Arcen Games

Games => AI War II => Topic started by: x4000 on June 05, 2019, 04:20:20 PM

Title: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: x4000 on June 05, 2019, 04:20:20 PM
Okay, so that's probably a pretty provocative title, but I hope I've laid out my case pretty well here: https://docs.google.com/document/d/108hBRHf3viOgyIS-Wp6l9bBMWPqsjkS1uUjNKrv-2CM/

It took typing all that up, plus all the stuff I typed up to MatthewYCR in our discussion on mantis, for me to really solidify all my thoughts.  I think... this may be a one or two day sort of job?  If that?  Under the hood, not a lot is changing, which is good.

The core thrust of the document is that metal (and to a lesser extent energy, the way that energy is used right now) have really been messing with my vision of how fleet economies would work.  Basically I've been thinking of fleets as these independent things that you have, and once you have them they're yours and its a matter of how to best use them.  Not IF to use them, or what PART of them to use.

I make a lot of comparisons to Chess and Risk (as I typically do), but also to the economy flow of traditional RTS titles and why this is divergent.  Essentially your economy IS combat (and by extension AIP), which is similar to Risk and Chess.  Making you have a second (metal) or even third (energy) economy on top of all your units after that is just egregiously complicated for no real purpose.

And it's definitely at odds with a lot of my goals from the original unit construction and fleet design doc, where I'm wanting players to see what they can get up to with units that they might otherwise find useless.  I also want refleeting time to be a time where you can be thinking about your fleet DESIGN (via some upgrades or tweaks or where you position them), not a time where you have "doing your taxes" busywork.

Interestingly, this harkens back in some ways to some of the very earliest ideas I had for AI War 1, back before 1.0 of that game came out.  But this is vastly more refined, and is in most ways just an incremental change to the design as it currently exists.  This is basically trying to get you to be thinking most of the time in the mid-level strategic mindset, rather than getting bogged down repeatedly with individual trees and bushes in that forest. 

The fleets design so far has really pushed things in that direction to an excellent extent, but the metal and energy design are holding it back at the moment.

The document also includes some ideas for potential future additions to the game, probably for post-1.0, but basically to illustrate that this in no way gets rid of the good kind of complexity that you might want to have.  It gets rid of busywork, and in so doing kind of opens the door for a more clear view of your real strategic challenges, and also opens the door to other potential complexities that would have previously potentially been too overwhelming.

Most likely, this is what I'm going to work on next, then circle around to the fleet management screen.

Best,
Chris

Discussion also here: https://steamcommunity.com/app/573410/discussions/0/1638661595043362159/
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: BadgerBadger on June 05, 2019, 04:47:48 PM
This makes me really leery, in part because I like games with economies, and it sounds like this is removing most all of the economics.

I'm not sure exactly what problem you are trying to solve. I worry that  you're basically trying to automate the player out of the game, like what happened pre-pivot. If I have basically no need for Energy or Metal then I don't need to have an empire at all, just some combat factories and a heavily defended home command station.

Also, removing metal is a giant giant buff to the player. If I have infinite metal and 2 or 3 fleets then I can pretty casually kill all AI defenses up to 2 hops from a factory. I just leave the fleet centerpieces off the planet and send an endless stream of ships over (and now I really want the ability to rally units to a location again.....) A ton of new balance work would would need to be done, including probably reimagining how the AI should work to deal with the player suddenly being able to win attrition wars.

Edit: I just tried an existing save game battle for fun and just used the "debug: more metal" button over and over, and it was really dull. I just attritioned them out, since there's no cost to having a total fleet wipe. I can just rebuild everything almost instantly. I'm not sure how to prevent the "attrition is OP" strategy without the cure being worse than the problem of metal.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: Lord Of Nothing on June 05, 2019, 06:09:56 PM
I agree with a lot of what badger is saying, to be honest, so I won't bother repeating that.

I did have one slight thought that, under this model, perhaps factories need to be more limited and on a per-planet basis, letting build time and capacity act as a less complicated, more streamlined, pseudo-metal.

However, I do have a somewhat more serious point. It is perhaps not my place to raise this, but I'm going to do it anyway, because this is something that's been bothering me since you announced the fleet rework:
Anything along these lines is going to radically alter the game's balance, and it's flow, as Badger said. Now, perhaps these are solvable, and perhaps they are not. Perhaps we're both wrong, and it would make a better game at the end. But what is true is that iterating out a new balance and flow again is going to take time.

This makes me very worried that you're sliding towards a stars beyond reach situation again, where you simply never reach a release version of the game.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: Draco18s on June 05, 2019, 06:25:22 PM
I....am not opposed, but I definitely respect Badger's concerns.

"Standard" RTS games (Age of Empires, etc) weren't so much about controlling your flow of gold/meat/wood/whatever, but the fact that those materials were fundamentally limited in some way. There were only so many trees on the map, only so many gold ore nodes, and so on.

