Arcen Games

Games => AI War II => Topic started by: zeusalmighty on April 07, 2018, 08:37:46 PM

Title: My comprehensive feedback
Post by: zeusalmighty on April 07, 2018, 08:37:46 PM
This is my comprehensive feedback thus far. It’s a long read. I don’t focus on quantitative balance, I’ll leave that to the more insightful. My focus is on game philosophy, playstyles, problems, and gameplay proposals

Experience/Playstyle:

550 hours classic/50 hours AI 2

My playstyle is generally high AIP (exploit all AIP reducers to keep low as long as possible), turtling (invest in turrets, mines, fortresses), surgical strikes (advanced transport raids/spire transport co-processor rush), divide and conquer (strike out large territories finding ideal chokepoint and absorb/neutralize remaining territory)

The reason I regard AI classic as my favorite game is how spectacularly it executed the premise: humanity is all but wiped out and faces overwhelming odds—success depends on exploiting the AI’s disinterest. The replay-ability is ultimately what made this game so impressive, every campaign demanded a unique strategy catered to the galaxy and the personalities of the A.I.


Philosophy of success
For AI 2 to succeed classic it must capture the same spirit—humanity have the means of thwarting and exploiting the AI’s attention to wage a successful war. A major element I would like to see more of is the ability to manipulate/exploit the AI’s knowledge of human activities. Basically, I would like a toolkit that would deceive the AI to deal with fake threat elsewhere in the galaxy—in effect distracting the warden/hunters/sentinels and manipulating AIP through subterfuge


Problems:
?   Currently the means of waging a guerilla war are too limited. Badger and Chemical have already observed that tech in AI 2 lacks versatility and players are effectively forced to use the same tech every game. This is particularly true of defense options. Mutually exclusive tech trees (I’ll give proposals later) could address the replay-ability issue by making certain game styles more viable than others in a given campaign.
?   Movement of fleets is tedious, as others have indicated.
?   Turrets aren’t effective by themselves. I’m not sure they can be without overcompensating. I think the issue is that we are expecting them to fulfill defense by themselves. They need to be supplemented with something else (I have proposals)
?   Ark needs more functionality; it’s too important of a unit to just have sit in a corner. Hacking could be interesting as a defensive tool.
?   Controllers feel too vulnerable should the AI decide to target them explicitly. Given the current targeting priorities this hasn’t been a major issue but I suspect this will be more of a target in the long-run.
?   Defense Nodes are good in theory (opens up interesting tactical choices) but in practice feel relatively unimportant. Should invest more in this idea (I have a couple)


Proposals/Ideas: (These are largely inspired by Classic and the Dune prequels)


Defense:

Traps: I want a tech tree for a variety of traps. Humans need to rely on cunning to overcome the AI and the fact that we don’t have some sort of trap system in place strikes me as odd from both in terms of lore and mechanics. (I guess tractor beams count). Here’s a few that interest me

1.   Mines from classic
2.   Attrition field: Zone trap, when active, that causes DOT (% based) to ALL fleet ships (friend or foe)
3.   Sensor disruptor: Generator that reduces nearby AI to melee range
4.   EMP mine: trap that paralyzes ALL fleet upon activation (no movement/no targeting)
5.   Stasis Trap: Trap that makes fleet “ethereal” and frozen (same as paralysis but also ships can’t be targeted)

Fortresses: Bring these Back! They were awesome in classic for turtlers and add much needed variety to defense tools

Defense Nodes/Guard Posts: Double down on the nodes. Make a tech tree for them that would result in something like Guard Posts from classic (hell, would be cool if the AI had these too on higher mark planets). Maybe even make the tech for these mutually exclusive. Here’s a few applications

1.   Stealth Field: Defense node has basic weaponry but also produces a stealth field to conceal the controller. All of these need to be destroyed to reveal controller or and otherwise insane amount of detection
2.   Drone Deployer: Defense node has basic weaponry but produces drones based on tech level (of the triangle fleet).
3.   Supercharger: Defense node has basic weaponry. Passively and massively increases the defense capabilities of all other defense on the planet, based on how many are still operational.
4.   Superior Armament: Defense node has ADVANCED weaponry and “Artillery” range (even longer than sniper)

Ark Tech

The ARK is presently disappointing; I expect more from our King. Hacking is fine and dandy but this doesn't get enough usage. Insofar as the idea is to keep this away from the AI generally speaking, I think there needs to be some ways to use hacking defensively. I would also be interesting a tech just for the ARK to make it more of an impactful unit, potentially even mutually exclusive tech options. Here's a few possibilities:  Advanced weapons, cloaking, drone spawning, passively boosts fleet/defense systems, etc.

As for defensive hacks, something cool would be the ability to disrupt wormhole transit (e.g. blackhole generator). Another one would corrupt AI targeting--certain proportion of AI strength views other AI fleet as hostile and attacks. Defensive hacks should get a lot of bang for their buck

Utility/Logistics/Guerilla:

Controller Tech: I miss the different types of command centers. Controllers feel uninteresting and are liabilities. This applies to the AI as well. I really liked how in classic there were special command centers that dramatically changed how a player should deal with them. Mutually exclusive tech similar to my proposals for Defense Nodes would be applicable to them as well.

Saboteur (Special starship): Non-combat unit, advanced cloaking, specializes in subterfuge. I think this is needed (or something comparable) because the humans are lacking a significant way to “fool” the AI, which is arguably the key to their ultimate hope of defeating the AI. Some possible functions:

1.   Uses hacking to manipulate waves/hunters/wardens (e.g. sends them to deal with fake threat on a distant planet).
2.   Can hack “sensor scramblers” to gain vision of nearby planets (maybe only temporarily)
3.   Can EARN hacking points (e.g. a unique twist on hacking)
4.   Temporarily (10 min?) reduces AIP. Maybe raises AIP proportionally afterwards, temporarily.

