AI War II: Design Document Updates Round 2: Space Platforms

[x4000]

Hey folks,

A variety of new things today: https://docs.google.com/document/d/1IdzU90psGas_3UFe23BLvsGQ8fclec49NmnbHfwkZ8w/edit#

You'll have to do a Ctrl+F search for these headers, apologies.

Minor stuff:
1. Added "Rationale For Scope Decisions"
2. Added "1.a.v. Drop-In Co-op"
3. Added "1.g.iii. Deeper Steamworks Integration"
4. Added "Clarifying A Few Points About Modularity"
5. Added "4.a. Temporarily Added AI Progress"
6. Barebones start to "Section 5: Ship(Squad) Mechanics"

The biggies:
1. Added the bulk of "2.b. Space Platforms and Ship Bays", although more is still needed there.
2. Added "Section 3: Graveyard of Discarded New Ideas" with in particular "3.a. Replacement of AI Progress" in it.

[Captain Jack]

Space platforms sound REALLY COOL, and if they weren't going in here I would have suggested something very similar!

What do you think about making hacking a function of a specific platform type?

EDIT: You talk about planets having a certain number of platform support points. Wouldn't we just call that... supply? 

[Tridus]

re: Space Platforms.

On the one hand, I see a lot of upsides to this design. It could do a lot of pretty cool things. In particular it leaves room to add things like Spire Cities back in, because that can be done as a special platform you gain the ability to build when you do whatever it is to unlock it. It gives you the customization in a less fiddly way, makes turrets work somewhat sanely, lets people have their movable defenses if they want them, and such. I like all that.

On the other hand:

• What determines how much space there is for platforms on a given planet? Is this something players can control somehow?
• Is the only thing stopping me from dropping fortress platforms on every planet that some planets just don't have enough platform space to do it?
• Similarly, there's no power concern about doing that, because the power is all self-contained?
• I know you mentioned people wanting templates for this, but if I'm dropping fortresses on 15 planets, I do want that. I don't want to play Power Grid (which is a great board game) every time I build one, when what I really want is the same setup.
• While having universal cap defenses like Heavy Beam Cannons no longer work differently from normal turrets is great, doesn't that mean balancing them is harder without a per-space platform limit instead? If I have to unlock them with K, I expect them to be better than base stuff in some way, which means I'm going to want to use them.
• I'm having a hard time visualizing how this would all work.

I'm positive on the idea overall, but I'm having a hard time understanding how it's all going to work at this point. Maybe it's just too early for that and it'll make more sense once the game is playable.

[Orelius]

So we're getting a city-building style space platform central ship that replaces turrets?  That sounds like a fantastic idea to me.  I've never really liked the whole turret placing metagame.

To build on top of that, I think that making space platforms comparatively huge compared to ships and the planet would be pretty cool.  It'd feel more like you're actually raiding a base rather than just a small collection of scattered buildings.  You'd be able to get a pretty good picture of how strong or a weak a system is based on its size, and you can make very noticeable and visible progress as you pick a station apart.  Adding collision to components or adding forcefields along the way would make it difficult to simply snipe a command ship or other critical components like ion cannons, making them actually a significant threat instead of a micro sink.  Maybe a mark 1 AI planet could have a small station roughly the size of a mark III forcefield in AI War classic, and a core world or an AI home planet would have a station at least half the area of the usable area in a planet.

[Sestren]

The space platform concept is something I would very much enjoy playing around with, but I do have some questions.

9. Other types of modules of note.
...
c. You’d want to have shield-style modules, but these would work differently from forcefields in AI War Classic.  These would be a lot more interesting and precise, because they would protect only the other modules adjacent to them (along struts).
d. Other modules might provide similar adjacency bonuses to modules that are adjacent to them within the mesh of modules on the platform.

If adjacency bonuses are an important factor then people are going to want a fairly powerful and comprehensive interface to arrange *exactly* which modules go where on a space platform. Combine this with the ability to swap modules out on the fly and I know you wrote...