While it was true that some resources were infinite (food) and you could nominally trade for the remainder (with a market, at insane prices) it was not viable to just produce oodles of food and convert to gold/wood/whatever after the nodes ran out, as generally speaking if there was no more on the map, that also meant your opponent was facing the same scarcity you were and if that meant that gunfights devolved into sticks and stones using basic infantry in order to establish supremacy, then, well, it happened.

There were also defensive structures (ie walls) that were time consuming (but not exactly resource expensive) to build, but that once constructed were trivial to maintain and defend (4 archers, costing only food and maybe a wood could defend a 12 tile long section of wall from an arbitrarily large force of basic infantry pretty much indefinitely) and forced your opponent to invest heavily into those expensive gold-consuming units in order to knock down in a hurry (and they didn't generally need to take out every wall, just three or four sections so their infantry could get through).

AI War doesn't have this paradigm. Both in that resources are infinite while controlled (just time limited, as if you could only ever have 4 workers gathering stone, sorry mate, can't have more than that!) and that defensive structures are often weaker and more expensive than their offensive mobile counterparts, with very few exceptions.  For example, fortresses. I've often had to spend four hours worth of resource income over the span of two hours to build one, only to have the AI come knocking and break it in under five minutes. Oh, they have bombers? Thirty seconds.

Chess: AI War will never be fun in this paradigm. It can't function. It would be like pitting one player with the standard set of units against another player who has 14 boxes of spares. And three more boards and two kings. And a cat.

Risk: AI War can't match this paradigm. Risk functions by giving units for both territory controlled and territory most recently conquered. AI War will not be able to doll out reinforcements based on the fact that I took a new system in the last 30 minutes. It wouldn't make any sense. It also couldn't just award me reinforcements for controlling 7 systems because Risk is all about controlling a specific set of territories, not your total (although there's often a small bonus for total, most, or similar).  It might be the kind of paradigm that matches most closely with what AI War wants to be, but the mechanics don't translate.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: Democracy? Democracy! on June 05, 2019, 07:26:59 PM
After talking about it on the discord, one of the main concerns we have with removing metal is the removal of the challenge of assaulting worlds and defending.

The metal income is a kind of limiter in how much you can commit to an assault on a planet. Let's say for example there is a mk5 planet in my way and I'm forced to go through it/really want a capturable on it. I can only throw so many ships at it before my income starts tanking and ships start to only trickle in like drops of water instead of coming in like a river. Therefore I'm forced to retreat and let metal build up again, while the AI can launch a sizeable reprisal wave. And since I'm low on metal, if I made the mistake of not reinforcing where the AI can attack, it can just easily roll over me since I have no metal to build defenses.

If I had infinite metal, I could just simply keep pressing the attack with my river of reinforcements, and eventually win through attrition, since the AI can't reinforce as fast as i can. When the reprisal comes, i can just build turrets and station keepers fast enough that the AI doesn't raid in time. Even wormhole invasions could be stopped by sending a transport flagship, to hold the line while turrets construct in the background. The AI is supposed to win at games of attrition, considering it has tons of stuff being produced outside this galaxy.

Another thing pointed out is that with flagships providing the "spawn point" of strikecraft, the only way to stop the tide of human ships would be to cripple the flagship. And crippling the flagship can be difficult if the flagship is hiding on a different planet. With infinite metal, one can simply group all their strikecraft and send them into a planet, wait for that group to die, rebuild and send the next group. The AI would have to be wired to target flagships and hunt them down, which can be challenging for the ai considering most flagships pack either considerable heat (golems) or a gimmick that makes it hard to kill (zombifying allies, spawning even MORE ships, self healing, etc.)

I don't wanna pull a "back in my day" thing, but AIW:C was a very difficult game intentionally. And it was fun, since other RTS didn't have an AI that already started with everything, and gave a good challenge to people who want to PVE. With the removal of a limit put on human players, it just feels like I'd be playing as an AI, since i have unlimited resources (OK still limited by energy and ship caps) to work with. I'm sure Kesseleth or RocketAssistedPuffin can explain better than I can.

TL;DR infinite metal is buffing human players so they only loose on either difficulty 10 or if they have no idea what they're doing.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: Kesseleth on June 05, 2019, 07:52:01 PM
I do not have an elegant way to say it, so I am forced to put it in a blunt way instead: I don't like it much at all.

I actually don't mind the energy changes, which, based on discussion in the Discord chatroom, actually makes me closer to agreeing with this than most people. In Classic, energy management for me consisted of getting close to max energy for some reason or another, pausing the game, and trashing some turrets on an out-of-the-way planet that didn't need them anymore. It never much affected my decision making at any point in the game and I would be more or less ambivalent to seeing it go. Only ambivalent, mind.

But metal? I don't agree even remotely. You refer to metal as causing problems by forcing you to pay for ships twice. That... kinda makes sense to me? However, metal and AIP deal with different facets of the game and serve two very different purposes.