Fast-travel Carriers—Special Flagships: (Inspired by the Guild Heighliners in Dune):

MASSIVE Flagships (Spire ship?) expressly for fast travel. Takes time to load ships (proportional to fleet size) & plot course (proportional to hops travelled) but travel itself is instantaneous. Unloading would also take time based on strength of passengers. No offensive capabilities (maybe drones?) Discoverable like flagships; 1 ALWAYS ? 3 hops from home planet. Can load other flagships (max 2) but not golems.



Healthy optimism


The beta has come a long way and has lots of the elements that attracted me to classic. While there’s room for improvement there’s nothing that would require a fundamental change in game design. I hope these suggestions (or something comparable) can be implemented without requiring a major overhaul. My basic contention is that the human players need more tools to wage guerilla warfare, ideally with respects to subterfuge.

Keep up the amazing work!



Title: Re: My comprehensive feedback
Post by: keith.lamothe on April 07, 2018, 09:22:42 PM
Thanks for the feedback! :)


Some things that could be done fairly "safely", both in terms of being close to the AIWC design and being fast to implement:

1) Add Mines
- already defined in the game, just currently not buildable

2) Add EMP Mines
- not implemented, but shouldn't be too bad

3) Add AOE Mines
- when triggered, they cause an explosion that damages more than just the triggering ship

4) Add Fortresses
- already defined in the game, just not currently buildable

5) Give Stealth Starships/Flagships a passive cloak-boost aura that builds a temporary cloak around nearby friendly units
- the mechanic needs to be adapted from the planet-wide cloaker, but not too bad
- the idea here is that the cloak would diminish over time outside the aura, and would not generally be as strong as a normal cloaked units'

6) Make different distribution-node types that are guard-post-like, and seed those instead of the vanilla ones. So like AIWC's guard posts, except you "inherit" them when you take over the planet, and they help you defend.


Things that would also fit design-wise (probably) but may need to wait until beta proper on implementation:

- Researchable upgrades for controllers
-- I'd initially intended to not do this, and instead focus on "how you spend your power is how you customize your planet", but with the removal of the resource converter buildings I can see that there's less and less argument for that being enough
-- This would need some additional UI to allow for switching the controller type
-- Incidentally, if the AI reconquers the planet, it gets your upgraded controller too :) Otherwise the game would have to reset it to the base type, and you'd have to switch it in again after taking it back.

- More Ark utility
-- one thought I've had lately is to make it immobile and Structure (a generally-weak defense type) to start with, but you get a line of techs like:
--- Fix the engines (can move, and thus hack)
--- Fix the hull (increase max hp)
--- Fix the armor (switch to Armor)
--- Fix the guns (short range)
--- Add more guns (medium range)
--- Add MOAR guns (long range)
--- Install deflectors (switch to deflectors, the best defense type)
--- Install Warp-Grid-Jammers (suppresses the AI's ability to warp in tons of extra reinforcements whenever it sees your Ark)
-- normally we can't do much of this sort of thing because of how stats are defined (making it much more flexible would be cpu-murder at this scale), but if it's a linear path of upgrading then it just replaces one unit with the newly unlocked type
-- anyway, the upshot is that you have to treat the Ark as a stationary King like AIWC to start with, but with major science investment you can use it defensively, and with even more science you can use it offensively. At that point you've got enough of an empire that the "the player needs to have something they have to defend" need has already been met, and AIP will be so high that you'd better know what you're doing if you leave home.


On fast-travel, I don't think we've really defined the problem yet:
- If the issue is "I don't want to wait that long, in wall-clock time", then just press Ctrl+plus a few times and the wait will be over very quickly. It's much more effective at that than AIWC's speedup feature.
- If the issue is "I don't want to wait that long, in game-time", then: why? Is it because the AI is able to hit you while you're out of position? That sounds working-as-designed to me ;) Though we can adjust the balance there.

I folks want to discuss fast-travel further I think a separate thread would be good, so we can nail down what specifically is bothering specific folks.
Title: Re: My comprehensive feedback
Post by: chemical_art on April 07, 2018, 11:29:39 PM
Zeusalmighty, I am very happy to see you back. I have never knew you by name explictly, but I knew there was "Transport/ surgical strike expert" from AIW 1 out there who I loved reading about and your expertise and perspective is always valuable  :)

My first thought is that your thoughts on the Ark remind me of the AIW 1 champion, which I feel is the perfect blueprint. The champion was always unique by the fact it could be like a phoenix in that it could never die; Having such a unit be assigned the role of the King as a core unit yet later lose its ability to resurrect due to <reasons> is really lovely lore in a sense (I'm getting Battlestar Galactica 2 vibes following this path).  It only makes sense it is the blueprint the King should be, a fallen unit who still retains the ability to upgrade over time. So I feel the ark should draw inspiration from it, and having the parallels would be a perfect Easter egg of sorts for players. A lot of fun lore could be done with having the ark once being the champion.

From a gameplay point of view, the Ark doesn't have to have all the story that went with the champion, but the concept of having player tech strengthen the Ark, in addition to direct player choices on the "tech tree" of the ark, would DRAMATICALLY address the worry that games are samey.

Sleep is calling, so the rest of your excellent post I will reflect tomorrow, but that tangent alone of having the ark grow with the player in a strong way would solve a lot of issues.
Title: Re: My comprehensive feedback
Post by: chemical_art on April 08, 2018, 01:37:42 AM
Yep way too late, just excited at the idea of trying to make the ark more like the champions of yore (which in AIW2 terms would be how they would have been viewed as). I know the code to try to drag out all the code to make slots and such probably isn't feasible so I expect the cold water treatment...but I can dream, can't I?