10. One of the things that people are potentially going to immediately ask for with space platforms are blueprints for those, but we’d really rather avoid that.
a. Rather than thinking of space platforms as a form of modular ships that you build directly on the game map (though they are, really), we’d like to think of these as much more situational, somewhat-city-building-eqsue tasks.

...but, at risk of turning this into an unwanted discussion that wouldn't die, it seems like you said the interface for adding and managing entities on a 'modular' ship would be the hardest part of making those work and here it seems like you are planning on doing that kind of work anyway. Is there a technical distinction between the two features here that I'm missing?

On another note, you could possibly maintain the utility of the 'raid' style of ship a couple of ways in a framework like this. They could bypass the protection of adjacent shield modules, thus allowing the Raid Starship to surgically remove that heavily protected alarm module for example. Or, since these platforms start with a module on the central node (presumably some sort of command module?) you could have the node be untargetable until everything else is destroyed normally but allow raid type ships to attack it directly. Let its destruction impart some sort of penalty to the rest of the platform (total shutdown until repair, reduction in power, reduction in RoF, loss of planetary control for a command platform, whatever) until repaired.

[chemical_art]

I find the space platform idea interesting, but for me it runs until the same gripes that I have with modular ships. Eventually the novelty of new designs will be washed away and then I am stuck following a few ideas, but have to do extra steps to actually implement the idea.

There has to be blueprints, because this idea while flexible tests skills not seen anywhere in the game. In addition, this is pretty overwhelming for a veteran like me, I pity a new player.

There should be pre-made platforms already made, similar to how the current stations are set up for now (economy, logistics, military).

In short I see the gains of combining all the things together in one platform. But I do not enjoy the approach this is going from. The idea should not be they are puzzles to solve because it is too much of a departure from the skills the rest of the game does. This is not a puzzle game, this is a strategy game and there is a key difference. A puzzle game would enjoy the process of making each one optimal. A strategy game finds the optimal layout, preserves it, and then tweaks it as local situation dictates. I feel that is the approach stations should have: a few generalized blueprints that the player can tweak if desired.

edit: To be clear, I love the concept and the options it unlocks a player. What I do not like is how...time consuming it would be to tweak each individual one. It sounds like a great idea for an expansion because at that point I would be experienced enough to actually take advantage of changing individual modules. But the idea of doing that on the first planet I conquer (so within the first hour of the game) sounds too overwhelming.

[Tridus]

Also - one of the things with unlocking new defenses in AIW Classic is that you get more of them, on top of whatever it is they do. Heavy Beam Cannons again, as an example. In addition to being pretty good, they don't supplant anything. They're an extra thing.

In this setup, I have limited space for modules on my space platforms, which means if I'm going to unlock and use anything, it's going to have to *replace* something I could have already built in that limited space. That means the unlocks are going to have to be good, or they won't be used.

That's a good thing for reducing clutter, in that I won't have 80 Laser Turret mk I, 80 Laser Turret mk I, 80 Laser Turret mk I, etc, but it means anything that isn't one of those had better be worth swapping in.

Also also - if upgrades stick around, I'm going to want a way to do it in bulk. Going around to every platform and swapping mk I turrets for the mk II versions will get old real fast.

Also also also - Isn't the work to do this also most of the work required to have modular ships designed in game? I'm not clear why we can do this but can't do that, since they don't sound all that different to the end user. "I can design my space platform in game but have to edit XML to swap the lasers for shotguns on a Riot Ship... why?"

[Nuc_Temeron]

On Space Platforms:

Whew, deep breath indeed. I had to read this twice.

This sounds a lot like another game that I play occasionally, Distant Worlds: Universe (DW). I play DW in expert mode (no automation, no suggestions) and I delete all the starting ship designs and remake them from scratch. If you're familiar with that game, you understand the massive burden that entails.