Here's the best way I have to explain it: There are, broadly speaking, two major aspects to AI War: Fleet Command and its sequel - the strategic aspect and the tactical aspect. I hear the Total War series has these aspects as well, but I have never played any. From what I hear, though, in Total War there is a very clear dividing line between them. I'm in tactics mode! Oh, now I'm in grand strategy mode, put away all those soldiers and get out your calculator, it's time for trade routes. One thing that made AI War so interesting was how it blends the two very well, but removing metal would eliminate this.

AI Progress is a strategic stat. When you look at the galaxy map, deliberating which planet is the most appetizing to conquer, AIP is a very important stat. Is it finally time to take the superterminal? Or, hmm, this planet over here has a very nice fleet on it, but getting there means I have to go through some Mark IV planets... Oh, but along the way I can take out a data center! These are the sorts of considerations I have as a player when thinking about my strategy. Metal doesn't really matter to me when I make these grand, sweeping decisions, aside from "is my metal flow enough and if not how can I get more".

But when I'm a couple hops into AI territory, embroiled in a fierce conflict with the Warden Fleet, metal becomes far more important. Do I have enough? Can I build these turrets? Are my central fleet ships in too much danger? Is it time to retreat back a planet or can I keep holding out? When in a battle, AI Progress becomes almost irrelevant. AIP chose the difficulty of the encounter, but it's up to my wits and my metal reserves to resolve it. I care about the AIP here no more than I cared about metal when I decided to make this journey.

So, put another way, we are indeed paying both metal and AIP. As well we should, because the payments are for two completely different parts of the game. The AIP is the scale of the forces the AI has arrayed against me, and in a strategic view I need to keep that at a minimum. The AIP is the measure of the armchair general, giving grand commands to move his army. But metal is a fragile sword and shield, a reserve to block each swing of the AI and a blade to strike back once boots are on the ground and the arrows start to fly. It's reforged between fights, indeed, but in the heat of battle, what matters is not the grand strategy, but the survival. At this point, all you can do is grit your teeth, hope your general knows what he's doing, and fight.

To eliminate metal is to break this relationship and, frankly, gut a large portion of the game.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: RocketAssistedPuffin on June 05, 2019, 07:58:06 PM
I feel like I have racing thoughts, so this might be a bit confused and/or forgetful (and I apologise for that), but I will try.

I really don't like much of this.

To me, this feels like removing bits from the game and kind of...dumbing it down? Why do I need to be careful with my units? They're free, I could spam units out at an AI planet far faster than it could reinforce, faster than the Warden Fleet could.

It feels like there'd be almost no permanence anymore. Currently, we cannot lose things like a Golem or a Zenith Power Generator permanently. We cannot lose a Fabricator, we cannot Corrupt Design Backups, we can't destroy a Special Forces Base to weaken the Special Forces income. With this, once I've captured a planet with a Fleet, why keep it? It holds nothing else of value, since most planets don't actually hold more than 1 item. Yes, the AIP Increase from a ZPG being destroyed is something, but that feels so little, because I still have the ZPG.

I don't know how we're meant to lose with this. We can barely be permanently damaged - the only path to defeat is just one big strike that kills the Home Station. If I don't need to really hold planets for more resources, my behaviour at this point would be just castle up around my Homeworld, as it's the only thing that matters anymore. Holding a ZPG isn't worth having to keep a Citadel or Battlestation there anymore, since under the new Energy system, all it'd grant me is the ability to power a Forcefield or something like that at home...or I could keep that Citadel and/or Battlestation at home and not pay the AIP for the ZPG.

Losing a Command Station doesn't really hurt (outside of Economics). A Fleet Wipe barely hurts, and if Salvage gives EXP...I might want that? Fleet Wipe = no cost to me, but the Reprisal gives me a permanent power increase in EXP. (Speaking of EXP, if it's gained entirely by just fighting...then that feels like a return to the "optimal thing is to slowly scout the entire galaxy, bit by bit, as the first thing you do" problem - just sit there on low AIP and let the Stations rank up. Hopefully I'm wrong here).

Back to metal, if it's free and we can spam ships endlessly, then gameplay feels like it's now just "spam ships at planet, get Fleet, spam even more ships at next planet, get Fleet, repeat". Since almost nothing can be permanently lost, what do I have to think about? Sending the correct units so I can win the battle, do some minor tactics to preserve my ships? Why bother, it's work for nothing. It's...now busy work. We'd also lose things like Neinzul type ships being dirt cheap and spammable as their selling point and unique trait.

If metal is free, why do we have personal shields? A big point of those is they recharge for free, or can be sped up for really cheap (but still a cost). If metal is free, then repair is free and shields lose that point. It might still work for non-player factions, but I've not cared about that so far. At this point, it's now just a downside, since it means some units kill you far quicker.