To clarify: The actual slots and all that jazz were just the method to having the unit grow with your tech and to your liking, I don't expect a 1 to 1 transfer or anything. Just the spirit of the concept makes me excited.
Title: Re: My comprehensive feedback
Post by: keith.lamothe on April 08, 2018, 08:24:41 AM
Any stat variances on the Ark's entity record itself would have to be accomplished by separate entity records in the xml that the techs switch between. In other words, you can have forking paths but not a "talent tree" in the normal sense as that would require an exponential number of entity records. So that's things like which defense class (structure/armor/evasion), speed, hit points, resistances.

Varying by which systems are on the Ark is much easier, as the techs can simply swap systems onto (and off, if necessary) the in-game entity. So that's things like guns, cloak, tractors, and gravity.

So the implementation could go something like this:

1) Add a science category for "Ark Core" that lets you get it mobile again, then choose which defense class you want to use, then add more hp, etc
- this would probably start with just one path at first, but forking paths are ok

2) Add a science category for "Ark Modules" that lets you add guns, gravity, tractors, etc (and cloak theoretically, though cloaking your King and hiding in the back of beyond might be a bit OP).
- in theory we could have the guns actually be put on as you research normal turret techs, but I'm not sure that would be intuitive. Perhaps the Ark-gun-techs would have the turret techs as prereqs.
Title: Re: My comprehensive feedback
Post by: chemical_art on April 08, 2018, 09:55:29 AM
Any stat variances on the Ark's entity record itself would have to be accomplished by separate entity records in the xml that the techs switch between. In other words, you can have forking paths but not a "talent tree" in the normal sense as that would require an exponential number of entity records. So that's things like which defense class (structure/armor/evasion), speed, hit points, resistances.

Varying by which systems are on the Ark is much easier, as the techs can simply swap systems onto (and off, if necessary) the in-game entity. So that's things like guns, cloak, tractors, and gravity.

So the implementation could go something like this:

1) Add a science category for "Ark Core" that lets you get it mobile again, then choose which defense class you want to use, then add more hp, etc
- this would probably start with just one path at first, but forking paths are ok

2) Add a science category for "Ark Modules" that lets you add guns, gravity, tractors, etc (and cloak theoretically, though cloaking your King and hiding in the back of beyond might be a bit OP).
- in theory we could have the guns actually be put on as you research normal turret techs, but I'm not sure that would be intuitive. Perhaps the Ark-gun-techs would have the turret techs as prereqs.

In my ridiculous head canon, I was thinking along the lines of:

Spire tech: Strongest defensively with strong single target weapons. Thinking of the spirit of multiple shields (so gobs of HP) and the spirit of the lance or plasma cannons, weapons who deals gobs of HP to one target at a time. These weapons and shields do crowd out economic power, however.
Offense: ***
Defense: ****
Economy: *
Special: **

Zenith tech: Medicore defenses with strong multi target weapons. This could give the ark armor for example for defense but doesn't give much extra HP but the weapons front is where it shines: It slugs multiple projectiles rapidly causing AoE damage which shreds light HP units.

Offense: ****
Defense: **
Economy: **
Special: **

"Human" tech: The most balanced and easiest to use. It has good HP (thinking of the weaker human shields) and its weapons while not flashy have no weakness. It's really used more for economics or otherwise behind the frontlines in terms of use which makes it easy to use.

Offense: ***
Defense: ***
Economy: ***
Special: *

Nenizul tech: The "other" one. This one can go multiple directions but here is one. You can have it cloak, have no direct weapons, and still support the battlefield (I did this with the drone launchers in AIW 1. Did I ever report that "bug"? I don't think I ever did.). In a straight up fight it really won't do anything special but the cloaking device and other unique opens up new possibilities.

Offense: **
Defense: **
Economy: **
Special: ****
Title: Re: My comprehensive feedback
Post by: keith.lamothe on April 08, 2018, 01:55:08 PM
I don't think we'd be going Spire/Zenith/Neinzul/Human with it at this point. The Spire content itself might have something along those lines later, though, when we get to it.

Each of the different backer-commissioned Ark "skins" could have its own progression options, though (later in development). They'll have to have their own entity records, anyway (they already do). And if someone really wants progression X with skin Y it's a very simple mod to swap the visual models.
Title: Re: My comprehensive feedback
Post by: chemical_art on April 08, 2018, 02:11:12 PM
Yes that all makes sense. I'm just having fun  ;)

For simplicity I'd say just start with as you unlock a turret give the Ark some kind of feature as well. Such a concept was already common in AIW 1 from all the units that had customization slots. Rather then get all fiddly with those menus, just have it add onto the Ark. On the intuitive front all you need to do is for when the ark is selected have a button next to the fleetship options that says "details" which when clicked can list all the various upgrades it has. That would be a nice synergy to give turrets and the Ark more value. If the idea gets traction you can add on other features as well.
Title: Re: My comprehensive feedback
Post by: zeusalmighty on April 08, 2018, 04:59:38 PM
Thanks Chemical =)

From a gameplay point of view, the Ark doesn't have to have all the story that went with the champion, but the concept of having player tech strengthen the Ark, in addition to direct player choices on the "tech tree" of the ark, would DRAMATICALLY address the worry that games are samey.

I loved the Champion from classic. The nebulae side quests were a cool diversion that could tremendously benefit your late game strength. I certainly don't anticipate bring the Champion back such as it was, but I agree that it could be a good benchmark to compare the Ark with. It just seems so natural to have it evolve over a course of a game in connection with your knowledge purchases, and I think the synergy of having both the turret tech and fleet tech also influence the Ark's development is a sweet idea that is also simple and intuitive to the player
Title: Re: My comprehensive feedback
Post by: x4000 on April 08, 2018, 05:36:37 PM
Thanks for all of this, everyone involved. :)

This thread did partially inform my recent post here: https://forums.arcengames.com/ai-war-ii/brainstorming-about-offense-vs-defense-and-refleeting-speeds/msg218579/#msg218579

But I also wanted to directly respond to a few things in this thread separately, too.  The conversations have overlap, but are definitely separate.