Now, that's how I prefer to do it in that game, even though I don't have to. So, in that way, I like what you're saying here because your description of Space Platforms sounds an awful lot like the ship design screen in DW, and I do enjoy that somewhat because of the freedom that offers. If you pull it off correctly that could be pretty great.

However, it can become rather tedious building all your ships from scratch. One of the reasons I do this in DW is because the AI's designs are terrible. I know I can't trust it to do a good job, so I just do it myself. At the beginning of the game I enjoy the freedom, but as the game starts to go longer, it becomes kind of a chore. Every science upgrade, you have to load up EACH design and rebalance it with your new power generator or whatever. Sometimes I wish that the default designs were already min/maxed so I could get along with moving dots and lines around.

In AIW2, I'm guessing that when the player spends Science Points we're now going to usually be unlocking Modules for our Space Platforms / Turret Pods / Command Stations / Fortresses instead of another specific ship/building. Making this a non-tedous process, like it is in DW, might be a challenge.

When I play AI War Classic, I don't bother making modular ships (Riot Control etc) because I already know that the zillions of ships and buildings are all happily balanced. It's almost like a collectible card game; everything has an interesting unique ability and is recognizable by a distinct icon (like a card's picture). I can look at a system and know what's there AT A GLANCE, because I know the iconography. I don't feel the need to make my own cards, because there's already so many interesting ones to play with.

The diversity of buildings and ships is one of AIW's greatest strengths!

If all of our buildings are now one of these 4 types, what happens to the iconography? How will we tell our ships, the enemy's ships, everyone's ships apart? Will I be able to tell what's at a planet with a simple glance? I don't want to have to open up each ship's design in a window and scrutinize it. Manual design is not at all like a collectible card game, it's more like Distant Worlds Universe (which is inferior to AI War, in my opinion). Will players tend to build a single homogenous ship type and just ball-attack with it the whole game? Will a "best build" develop that everyone just does every game, rendering all this freedom pointless?

And I can't place turrets anywhere I want anymore? I do see some merit to grouping blob-turrets into one entity, frankly I do usually just dump all the turrets I can of a given type in one spot.

I feel like the proposed changes here are massive. Massive! Completely new game massive. I super-duper respect Arcen's creativity and vision, if anyone can pull this off I suppose they can, but I feel like this is drifting very very far away from AIW very quickly. Why rewrite all these systems that already work so beautifully? The math-based gains seem wonderful and all, but I'm skeptical of a complete overhaul. AI War is a fantastic game as it is, if it's not broken, why fix it?

[Draco18s]

I am loving the space platforms concept here, and I think the above folks addressed my main concerns/questions.

But one thing occurred to me:
Can the player get an alarm "module"?
I realize that these things getting shot is already going to throw alerts, but I think it would still be a worthwhile tool to have an...object that goes "OH GOD, THERE ARE SHIPS HERE" before anything starts shooting.  I'd drop one on every Mk4 fabricator, z-generator, etc. world.

[chemical_art]

...

I certainly agree with the DW vibe of creating ships. It is entertaining but also tedious. I also enjoyed making ships in Master of Orion, but in that game it was almost a necessity because technologies changed so much then (so while ship ideas could go game by game, the ship designs were always slightly different.)

I will also say that the ability to at a glance tell a command station's function is also important. I would not like to have to zoom in to the unit level on a command station to understand it's situation, and until I am extremely familiar with the game I will need to read the tooltip of the station so it is imperative that is clear too.