Like Badger said, this is a fairly big buff to the player, and...kinda in a way ruins what I think AI Planets are: Roadblocks. Time is needed for the AI to send waves, to get CPAs, etc. Kind of like in Starward Rogue, the enemies and bosses need time to get their patterns out, they have to survive to do that. If I can spam ships forever through planets, then I don't think they'll really be proper roadblocks anymore, and I could likely win the game in a much shorter span of time. Boost my Fleet so I can have a chance at attacking an AI Homeworld? Meh, send stuff into the grinder, I'll wear it down eventually with no weakness caused in myself. Reprisal doesn't matter, since I bunkered up hard around my Homeworld and hey, I benefit from it? I can just keep doing this and I don't get any temporary weakness or anything, really.

Another behaviour that I'd probably do is just bring along a Citadel or Battlestation, and freely set up turrets all over enemy planets. You can do this currently, but you at least have metal keeping you in check.

Energy change feels like the bit I'm least bugged about, but the point above of the ZPG and Citadel/Battlestation and AIP is still here, and...kind of makes it still less likely in general for me to ever capture something for energy. I get the idea of Metal, not having that old..."Netflix Time" and the like, but I just...feel like it'd change the game into something I would never enjoy playing.

I would point to two things in particular that exhibit something like this: Command and Conquer 4, and Zeratul from Starcraft 2s Co-Op Mode:

 C&C 4 has no resources at all, units are built for free up to a generic total popcap like many RTS games have. There's not much use for map control, there's no permanence, almost no delay in your actions, almost nothing to do to the enemy except kill them outright. This game effectively killed the series for over 9 years, with still nothing really new in sight.

Zeratul has gas be automatically harvested, his upgrades are automatically researched, and his units automatically do their abilities. The units also cost so much, you barely have to actually order any to be built, and your supply cap is half that of normal as well. He is so incredibly boring, because you really have been automated out of the game.

That last bit I'm worried about now...truly pointless things like Science Labs being needed to gather it I'm fine with, but without metal...the game just becomes crank the speed up and spam units, which whenever that state is reached for me in a game, instantly dooms that save/run/whatever to boredom.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: AnnoyingOrange on June 06, 2019, 05:08:01 AM
My general impression is that this is a bad idea, that it would exacerbate the problems the game currently has, and that finishing what's already in game is more important and will give a better view of the state of the gameplay.

Let's go in detail.

First, the energy suggestion: the role of energy right now and in the past is exactly that of imposing a secondary cap on things like turrets, so that the optimal strategy is at least a bit more interesting than "full caps everywhere" and so that there's a strong pressure to allocate your defensive assets efficiently instead of keeping hundreds of turrets on every plane "just in case".
Making turrets cost no energy would result in incredibly boring battlestation micromanagement every time you decide to move a chokepoint (and battlestations are already really boring and micro-intensive compared to the old system in my opinion).
Making strikecraft cost no energy wouldn't stop me from wishing I could simply not build some of them (such as telsa corvettes in an otherwise long range fleet): it would simply make mediocre fleets a bit less counterproductive to capture, but that's better fixed by fleet controls.

About metal being removed: but for what purpose?
If my main attack fleet has a metal sink ship in it (such as raid starships), I'll still spend a lot of factory and engineer work on rebuilding those metal sinks (instead of building useful ships) even if the metal itself is no longer an issue: again, this is better solved by fleet controls being implemented so I can stop construction of things I don't want to construct.
The removal of metal cost increase on mark increase already felt like a bandaid solution to economy issues, and the inability to upgrade metal collectors certainly didn't help as that was the easy way to let players balance military and economy.

About the fleet EXP via salvage suggestion: that's even less interactive and interesting than classic scouting, dare I say it might be even less interactive and interesting of my 120 plantet full conquest game in AIWC (and that was 90% AFK waiting for spire civ leaders to lower AIP).
Of course, that could be wrong, because fleet EXP is currently not used in game and so many important units (engineers, forcefields, command stations...) are stuck at mark 1 no matter what: once that changes the gameplay will probably be barely recognizable compared to right now.

I agree about distribution nodes, as of now they're kinda useful occasionally when the stars align, and the idea of hunting both data centers and distribution nodes sounds good.
I also agree on mercenaries, as of now they're a very expensive and low impact option even compared to their endgame metal sink role in AIWC.


TL DR: finish the fleet controls and the fleet EXP mechanic before considering massive changes to the core mechanics, the mercs and distribution node changes sound good but waiting on them won't hurt.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: WeaponMaster on June 06, 2019, 10:35:32 AM
My thoughts are definitely more half baked than anything people have brought up in here as I don't tend to think as much about mechanics when playing games.

First thing I'd like to say: I don't think any of my losses in AI War 2, and maybe only 25% of my losses in AI War 1, came about because of a lack (or mismanaging) of metal. That's not to say that metal plays no part in how the game plays out at all, but I never really think to myself "Damn if only I just had more metal, I would have survived". The three ways that lead to death in my case are:

1. Backdooring - When you send all of your mobile fleets to that sticky Mk 5 planet that's two hops from your territory, and 40 strength of threat ships bum rush my home world. I would consider this positive gameplay, you shouldn't be able to just mindlessly steamroll through planets, hoping that turrets can stop threat fleets from rolling you down in return.