1. One of the things that I keep coming back to in my head lately, and this thread only reinforced it, is that there isn't enough going on.  There aren't enough layers.  We've streamlined out so much micro that the game feels simplistic compared to the first one.  And in general what is there is very good, and the streamlining helps what is there enormously, and I am not advocating for reverting to microing engies around or something.  I also think there are still some things that need MORE streamlining, so that will improve the experience that we have going on here in one way, while unfortunately digging the "too simple" hole deeper.

- Now this might sound insane, and maybe it is... but I think that we need at least one more vector of competition with the AI.  It's just something that has been striking me, of late.  The only thing that we compete with the AI on right now is tactics, which any AI can only be so good at, and strategy, which is hugely asymmetrical.  There's a defensive "minigame" as it were, and the streamlining has hurt that a lot.  The tech tree was always weak in the first game, and has only gotten moreso here -- see my related thread.

- So that got me thinking, though.  In what other ways could we do something to make science, hacking points, and AIP, more interesting.  Could we build "minigames" of sorts into those?  Those are already part of the meta, but what if those required some skill and planning and thought in the same way that setting up effective defenses in the tower-defense part of the game does?

- A "proper minigame" for science is, naturally, a more involved tech tree.  Visually we could do it, if we wanted to.  We could do something like Civ, or something strange like the Sphere Grid from Final Fantasy X.  You get science, and you choose how to advance along these paths, and it's interesting and fun.  It's another little screen to occupy some of your time and interest, and it's a change of pace from the rest of things.  You're not competing with the AI on this, per se, but in a lot of ways it does feel like you are.

- A "hacking minigame" seems like it could be more adversarial, and indeed more of a direct "turn based minigame," so to speak, in the same sense that planet defenes is a "tower defense minigame."  We already have the perfect "board" for this, too: the galaxy map.  What if hacking points, or some analogue, were used in a tug of war of sorts on planets here?  This would involve no units anywhere, it's all just numbers in the simulation and some new bits of interface on a "hacker view" of the galaxy map.  You'd "deploy hackers" anywhere that you can see, and have them do something for X amount of gametime.  The AI would have antihackers.  You win simply in a king of the hill sort of style: hold the place for long enough, by having more hackers than they have antihackers, and you complete the objective.  You set off an EMP at that planet, you stall the next wave, you get permanent scout intel there, you destroy prevent reinforcements there for X amount of time... all sorts of things could be done that don't require hooking up to specific individual units but mess with the AI and give you interesting defensive options.  Perhaps for every hacking point you get, you get 1 hacker to be used this way, and for every AIP the AI gains, they get 10 anti-hackers.
-- There was this mobile game with lots of little planets, and it had little arrow-like ships circling them, and you'd get more the longer you held a planet.  You could send certain numbers of them from one place to another, and at the other end they would fight the guys at the other side.  You couldn't move them while in transit.  My dad plays something that's a clone of that in VR.  Anyone know the game I'm thinking of?  I think that using similar visuals to show hackers and anti-hackers, we could have a pretty cool thing going there really fast, and it would be on my plate, not Keith's.

- The overall goal of having hacking work like this instead of the current way is that then there is another layer to the game, a third thing that you're considering, which is good.  It doesn't require micro, per se, but it does require attention and is something you might focus on while refleeting, or you might hold territory for a while in order to let your hackers do their thing from that space, then let go of the territory, etc.  Maybe a special building at that space lets you have extra hackers as long as you hold it, or just for 30 game minutes, whichever comes soonest, etc.  Since you can't compete economically with the AI here, given that the economies are unrelated, this sort of extra cat and mouse layer would make it feel more 4x-y to me.

- The minigame doesn't have to be that sort of thing exactly, but the idea is that it's something that is simple, easy to grasp, has a visual component, doesn't involve actual ships, does involve actual planets and territory, and gives you and/or the AI bonuses or penalties based on your successes there.  It gives you an outlet for things that isn't purely military, and makes the galaxy map have more than one meaning to it.

2. I'm partially wary of the Ark getting more useful, but I suppose it would be cool if it got more guns, etc.  It automatically has a "don't go into AI territory or you aggro them majorly" mechanic, so if we kept that but let the Ark become very strong, then that would be our roaming special forces unit all on its own, potentially.  And it WOULD work in enemy territory, which is cool, but it would just have a (very easy to digest in a lore sense) penalty to doing so, rather like aggroing a superterminal for too long makes sense in the first game.
Title: Re: My comprehensive feedback
Post by: zeusalmighty on April 08, 2018, 07:30:51 PM
-- There was this mobile game with lots of little planets, and it had little arrow-like ships circling them, and you'd get more the longer you held a planet.  You could send certain numbers of them from one place to another, and at the other end they would fight the guys at the other side.  You couldn't move them while in transit.  My dad plays something that's a clone of that in VR.  Anyone know the game I'm thinking of?

Was it "Eufloria"? I played this a while back but seem to recall it fitting your description. Can't remember why I stopped playing; it was enjoyable.

- A "hacking minigame" seems like it could be more adversarial, and indeed more of a direct "turn based minigame," so to speak, in the same sense that planet defenes is a "tower defense minigame."  We already have the perfect "board" for this, too: the galaxy map.