[ptarth]

• Do stations really need to have slots? Why not have infinite slots?
• Do stations really need to have power? Why not just pay for the cost of adding "modules" by having to buy engines/whatever.
• Being able to save blueprints and order conversions of unit X into unit Y, sounds really important.
• Hull size seems important, but is it? Could ships be defined simply by what is on them? I.e., if you have small gun, engines, bridge, storage, isn't that a fighter? If you have a bigger gun, engine, bridge, etc a corvette? And so on?
• If that's the case, then you could swap out hulls and ship limits for something like global supply production and usage. If you are over your supply limit, your ships perform worse. Perhaps a hard cap? Or perhaps even allow boosting ships that will temporarily boost your supply.
• Following the above, then the "core" of the game would be a series of predefined blueprints. Mark improvements would then have higher efficiency parts. And then the player could add/update their own variants of blueprints.

[tadrinth]

Can the player get an alarm "module"?

Isn't this an option under CTRLS => Planet-specific options? Alert if > X ships present?

On ship platforms:

In AI War classic, the human player generally has a lot fiddlier defense setups than the AI.  That's fair because you're countering much larger numbers, and because the AI doesn't need to understand the human defenses to throw gobs of ships at them.

When attacking a guard post or fortress, though, you're usually dealing with just one structure that frequently has an obvious triangle counter.  Put two together (fortress or FF on top of a guard post), and now you might not have a triangle counter, but perhaps there's a bonus ship that would be perfect if you had it.  Or perhaps not, a post that wrecks polycrystal next to a fortress is just going to be painful.

Are mod fort command stations interesting to kill as a player? Or are they just a pain to kill no matter what you do?  Because a pile of random modules can mean there isn't really any counter.

So, I'm worried that AI platforms could become very hard for a player to read and not particularly counterable.  Either they should have very few modules, or the logic for picking AI modules should be careful to create platforms with enough personality to be counterable.

Overall, I think the city-building aspect of human space platforms might be a rabbit-hole of complexity.  I'm all for abstractions to streamline setting up defenses, but staying at a pretty high level of abstraction might be fine here. 

This does also sound suspiciously like modules.

Oh, and swapping engines onto a platform temporarily to reposition it and then swapping back to an all-offensive loadout sounds like a recipe for annoying micro.  Platforms should be able to move to allow repositioning, but moving should disable their attacks; an engine module might make them able to move and shoot (roll your own Sentinel Guard Post).

I was more expecting space platforms to be  MOSTLY about the core, and then have a very small number of support slots. Each slot  can get a FF module or a range boost module or a damage boost module. Gravity/tractor/tachyon could be support modules on regular turrets or be options on a special support-turret platform.  This is probably also a rabbit hole, but it's not nearly as deep. I think it's roughly equivalent to replacing human turrets with Mod Forts. 

Or just have variants (like for ships) that either have a support part or don't.  IE, this is a lightning turret platform, and it comes in regular or FF protected flavors. Pick one.  I feel like for most turret types, and really most defensive structures, it's pretty clear whether or not you want FF coverage, though you don't have enough FFs to cover all the turrets you'd like to.

[Pumpkin]

Here are my thoughts on Space Platforms. I need to read others' comments more thoroughly but for what I briefly read, I think I'm in the minority, this time.
EDIT: Uh, maybe not...

Personally and for now, I disagree with the addition of Space Platforms. While I would love to see more streamlined buildings, I think this space platform idea, as it currently is, has two problems. First, I feel it go to the opposite direction of where "no modular ships" were going. Then, I feel it has the same problem of several SBR mechanisms: it feels disconnected from the rest of the game. Basically, I understand that platform construction / filling as a sort of mini-game puzzle thingy. It would be added complexity where the game beg for simplicity and coherency.

However I think there are very interesting ideas, or more accurately goals, in there. "Squads" for buildings is interesting. A more versatile base on which players can create a defensive strategy is a very interesting goal. More diverse and interesting "guard posts" or fortresses things too. (But modular puzzle minigame, not in AI War, please.)

I like the idea of predefined spots (metal deposits) where things could be built.
However I'm afraid that would kill the Kahuna defense style where tractor turrets and gravity turrets must be placed in precise spots between wormholes and AI targets (command station) within range of damage turrets.