2. CPAs - Less scary than in AI War 1 where at level 7, 100 AIP there could be over a thousand ships, but still plenty deadly. The AI code from Badger that keeps threat ships on the move so you can't corner it and take it out in one sweep can make it very hard to deal with. The CPA can be an exciting event if your territory is large enough that the AI makes all these tiny attacks before the main force groups and breaks in, but if they just sit outside your territory running away the whole time... not very fun then. Good gameplay, but it might need some way to encourage the AI to stand and fight for a while during the CPA.

3. Reprisal waves - AKA Big Ball of Death, AI Edition. Probably the least interesting way to die. Reprisal waves have such a long startup time that I don't feel metal plays a factor in this kind of death at all. So you're at 70 AIP and you just lost all 3 fleets on an Mk 5 planet assault. It'll take 3 million metal to rebuild (not counting the metal to get flagships up to 10% hp), but now that the metal cap goes into the high 3 - 5 millions that doesn't matter at all. The wave gets scheduled for 10 minutes later, more than enough time to rebuild everything except for maybe a few frigates, and then the only thing left to do is pray your 40 strength can stop their 40 strength.

Maybe it's just the way I play, but I tend to only make moves when I'm approaching the metal cap. In that sense, the limiting factor for how well I can defend myself is always positioning fleets correctly and the replacement rate of crafts. Sure I sometimes run out of metal and it slows things to a halt, but its almost always during "peace time", like when claiming a golem right after capturing a planet. There'll either be no planned waves, or a planned wave in a few minutes, and it just ends up slowing gameplay to a halt.

Assuming you don't anger the reprisal wave gods and die instantly, my peace time ends up being 5 minutes of planning turret defenses (scrap old turrets and move battlestations around to cover the new territory), 5 minutes of rebuilding and repairing my fleet, and then 20 minutes of just sitting and watching turrets build and golems/arks be claimed while in metal deficit. It's hardly engaging to have to babysit a new territory while defenses slowly build.

I'm rambling and I'm sleepy, so I hope my text wasn't too useless.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: x4000 on June 06, 2019, 11:07:03 AM
There's a new note in here: https://docs.google.com/document/d/108hBRHf3viOgyIS-Wp6l9bBMWPqsjkS1uUjNKrv-2CM/

I've been convinced not to do this, but have some thoughts.  Thanks for all the feedback, it was super useful!
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: RocketAssistedPuffin on June 06, 2019, 04:27:20 PM
Hey, sometimes there’s a unit that is crappy and nonoptimal but it’s so much more fun to try to find an interesting way to use it rather than just turning it off because it doesn’t fit with your optimal economy.
How do we deal with this problem?  I dunno.  We could make the crappy unit better, but then we will sometimes get into situations where all units are basically interchangeable.  And sometimes the problem isn’t that the unit design is bad, it’s  that the unit is just un-upgraded.

A big question. A while ago when I poked Classic again, I had found a number of units there blended together, just like in 2, contrary to memory. I would really like to fix this...somehow. It bugs me greatly to see things like "I never want to build this unit, ever".

I think it's...mainly, we don't provide reasons to use it. There's not really any scenario in which a Raider is worth using by a player. With Fleets, you cannot deepstrike them on their own to hit something deep, AI units don't kite outside of Warden/Hunter (and kiting doesn't seem to do much anyway, for some reason). They're pretty much just "generic shooty unit that happens to go slightly faster than most".

Etherjet is similar. Woo, I have Tractor Beams!...Okay, but it's just better to shoot the things instead of moving them since they die so easily.

Maybe if mark 1 units are a liability, they wind up getting cheaper as the other units level up.  It used to be that there was a difference in cost in the mark level of units, but that turned out to be punitive in this game because your units would get too expensive and your economy would tank  just because you unlocked a tech. So unlocking a tech was a dangerous thing.  But the side effect of making all mark levels cost the same has been that now the low mark units are a liability in the late game, rather than cheap chaff that you can do interesting things with.

It...feels at first thought that the difference between Marks is just too great. Double the damage, then double the durability, so you last twice as long, doing twice the damage, for just Mark 2. Now that Mark 5 isn't restricted, and Mark 7 is possible with no downside like increased cost...yeah. The first game was balanced with that in mind possibly? Things had to be killable with normal, low Mark units. Here, things have to stand up to hundreds of Mark 5-6-7. But the Marks is kind of an important thing to the game so...not really a useful line of thinking.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO
Post by: Draco18s on June 06, 2019, 06:41:16 PM
There's a new note in here: https://docs.google.com/document/d/108hBRHf3viOgyIS-Wp6l9bBMWPqsjkS1uUjNKrv-2CM/

I've been convinced not to do this, but have some thoughts.  Thanks for all the feedback, it was super useful!