Hacking the AI, from both a lore and mechanical perspective, strikes me as the most fruitful area to develop a non-combative way of engagement. Particularly from the lore perspective, humanity NEEDS to exploit the fact that their opponent is an AI and since humanity is puny in comparison, they need to rely on other methods besides brute force. Hacking just fits the bill so nicely, but it's current iteration doesn't feel like a necessary stratagem to humanity's success but more like another thing to do when the opportunity presents itself.
 And I'm all for making the galaxy map more significant in planning strategy. Homeworld series does this sort of combination of perspectives superbly.
Title: Re: My comprehensive feedback
Post by: x4000 on April 08, 2018, 07:57:06 PM
Had to go back and look, now -- it was Galcon: https://www.galcon.com/iphone/

And I agree, having the AI be something you engage in a hacking sense, in a non-battlefield battlefield that affects your physical ships, munitions, and battlefields, seems to me like it would cover a lot of things.  It has to be simple enough that the learning curve is not insane, and so that we're not spending weeks or months in prototyping it.  I think it's doable, but would need to be just the right design, and would have to take advantage of a lot of things that are pre-existing but just under-utilized right now.
Title: Re: My comprehensive feedback
Post by: chemical_art on April 08, 2018, 08:51:53 PM
My thoughts on the idea of a hacking mini game:

https://www.youtube.com/watch?v=i1XZMJwDRjM

Which is to say I am not opposed I will flat out say I am opposed to it for inevitably any sort of experience that is not a part of the "main" game becomes just a skill that a player must learn. Which is to say that if you want to play in the big leagues, you learn how to play it. It isn't really optional. If you ignore it, you suffer for it. I didn't have a problem with it in the AIW 1 experience for it was an expansion. But as a part of the base game? No.

I remember in bio shock there was hacking, which was a simple puzzle game. The game designer had to make it that if you bought a simple upgrade you could win the mini game. While at the time I thought it was cool that my skill could determine if it worked, in practice it meant I would reload the game until I won every time. Since then I have gone toward the %chance of success route which still isn't ideal but better in the long run.

If it really is simple to win, it is just a chore. If it is a challenge, those who are not skilled in it are punished, and those skilled at it view it as that skill check you got to cross. It doesn't really at a core decide you skill at strategy, just a test of patience.

Now as resource to use as you wish, like hacking? No problem. But the instant it moves away from that I give the shoulder shove, because I don't want a game of yawndom. The story and defense mini games from the champions were not at all what I loved about champions, they were quite the chore. As much love I have given to champions this weekend, 0% of that came from the story bits and the mini games that went with them.  I hated them and viewed with the enthusiasm I do doing the dishes and laundry.

It could very well be I am over reacting to the concept and it is just a resource, which I would prefer. But if it is anything but that...smack!

Title: Re: My comprehensive feedback
Post by: eRe4s3r on April 08, 2018, 11:51:25 PM
Just as an idea, you could add resources or "thingymabobs" found in battle that go in a inventory like system after you conquered a planet that you can put in a hacking engine and depending on the things you put in that, and where you put them (maybe a complex grid that you can use min-max or just casual, you could influence something (you are the writer, you think of names ,p) with it. (or maybe it's just types of ressources where collect to a certain critical mass, or to certain masses)

Essentially like a magic system, except you don't make magic with elements, you initiate a hack with them. And battle and conquest gives you these elements

Downside: Requires a ton of new art and coding. Needs to be visually pleasing to use, and there should be a timer and seed attached so it can't be randomly rerolled or cheesed.
Upside: By definition works great in COOP, SP and otherwise. Also has the benefit of some emergent stuff going on (if you don't make it an element, but rather items, then you could also make them have unique effects and interactions...)
Title: Re: My comprehensive feedback
Post by: TheVampire100 on April 09, 2018, 03:50:52 AM
Okay, this is a small image, very sorry about this couldn't find something larger.
(http://mediang.gameswelt.net/public/images/200208/1a515917e3a44b0b6047c5c3b50e7b9a.jpg)

So, this is from Largo Winch, a very old computer game based on the comics with the same name. The game was mostly an adventure game with occasional fighting and hacking minigames thrown inbetween. Those hacking mini games were actually very well made and I just want to point out an inspiration.
Now I can understand that Chemical Art is concerned about a "hacking minigame" because this is somewhat different from AIWC. However, I never found hacking any fun in the frst game to begin with, it was basically "slap your hackign device here and wait until it is done. Oh, look, countermeasure, better have some ships nearby to protect it". It was just wtching a timer ticking down and eventually defend against some ships. This is in my very definition a chore.

I don't expect that you copy the system of Largo Winch, in fact, you can't. The problem is, AI War is still a RTS game, the hacking system of Largo Winch is turn based. Any turn based element involed would either have to pause the main game (which destroys the flow of the game) or it would go on (which means you always have to make sure that everything is okay before you go back to the mini game).
Now back to the mini game from Largo Winch, as you can see there are multiple nodes and some computers. The computer with the W-Logo is the starting point, your "avatar" spawns there and each round you can move him around and activate nodes. You have to activate all nodes before you can access to the next point, basically you build up a connection to the other PC. The pc with the skull is obviously the "opponent", it's a firewall that tries to keep you out, it sends out its own avatar (let's just pretend that's how hacking works) which has to move also on the nodes but also destroys all nodes you have activated. It can move freely on nodes you don't own (because it's own system) but has to deactivate the nodes from you one by one before it can move on (because you hacked that part). When it reaches you, it throws you out of the system and you have to start again.
Your goal is to connect all the nodes to the pc (not the firewall) and then access it before leaving from your own pc again. If you get caught on the way back out,t he firewall deletes all your files and throws you out again.

This sounds awefully complicated, so have  avideo that shows this better: https://youtu.be/DlWgBnzkt7M?t=36s

Now like I said, I cannot see this in AI War 2 but I can see a similiar system. Currently we use the Ark for hacking (instead of hacker units like in classic) which is also risky because you have t move the Ark to potentially dangerous territory). We could either leave it liek this or create dedicated hacker units again.
My idea would be that each planet has now "hacking" nodes even if there is nothing itneresting on it, you have to connect all nodes from your starting planet to your hacking goal, this sounds tiresome but this gives the player something to do while for example he refleets or prepares defenses for an attack.
You could either hack these nodes by sending yor Ark directly or sending replaceable hacker units. Maybe each node coss exactly 1 hacking point to activate?
As Ai progress goes on, the Ai gets Anti Hackers that go on the node and destroy some of your nodes before they vanish. In the video you can clearly see the timers on the firewall avatars that count down with each turn. The Anti hackers wil disappear after they destroyed X nodes connected to their system.
Hacked system are kept (you downloaded already the backup files or whatever) but the hacked planet has to be hacked again if you want to go further intot he hackign grid.