All in all, I think there are interesting ideas and goals pursued here, with this new kind of non-ship unit. But as it is presented here, I feel it's very poorly integrated with the rest of the game.

Or else, add champions back, make every ship modular and Knowledge would allow research of parts (guns, thrusters, hulls), à la Star Ruler. But that won't be an AIW2 that I, personally, would kickstart.

[kasnavada]

Hum.

Before, we had to place manually, for each planet, a set of turrets / energy collectors / ressources.
Now, for each planet, we have to design a modulable command station, basically.

Note : I don't like this idea as is currently written presented.

Before we had the possibility to have turrets & stuff positioned at multiple place, beachheads, I don't see any of that being possible with the platforms. Is that intended ? Points below:

- Is there a limitation on the platform size ? platform number / planet other than the limit / planet you're envisionning ? Because else, just building the biggest there is the optimal design.
- The limitation per planet from 10 to 80 also sounds like a terrible idea. Like "Woah this planet is a perfect chokepoint... NOOOOOO 10 platform point ? F**** RNG !!!!!
- Is it possible to build platfom on enemy planets ? If yes, isn't it kind of OP to have the same defensive options ? Wasn't that kind of an issue, balance-wise ?
- Is there a limitation on the size of platforms ?

Next, there would be players that actually hate rebuilding the same station all over again. With station size from 10 to 80, there is basically no ways to store designs, so designing an auto-build option seems mandatory to me.
- Defence => automatically places defensive modules + energy as needed
- offense => offensive modules + energy as needed, and so on)
- ...
- with possibly more "defense options" like, shield def, armor def, and more "offense options" like "long range, short range".

Minor nitpick: you're assuming that people defended the ressource collectors on the front lines, or even other platforms. That was never my experience. At best they'd draw some AI while the others ran to the command station, usually the AI didn't even bother moving to shoot them. Oh, they'd shoot IF they were in the way, but it was rather easy & fast to come up with a solution, don't put those in the path between the command station and the wormholes. As far as shooting priority are concerned, command station >> whatever else there is. Possibly a minor exception => fabricators & AI only buildings. Anything else the player builds is next to irrelevant to the AI.

[Wingflier]

Personally I really like and am on board with the platforms idea.

Turrets in their current form were extremely tedious. On one hand you could just place a bunch in a sector and kind of forget about them, but on the other hand there were the costs, the repair fees, rebuilding them if your reclamation bots had been destroyed -- plus the thing I disliked the most: They died if anything so much as breathed on them. In most cases they were good for defending harassment and light raiding attempts, but when the enemy sent any kind of sizable force they would usually die instantly. There was also the counter-intuitive mechanic that putting shields over them made them do less damage, something which was both confusing and somewhat frustrating at the same time.

In my opinion, anything that reduces the unnecessary micromanagement in an already user input taxing game is a good thing. However, the options of these space platforms are going to have to be pretty diverse if they are going to allow and represent meaningful choices for the players. One idea that I liked was being able to assign "squads" to the platforms, which I kind of envisioned as like a "roaming defense". This is something very scifi-esque, and definitely fits well into the theme of space platforms, should you choose to go that route.

Also, I think this idea also fits better into your new design goals of "no instagib mechanics". The way Orbital Stations were before, it was not uncommon to lose them in a massive chain of events, resulting in power loss, resulting in shields dropping, resulting in everything popping at once, because of some silly Raid Starships or some other unexpected threat you had overlooked for 15 seconds. These same kind of threats could also affect your Home base as well, and result in an instant loss if not careful.

Anyway, in a game like this, I think it's wise to give players time to react. When the games go on for 12+ hours, people should be more worried about grand strategy and thinking several steps ahead instead of the twitch reflexes and panicked reactions that you would expect out of a fast-paced RTS game like Starcraft. Sometimes though that's how AI War was, especially on the harder difficulties, and I'm glad we're moving away from that paradigm.