Sure thing, Chris.
I still don't know if what we have is "the right thing" either, but at least its issues are one of player engagement, rather than actually broken. The problems with it are very valid, but finding the solution is going to be a lot more difficult than simply removing it.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: AnnoyingOrange on June 07, 2019, 06:14:29 AM
Some ideas about your new note.


About crappy and non-optimal units, changing the economy is hardly going to improve them in terms of player fun.

My bandaid fix would be to let players transfer units between fleets at the cost of some fleet EXP, to get rid of the whole "this ship doesn't play nice with the rest of the fleet" issue.
This wouldn't fix other sources of crappiness and it would require some additional work to stop players from transferring every unit to a single giant deathball fleet, but it would make it much easier to do things such as creating a good deep strike fleet or a good cloaked fleet and it would allow players to get more mileage out of the special traits of any unit.

A much stronger improvement (at a much higher implementation cost) could come from advanced fleet and ship controls, such as pseudo-rally points, target priorities, hold position, pursuit with return to position, and others: things like "all bombers in this fleet, including those that are still being built, shall attack this guardpost", "gravity battlestation, hold position here to slow down ships coming from the wormhole", or "shield frigates, protect the mobile factory" would massively reduce micromanagement and make certain micro-intensive units shine.
Of course, none of this is easy to implement, and certain ideas might even prove too taxing performance-wise, so take this suggestion with a grain of salt.

Unupgraded units should hopefully become less of a problem as fleet levels are implemented.


About economy management, a screen showing the metal drain per fleet (and letting you select fleets from it) is probably good enough: sort by metal drain, look at the top entries, choose which one to pull out of a fight or pause constructions for, done.
This isn't 100% foolproof because of things like capturables and zenith trader wares, so perhaps requiring capturables to be reclaimed manually could be a good idea, but it should be sufficient for the average metal shortage and shouldn't add too much complexity UI-wise.


More generally, I would say that right now the big issue is how it's harder and less fun to make a weird unit work in AIW2 than it was in AIWC.

Let's take Autocannons and Space Planes, in Classic I could easily build then into two "fleets" (control groups) and take them to daring raids against guardposts and even superfortresses, exploiting their peculiar abilities (armor damage and radar dampening) to overwhelm seemingly impossible odds; In AI War 2 their quirks aren't quite as interesting (nor as exploitable) and to stockpile those units I need to keep their whole fleet in a safe enough area, removing a lot of other good ships from the offensive.

Another clear example are Autobombs, in Classic I could build tens of docks, hundreds of engineers right next to them, and quickly send hundreds of Autobombs into the threat force massed up on the other side of a wormhole; In AI War 2 I would have to wait for a fleet to build them, which to be sped up requires me to fiddle with command centers to have the max amount of factories on adjacent planets, and then manually send them on the offensive, one cap at a time.
Also, in Classic I could decide to invest into higher mark metal extractors and engineers to improve my autobomb spam capabilities, and that would also allow me to support other offensives at the same time, and have dedicated MK3 engineer fleets for daring repairs in enemy territory or fast response to attacks: that way even a simple "lol, let's use autobombs" game would naturally open up viable alternatives, in a similar way to techs upgrading multiple units in AIW2.

Of course, Classic had its fair share of problems, and for every fantastic quirk on your units there was one rage-inducing immunity on the enemy units, so I'm not suggesting to copy it mechanically: it's more about recreating that problem solving feel it had, that way in which you wondered how to leverage all the tools at your disposal no matter how unusual they seemed.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: BadgerBadger on June 07, 2019, 11:53:33 AM
Orange, I worry that some of what you are describing is in the realm of the "too much fine grain control" stuff that Chris is worrying about.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: AnnoyingOrange on June 07, 2019, 01:02:35 PM
On one hand, I did say to take my ideas with a grain of salt and not as a wishlist.
On the other, the current amount of control the player has leads to every single sabot guardpost requiring micromanagement of your units, compared to the pre-fleets version allowing a single rally order to get the job done efficiently: perhaps giving some control back to the player is needed to allow streamlining in other areas.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: x4000 on June 07, 2019, 04:06:14 PM
A few thoughts:

1. I think that, when it comes to ships that really only work well under specific circumstances for the player, those should be out of the main fleet design lists.  Knowing what those are, and making hand-designed fleet types that show up more rarely would be good.  Aka having Raiders always be in an all-raiders fleet when showing up for the player, or space planes or whatever being in a combo with something else that's good just with them, should help.

2. Disallowing super-duper-expensive things from being part of the regular fleets might have some merit, I dunno.  The point is, there's a lot of flexibility with the fleet design stuff versus it just being "eh, it could be any old combo at all," specifically to deal with some of the more edge-casey units.  But if we're running into more cases of these than I expected so far, then that just means we need to adjust some xml.

3. I do think that potentially lowering the difference between the mark levels would be good, so that things are not quite such huge jumps.  That would then mean we need to definitely make the AI homeworlds mark 7, and their planets in general scale higher on average throughout the galaxy.  It's way easier to get higher-level units now, compared to the first game (and with the XP stuff that will only be even more true), so the AI going up in power to compensate makes sense to me, while downing the difference between all the power levels in general.  If mark 1 stuff is truly useless, well, that's just not good no matter how you cut it.