Now, there iS  aroblem of course: Map type. The map type can make hacking a lot easier ore harder depending on the map style you selected just like it can make the normal game harder or more difficult. However, I think with the hackign system I propose, the penality might be more sveere.
So my idea would be, the number of nodes depends on the number of connected planets. More connected planets? You have to hack more nodes on the planet to go on. So on map styles that are more linear and less branched out, you can easier hack through all the planets because there are less nodes to worry about. Planets with hacking goals obviously have always a high number of nodes to hack because there is something you want. Or maybe you just pay the usual hack costs on those planetg and keep the nodes as usual.

That's just my idea so far but it might be too complicated.
Title: Re: My comprehensive feedback
Post by: Magnus on April 09, 2018, 12:28:17 PM
My 2 cents.

I agree there is not enough going on at the moment. I started a short game on AI War 1 which I hadn't touched in years and I was immediately reminded of much more "stuff" is going on even in just a single planet.
I understand the idea about moving away from micromanagement and agree with it, it makes sense if the game wants to be a grand strategic, but the removal of micromanagement also implies a nearly complete removal of the tactical aspects. And if you're restricted to the strategic ones, this currently means: one offensive fleet going on some planet to bust the A.I., and turrets + tractors for defense.

You need, as a minimum:

 - the capability of more than one strategic initiative at a time. I know I sound like a broken record here, but picture this: you're playing an hex and counter WW2 wargame at the grand strategic level, and as the U.S. player you can only ever act on either the European or the Pacific theater, never in both at the same time. It doesn't work
- supply lines and logistic considerations are an essential part of operational and strategic games. There is currently very little of this in AI War 2: mainly which planets you take, and the limited capability of destroying Warp Gates in AI controlled planets.
- you can't act "behind enemy lines". No forcefields, no transports, non-operative cloaking (all cloaked units are decloaked nearly instantly by AI controllers and tachyons, which also have infinite range). This strikes me as odd in a game where you're supposed to engage in guerrilla warfare
- no strategical feint possible (only way would be to move the Ark on one planet to start hacking and move the fleet somewhere else; way too risky and hacking is too fast anyway)

If you don't solve these points at least, you will have, by necessity, to reintroduce tactical complexity to make the game interesting. In fact you will need to introduce some of it anyway even just to implement some of those (e.g. spec-ops deep strikes to significant enemy assets).

A few ideas:

- establish supply lines both for the player (most units can't act more than X hops away from your nearest planet, with exceptions being units specifically designed for deep strikes) and the A.I. (targets which if destroyed greatly limit the A.I. ability to operate in a certain radius until the line has been re-established)
- make cloaking (or some other mechanism like e.g. "disguising" units as A.I. controlled ones) effective again and allow it to be used to setup a strike even far away from player-controlled territory
- reintroduce some form of scouting, otherwise it's impossible to plan anything farther away than the immediately adjacent planets
- allow some mechanism to lure the threat fleet somewhere else


Luca
Title: Re: My comprehensive feedback
Post by: keith.lamothe on April 09, 2018, 01:29:40 PM
On cloaking, bear in mind that it's a very recent change that controllers had tachyon at all, so previously it was too easy to sneak around AI planets cloaked. More balancing can be done there, obviously.

There's also no cloak-boosting unit; the Stealth Starship would fill that role well.

Those two things would greatly change the picture.


On warp gates, what specifically is limited in the capability to destroy them on AI controlled planets? They were briefly autotargeted, but are already back to not being autotargeted.


On multiple attacks at once, what specifically makes it impossible? Just can't get a big enough fleet to split? Even with spending science to increase fuel production?


What sort of scouting do you mean? Currently scouting is "destroy sensor scramblers", which reveals the map in chunks at a time. Or do you mean actual vision? That's what Sensor Arrays were for, but they're currently not on the build menu due to incompatibilities with the new approach to mark upgrades; that could be remedied fairly quickly.
Title: Re: My comprehensive feedback
Post by: x4000 on April 09, 2018, 02:58:47 PM
Regarding the hacking "minigame," that is perhaps a bad way for me to explain it.  I view it as a "minigame" in the same sense that the waves are a "minigame," or the logistical challenges of AI War Classic were a "minigame."  They're a self-contained part of a larger whole.

I wrote this up for an internal email, and we've not discussed it yet internally, but I may as well post it here as well for feedback:

Why do we need hacking to be more of a thing?
In the first game we had a lot of things that were relating to logistics, and that was basically the third leg between offense and defense.  Logistics kept people busy, and gave them ways to feel clever.  Where do you allocate your engineers?  Which command stations do you make logistical ones, and by how much?  How do you route ships from the construction yards to the front lines?  Etc.  We have, happily, streamlined pretty much all of that out.  It was often leading to annoying micro, and it was getting increasingly streamlined-out during the progression of the first game from 2.0 to 8.0, anyway.  Things that were meant to be logistical challenges early on in, like having advanced factories be capturable and location-locked, eventually got stale and had to be changed to letting you build those mark iv variants wherever, or using those warp-gate pads that let us send things to far locations directly from the advanced factories.

The removal of this stuff was a really good thing, but it left a hole.  The game is now lopsided, and feels too simplistic because of it.

I don't think any of our existing designs cover this hole, and I feel like people will feel like this game is ultimately a dumbed-down version of the first one because of this hole.  My sense is that no amount of improvement of archers and soldiers is ultimately going to solve this issue.  People will get happier and happier... but you know what?  They're ultimately going to prefer the original game anyway.  This just has a Valley 2 versus Valley 1 vibe to me, it's too familiar.

