a special forces defended planet to make them completely irrelevant? If you could, in theory, get enough units to take and hold a planet that the special forces deems important, and do so for the rest of the game.. Would that not make the entire special forces ai basically useless on the defense?

I jump between ARS worlds fairly quickly once I get the ball rolling, so in effect this is basically what happens.  Going from ARS world to world to world with my blob, they never get enough time to rebuild their forces to anything remotely threatening to me.

This is partly because I have the resources to just keep plowing through the game like that.  Perhaps if I were stuck on MKI Harvesters, the special forces would have time to "regroup" between my ARS strikes.

I guess we'll see.

[KDR_11k]

Once upon a time, I was telling my friend about the game. My friend pointed out how many people must have died in every campaign of AI War. I honestly didn't have an answer, because I didn't know that the ships were manned or not. I figured the scouts were always unmanned, but the fleet ships? I guess since they're made to die essentially, they probably aren't manned. The human race probably doesn't have that many people left alive.

I assume that only the home command and those pods and settlements are manned. Losing manned items causes AIP increases (I guess the AI cares after all) depending on the casualties. Everything else is controlled by weak AI cores in the home command station. Well, maybe the Spire man their ships, no idea about that. I guess an AI designed to fight with and against humans doesn't care about non-humans.

IIRC the organic factions in Schlock Mercenary only use low-powered AIs for their ships now to prevent them from joining the fleetmind, a massive, godlike AI that owns pretty much the entirety of the Milkyway and uses its nearly infinite resources to fight omnicidal dark matter creatures in other galaxies after they tried to annihilate the entire Milkyway galaxy. The main difference between Schlock Mercenary and AI War is that Petey is a benevolent dictator since he has the ship design sensibilities of the Spire.

[onyhow]

The pre-4.0 manual said that all your ships are unmanned...

[Histidine]

Because nothing discourages something in AI War like giving AIP for it, maybe what we need to counteract blobbing is something like KDR_11k's Sleeping Giant AI idea except it gives permanent AIP. Makes thematic sense too!

(Of course the vanilla Sleeping Giant works for the purpose too, AIP is just a (probably) more evil lamothe way of doing it.)

[KDR_11k]

If you can unlock all MkII ships without tanking your economy by building them before expanding the first time that's a definite proof that the human economy is really damn OP. It used to be that you couldn't even build a cap of bombers or frigates without tanking your starting economy so you had to expand before you had your caps built up.

One thing with blobbing is that it's fairly inefficient, e.g. when a planet has a lot of frigs or railclusters or whatever you shouldn't take your fighters with you but it's easier to simply grab everything and let your economy replace what you lose. Even smashing an ion cannon with your whole blob is completely acceptable with an economy that can handle it.

Furthermore I'm advocating an energy economy cut, that would mean you couldn't even build full caps of everything without taking some extra planets. Obviously that would lead to a lot of scrapping and rebuilding so low-power mode should probably be reintroduced (or some sort of cold storage building), As it is I haven't seen energy matter even when using medium golems. I don't usually maintain a standby fleet of starships and don't use forts beyond mod forts and miniforts so maybe higher difficulties have higher energy consumption from building all those heavy units while taking fewer planets but I used to feel constrained even below 7/7 with the old energy economy. Probably because MkIII powerplants were something you couldn't plonk down nilly-willy but still had to use so you had to balance fleet cap vs build speed. The current power economy makes paid power a rather bad proposition, it does such a number on your economy while offering slim benefits. Then again the paid power is no longer affected by your territory control so making it produce more power for the money wouldn't work out well either.

Maybe a power cut plus an AI response cut at the 9er difficulties so you can actually survive with configurations other than a handful of worlds with massive fortifications on them.

Balancing the different difficulties there is going to be a nightmare since it alters the physical strength of the AI, not just its options and in an RTS like this that additional physical strength requires additional power on your end instead of merely more precise play (since you cannot negate that enemy strength with sheer skill). So there's a very narrow margin of difficulties where you actually face that promised overwhelming force without being completely unable to do anything about it. We all play with different AI physical strengths so we all see different options for changing the game balance and the range is currently huge (3x multiplier per difficulty level IIRC?) while we all get the same tools. I think the difficulty range may need to be narrowed so the AI's physical strength doesn't vary so much, just its aggressiveness, plots (stuff like the special forces and core reserve and things like that) and maybe other options (e.g. lower difficulties would make it harder to trigger a full planet wakeup and maybe go easier on the major obstacles like Raid Engines and Eyes). It's just not possible to balance a fixed force against a highly variable one. As long as a nerf could result in a supposedly viable difficulty being physically impossible balance is going to be extra hard. Forcing people into smaller forces and guerilla raids may make it impossible to deal with extremely well fortified worlds.

Generally I don't feel like we're using all the variables that the game has for balancing (e.g. we have two resources but it's common to consider them one resource! There's no clear line between the two like there is with e.g. minerals and vespene gas or mass and energy), there are too many values on a bonus ship that you don't even look at when determining which one to take.

I guess we're dealing with two balances here, unit vs unit and player vs AI.