4. I'm still inclined to make lower-mark stuff cheaper based on... something.  The average mark of ship that CAN be built in that fleet?  Or something more global? KISS principle, and just make it that as soon as you unlock your first mark 3 unit, all mark 1 units get cheaper, and after the first mark 4 unit all mark 1s and 2s get cheaper?  And so on?

5. Overall, when it comes to having some sort of sub-group that makes sense and is good for doing X task, that's really the entire role fleets are supposed to fill in the first place.  Sometimes it's supposed to be "here's an interesting combo," or just a smallish battle group of stuff in general.  But... maybe for the problematic units like autobombs, those need to only ever exist in fleets with just themselves, so they can be managed a certain way. Maybe... in some ways, maybe all of the strikecraft-based fleets need to be that way?  That would be kinda gross, though.

Quote
About economy management, a screen showing the metal drain per fleet (and letting you select fleets from it) is probably good enough: sort by metal drain, look at the top entries, choose which one to pull out of a fight or pause constructions for, done.

I like this idea, and anything that is being captured or purchased from the zenith trader is a part of a fleet of the planet they are on, so this would still capture that use case.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: RocketAssistedPuffin on June 07, 2019, 05:20:26 PM
1. I think that, when it comes to ships that really only work well under specific circumstances for the player, those should be out of the main fleet design lists.  Knowing what those are, and making hand-designed fleet types that show up more rarely would be good.  Aka having Raiders always be in an all-raiders fleet when showing up for the player, or space planes or whatever being in a combo with something else that's good just with them, should help.

That'd help with Fleets that have counter-active unit types (like Raiders and Tesla Corvettes), but doesn't completely fix it, since there's still some randomly designed fleets. It was a nice idea in hindsight but...the tactics side of the game just hasn't stepped up to that well - can't really find uses for many units beyond "toss it blindly at something". Raiders being tied to a Fleet is still a problem for them. If the centerpiece is slower, then they have to wait for it and thus defeats the point of Raiders, or if it's faster (somehow), then the Raiders speed is also pointless since they'd rather hang about inside the centerpiece instead. Either way, they're affected badly by being part of a Fleet.

EDIT: Thought I should say that I still kinda like the idea of the random Fleets (I did support it internally). I think it was just let down by some other parts, not that it itself is bad. Some units...might just need changes to really fit this whole thing. Raiders are just a particularly good example.

2. Disallowing super-duper-expensive things from being part of the regular fleets might have some merit, I dunno.  The point is, there's a lot of flexibility with the fleet design stuff versus it just being "eh, it could be any old combo at all," specifically to deal with some of the more edge-casey units.  But if we're running into more cases of these than I expected so far, then that just means we need to adjust some xml.

This feels a bit worrying. If we start making it all specific combinations then it's kinda...not really much to do in terms of trying to find uses for a fleet. It'd be kinda apparent? But on the other hand you can't combine things like Raiders and Eyebots in any useful way in the current system either.

3. I do think that potentially lowering the difference between the mark levels would be good, so that things are not quite such huge jumps.  That would then mean we need to definitely make the AI homeworlds mark 7, and their planets in general scale higher on average throughout the galaxy.  It's way easier to get higher-level units now, compared to the first game (and with the XP stuff that will only be even more true), so the AI going up in power to compensate makes sense to me, while downing the difference between all the power levels in general.  If mark 1 stuff is truly useless, well, that's just not good no matter how you cut it.

What were the reasons for Mark 6 and 7 originally? I don't recall seeing it, and it's always felt weird being able to out-tech an AI Homeworld en masse. Also weird that, because we don't have higher Marks locked behind Fabricators and Advanced Factories/Constructors, and no Design hacks left, that...there are almost no hack options that actually work on the AI. I think there's Science Extraction and...Super Terminal? The Dyson Sphere alone has twice that.

I wonder if not keeping the outdated ships when you increase their Mark level has caused this? With that you always had a sizable bunch, and they could still get something done. Or at least, give the impression of doing something beyond being the first to die.

5. Overall, when it comes to having some sort of sub-group that makes sense and is good for doing X task, that's really the entire role fleets are supposed to fill in the first place.  Sometimes it's supposed to be "here's an interesting combo," or just a smallish battle group of stuff in general.  But... maybe for the problematic units like autobombs, those need to only ever exist in fleets with just themselves, so they can be managed a certain way. Maybe... in some ways, maybe all of the strikecraft-based fleets need to be that way?  That would be kinda gross, though.

Yea, as said above that kinda starts to go towards being fairly straight forward and kinda boring. At that point Fleets are kinda just...bigger power spikes in less planets compared to Classic, and unit organisation. If there are Fleets entirely of AutoBombs or Raiders or whatever, then it's...I dunno, like capturing an ARS, but the unit has a centerpiece it has to follow.