We're very limited on time, though, so that means our options are limited.  And that's pretty good, in some respects, because we want to introduce something that is interesting to add to the game, but not so complex that now it's a mage-fest and soldiers and archers are falling by the wayside.  It needs to be just complicated enough to break up the tempo of the other two things, and all three components need to have lots of room for expansion in the future, but without adding something that feels too unconnected (like the nebula missions in the first game did, in my opinion, with the champions in the first game).  It also shouldn't just be yet more logistical MacGuffins like the spire quests in the main offensive terrain.  That can come later, but that's basically a variant on offense, not a third leg for the game to stand on (the original game had offense, defense, and logistics; this game has offense, kinda defense, and...?)

Hacking Alternative Idea
To me, the only idea I have that seems remotely feasible, is an addition to the metagame based on the concept of hacking.  We have the galaxy view already.  Taking a galcon-like approach with hackers and antihackers on there, in a specialized view, makes sense to me.  It's something that can replace the existing hacking mechanics, but accomplish all those same things in hopefully a more interesting way.  And it can accomplish a bunch of other things, too, such as allowing for debuffs or other things on enemy worlds, or buffs on our own worlds.  Bring in a bunch of hackers to aid in the defense of a planet by hacking the gravity in your favor for a while.

Then the hackers are basically augmenting either the offensive or the defensive power-based structures very directly, OR they are progressing the meta via means other than planet-hopping with your offensive fleet.  It also requires skill and planning: putting the right amount of hackers in the right place at the right time for the right length of time.  Blocking enemy anti-hackers, or tricking them out of position, or similar.  Just basically making the whole thing feel more like... a second battle front.

To me, it's always been the intersection of interesting and fairly simple systems that provides the greatest opportunities for player creativity as well as procedural magic to happen.  It's part of why I've been so happy with the way that the warden fleets and hunter fleets have evolved, and with the way the minor factions have been added.  Those allow for a bit of player creativity in dealing with them, and definitely contribute to the procedural magic effect.

But they don't do enough for allowing for player creativity.  I don't think that we can solve the fleet-ball problem, at least not with everybody.  But if an attack is inherently two-pronged because you're doing things with hackers at the same time you're doing things with physical ships... magic happens.  Suddenly the offensive game isn't just "put fleetball in place and win or lose."  You may have players going "dang I lost because my fleetball was there, but my hackers didn't finish doing xyz because my blocking hackers at planet A got overrun by antihackers from planet B, who came on over to the planet with my fleetball and interrupted me, so my fleetball was hung out to dry."

I haven't been to specific with designs on the hacking side of things yet, because I don't know what sort of buffs and penalties are possible to apply at a per-planet level.  Keith will have to answer that.  But, assuming that there are a fair number of things that we can do there, I can start providing some designs for the hacking game for people to tear apart and rebuild.

Techs And Hackers
The other area of meta is the tech tree, and that's something that is... tricky.  It may be that we need a lot more science, but then having a lot of that able to go to hackers, and really required to go to hackers, if you want to win.  The nice flexibility about hackers in my general design is that they can fast-travel very well, they don't exist on planets but only on the galaxy map, and they could have a bunch of techs associated with them if need be, which make them able to augment both soldiers and archers flexibly.

Hackers, TLDR
My thought is that basically these are logistical keys for doing all manner of cool things.  Yes, you absolutely wouldn't be able to play the game without them, same as you can't play without turrets.  But, as with turrets, this isn't exactly a whole new thing.  The general idea of the hackers here has a lot in common with the rest of the game:

1. A lot of it is just positioning numbers on a planet-by-planet basis, so it's strategic thought about terrain.  It's just adding more variance there.

2. This isn't some abstract minigame or door-cracking situation, so it's not disconnected from the rest of the game.  It happens right on the galaxy map, albeit in a specialized view of it.

3. There's a strong and immediate interplay between the battles and the hacking, because you can hack enemies to make them slower (I hope), or hack a planet to give yourself a range boost, or speed up starship construction at that planet, or whatever else.  A lot of these were things that fell under logistics in the first game.

4. There's a strong and immediate interplay between your long-term meta progression on the strategic front, and hacking.  You can hack to soften up targets, find targets, scout without traveling, create diversions, cancel or redirect waves, and all sorts of other things.  I expect.  So basically it's another resource that you're expending in order to make your main battles easier. 

5. The idea being that, on any serious difficulty, if you just leave this sitting around and never make any choices with it, of course you lose.  But it's not a grand new skillset to learn, not some disconnected "now we have go-kart racing mode" that you have to play, but instead it's a whole new toolbag that is very micro-averse that lets you do a lot of things that you could do in the first game, plus even more, but minus the micro the first game had.


edit: In general, obviously the hacking stuff I'm talking about here is pretty huge in terms of additions at this stage... but if people are on board with the general idea, I can write up a document that lays this out as just a few days of work, in the main, I think.
Title: Re: My comprehensive feedback
Post by: Magnus on April 09, 2018, 03:45:58 PM
On warp gates, what specifically is limited in the capability to destroy them on AI controlled planets? They were briefly autotargeted, but are already back to not being autotargeted.

If I need to bring a fleet strong enough to destroy a warp gate, I might as well raze the whole planet. The only decision left is whether I want to conquer it or not, based on the production advantage it will get me against the increase in AIP.
I can't blitz in this game. Nor can I setup an hidden strike. So destroying the warp gates becomes simply another necessary step in me systematically stripping the A.I. of everything it has in all its planets (except the controller if I don't want the AIP increase) up to the core planet I want to reach.

Quote
On multiple attacks at once, what specifically makes it impossible? Just can't get a big enough fleet to split? Even with spending science to increase fuel production?