One thing of Fleets that's came up is...regarding things like the Fleet Concentration AIP, or whatever "carrot" approach is possibly taken, Fleets make it a bit clumsy to do unit combinations unless said units are in the same Fleet. Spiders and Agravic Pods can kinda work together (Spiders slow down poor engine stuff, Agravics shoot up whatever is resistant), but if they're in different Fleets then you have to bring both together to do that. If there's incentive to spread them out then...you're really at the mercy of RNG if such interactions are possible. This'd extend to AI planets too. If this thing has a ton of Concussion units, then ideally I'd want to keep units weak away, and bring in units resistant, but if I happen to have only one type in each Fleet, then I have to bring them together...thus, I have fleetballed. Just, more a ball made of several smaller balls.

EDIT: Though maybe the above point isn't that bad. Sometimes Fleets just need to come together for a while, and that's alright.

"Clumsy" is a word I'd use to describe Fleets after using them for a while. The organisation is nice, being able to move groups of units that reinforce cleanly about is wonderful, and something I would've appreciated in Classic, but the things like being tied to the Centerpiece causes a lot of issues, like Raiders and AutoBombs being ruined by it, and the above mentioned awkward scenarios of trying to do combinations. Not impossible to change.

Regarding units...I might make a separate topic on that. There was a complaint in Discord while we were doing the original Fleet update about everything dying too fast, it all seems like generic hammer compared to scalpel, etc. Might be something in that...hard to notice things like MLRS doing more damage to stragglers if everything dies instantly anyway.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: MaxAstro on June 07, 2019, 11:53:09 PM
That'd help with Fleets that have counter-active unit types (like Raiders and Tesla Corvettes), but doesn't completely fix it, since there's still some randomly designed fleets. It was a nice idea in hindsight but...the tactics side of the game just hasn't stepped up to that well - can't really find uses for many units beyond "toss it blindly at something". Raiders being tied to a Fleet is still a problem for them. If the centerpiece is slower, then they have to wait for it and thus defeats the point of Raiders, or if it's faster (somehow), then the Raiders speed is also pointless since they'd rather hang about inside the centerpiece instead. Either way, they're affected badly by being part of a Fleet.
Other than verisimilitude, is there any reason the centerpiece of a Raider fleet couldn't just be a slightly more buff Raider?  I feel like some fleets - Space Planes also come to mind - really benefit from being homogenous, but I don't see any reason the current design doesn't allow that.

Just have a unit named "Raid Leader" or "Space Plane Commander" that has a slightly distinct icon and "centerpiece powers" but otherwise functions exactly like the rest of the ships the fleet is made of.
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: ptarth on June 08, 2019, 02:12:58 AM
I like the conceptual idea behind the re-evaluation of the economy, but like many others, I'm not sure the initial proposal was the right way to go. I do think that the core ideas were reasonable though. Here are my tweaks.


Noteable Results
Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: RocketAssistedPuffin on June 08, 2019, 06:25:52 AM
Other than verisimilitude, is there any reason the centerpiece of a Raider fleet couldn't just be a slightly more buff Raider?  I feel like some fleets - Space Planes also come to mind - really benefit from being homogenous, but I don't see any reason the current design doesn't allow that.

Just have a unit named "Raid Leader" or "Space Plane Commander" that has a slightly distinct icon and "centerpiece powers" but otherwise functions exactly like the rest of the ships the fleet is made of.

No reason that I can think of, it'd be trivial to do. There are things like RaidOffensiveMix and CloakedMix fleet designs, but they occur on any Ark or Golem currently, with low chances. RaidOffensiveMix seems to have only Raiders and Mirages (a Space Plane variant) in it, Strikecraft wise.

Title: Re: Chris's latest design thoughts: Metal Has Got To GO (Edit: nevermind)
Post by: AnnoyingOrange on June 08, 2019, 08:59:57 AM
1. I think that, when it comes to ships that really only work well under specific circumstances for the player, those should be out of the main fleet design lists.  Knowing what those are, and making hand-designed fleet types that show up more rarely would be good.  Aka having Raiders always be in an all-raiders fleet when showing up for the player, or space planes or whatever being in a combo with something else that's good just with them, should help.

Trying out unit combos to see what works and how is part of the appeal of strategy games, streamlining it is very VERY risky.

5. Overall, when it comes to having some sort of sub-group that makes sense and is good for doing X task, that's really the entire role fleets are supposed to fill in the first place.  Sometimes it's supposed to be "here's an interesting combo," or just a smallish battle group of stuff in general.  But... maybe for the problematic units like autobombs, those need to only ever exist in fleets with just themselves, so they can be managed a certain way. Maybe... in some ways, maybe all of the strikecraft-based fleets need to be that way?  That would be kinda gross, though.

Fleets have a lot more potential than this, they don't have to be a glorified control group and nothing more.
The current choice between micromanaging or sending a bunch of ships against targets they're weak to is squarely in the "optimize the fun away" camp already.