It's not a problem of caps, it's a problem of control.
Small ships in AI War 2 are incredibly fragile but also incredibly easy to replace; they die much faster than in the first game, but they also get rebuilt much faster. This means the only tactic possible is to send at least a builder on loop-building configuration, and the related death ball.
Due to how rebuilders work, if I have two fleets fighting at the same time on two different planets, I don't have any control on which rebuilder will rebuild any given lost ship; it's basically random on whichever queues it first. This will make it so that my initial ratio of ships in fleet 1 versus fleet 2 is subject to random changes. It's why I proposed that additional setting on rebuilders to only (re)build ships lost on their same planet.

Quote
What sort of scouting do you mean? Currently scouting is "destroy sensor scramblers", which reveals the map in chunks at a time. Or do you mean actual vision? That's what Sensor Arrays were for, but they're currently not on the build menu due to incompatibilities with the new approach to mark upgrades; that could be remedied fairly quickly.

If I want to setup a spec ops deep strike, I want to know where to send it first. If I'm limited to the nearby planets as my theater of operation, I might as well not bother with spec ops and keep going with deathball-based systematic A.I. planet-and-resource stripping.
How do I play guerrilla-like if I'm mostly blind and to see more of the map I need to conquer territory (or to have a fleet strong enough that it would be able to take territory anyway)?

Both this and the warp gate problem have the same root: it is theoretically possible to send a fleet and only destroy the sensor or the gate, but if you've gone to the trouble of assembling a strong enough fleet and moving it on the target planet, why bother targeting only those? Might as well kill everything important while you're there (which, btw, right now essentially means "all the A.I. fleet ships which will otherwise be added to future threat", since power nodes and other stuff do little to nothing and conquerables require either planet capture, or hacking).
The only reason to leave something important alive is if you need to recall the fleet in a hurry due to it being needed to fend off an attack. And we're back to "OMG I need to move back my fleet ASAP" which is annoying and not at all fun.

Basically, what I'm saying is: right now there's only one real way to play this game: build one single death ball, conquer the planets one by one until you have enough resources to never hit the cap, systematically strip the other planets of everything short of the controller to clear a path to the A.I. cores, and pray your turrets are enough (which is almost never the case) to not force you to recall the deathball and waste hours on playing the whack-a-fleet game with the A.I., with your fleet chasing its fleet all over your empire while Yakety Sax is playing in the background.
Title: Re: My comprehensive feedback
Post by: etheric42 on April 09, 2018, 04:20:14 PM
It's not a problem of caps, it's a problem of control.
Small ships in AI War 2 are incredibly fragile but also incredibly easy to replace; they die much faster than in the first game, but they also get rebuilt much faster. This means the only tactic possible is to send at least a builder on loop-building configuration, and the related death ball.
Due to how rebuilders work, if I have two fleets fighting at the same time on two different planets, I don't have any control on which rebuilder will rebuild any given lost ship; it's basically random on whichever queues it first. This will make it so that my initial ratio of ships in fleet 1 versus fleet 2 is subject to random changes. It's why I proposed that additional setting on rebuilders to only (re)build ships lost on their same planet.

This gets resolved by a UI feature that has been outlined but not implemented (or approved for implementation until after EA starts).  Control groups where you can set the quantities of each ship you want in them and they reinforce up to that limit.

Chris, do you want to move this up as a higher priority to address this issue?
Title: Re: My comprehensive feedback
Post by: x4000 on April 09, 2018, 04:36:04 PM
It's not a problem of caps, it's a problem of control.
Small ships in AI War 2 are incredibly fragile but also incredibly easy to replace; they die much faster than in the first game, but they also get rebuilt much faster. This means the only tactic possible is to send at least a builder on loop-building configuration, and the related death ball.
Due to how rebuilders work, if I have two fleets fighting at the same time on two different planets, I don't have any control on which rebuilder will rebuild any given lost ship; it's basically random on whichever queues it first. This will make it so that my initial ratio of ships in fleet 1 versus fleet 2 is subject to random changes. It's why I proposed that additional setting on rebuilders to only (re)build ships lost on their same planet.

This gets resolved by a UI feature that has been outlined but not implemented (or approved for implementation until after EA starts).  Control groups where you can set the quantities of each ship you want in them and they reinforce up to that limit.

Chris, do you want to move this up as a higher priority to address this issue?

I strongly doubt I'd have time for that sooner than EA, and that's the sort of feature that only advanced folks will use, anyway.  While I'm not averse to having the feature, I feel like we need to have a default non-advanced flow that is comfortable for people who don't dive this deep into the UI.  That should be bumped up in priority, and doesn't really involve me. ;)
Title: Re: My comprehensive feedback
Post by: zeusalmighty on April 10, 2018, 12:26:41 AM
  I don't think that we can solve the fleet-ball problem, at least not with everybody.  But if an attack is inherently two-pronged because you're doing things with hackers at the same time you're doing things with physical ships... magic happens.  Suddenly the offensive game isn't just "put fleetball in place and win or lose."  You may have players going "dang I lost because my fleetball was there, but my hackers didn't finish doing xyz because my blocking hackers at planet A got overrun by antihackers from planet B, who came on over to the planet with my fleetball and interrupted me, so my fleetball was hung out to dry."

This sounds so good to me. Imagining playing the as it is with this added feature makes me excited. But I realize that this would be a big undertaking this close to EA. I can't help but think it's worthwhile though, for the reason that fits the bill in terms adding the necessary depth that will make each new campaign unique as well as satisfying a key premise of the game--waging GUERILLA war against an AI OPPONENT.

Perhaps we need a new thread just for this question--what role should hacking have in AI2?
Title: Re: My comprehensive feedback
Post by: x4000 on April 10, 2018, 09:43:39 AM
Sounds like that would make sense, yes.  I might need to just make a document and have done with it.
Title: Re: My comprehensive feedback
Post by: x4000 on April 10, 2018, 11:17:21 AM
All right, here's a document explaining the design in detail: https://docs.google.com/document/d/1ecRBCt-pqZYT06ndFg9NkT0UM1b4CSTGBtzuHqZzdYc/edit?usp=sharing