Modular Units

Early game you pretty much had to make the army sensibly (some antiground capability, some anti air capability) and your tech was limited.

There was some choice (heavy hitting Tiger Gs or anti-aircraft Tiger-Rs), but you were kind of rail roaded into being sensible rather than revolutionary.

I found in singleplayer you didn't really get much of a chance to be very revolutionary - I don't know about you but at that point I was REALLY strapped for cash, so I made budget cuts. Granted late game there was a lot of choice..

I never had a lot of success in multiplayer though - picked up the game for $10 a few years ago, never really gave multiplayer much thought because the skirmish mode AI kicked my ***.

Perhaps we should try to arrange a game some time

That would be cool. It's now on Steam.

No way. Really? And multiplayer works via steam?

Yeah.

Er... perhaps? It asks you to sign in Earth net... which is defunct isn't it? But it seems to connect to server. . .

Cwarner do you have any idea how to EarthC? It didn't come on my copy.

Code:

Short version for TL;DR people:

Yes. Modular System.
1. Pre-build system like map-making instead of ingame. Units load with game.
2. Kerbal Space / Spore Creature creator type modular building instead. (More freedom, less fast.... can do since prebuilding.)
3. Get other player's designs by capturing unit and winning match with one in tact.

This makes metagame more interesting, more adaptation required in matches, more reasons for personal pride in team, and more fun in general.

I'm going to jump in on this debate. Modular units would be cool. But the complaint about micromanagment is also valid. Also, there's hypothetically greater server resources....

But you know what. Screw it. I'm getting behind this idea. It's not THAT much more work (It's not a small amount, but its not insurmountable either). Units already have moving parts. Just make every moving part switchable. Then units become collections of modules that move by set rules instead of having to design each animation individually.

Secondly, and here's where a "decently good idea" I have comes in...

Instead of designing new units on the fly... design them ahead of time. Like we have user-made maps that load up when the game begins, have part of the meta-game be designing your team ahead of time that then load up when the game begins. You have a standard stock based of regularly available units, but you can also construct buildings (or existing buildings have a few extra slots) for experimental units. Your experimental units are ones you design yourself, and doing so will add a LOT of diversity to gameplay. That's one thing very few RTSs have brought to the table, and it will help keep PA from suffering starcraft I syndrome. (Yea, yea, I know starcraft is popular, but that niche is already filled. And altough I love RTS's, I've always hated starcraft. It's always "Build this set number of units, set those units to do this ratio of tasks. Set this next group to do these set of tasks." It became very formulaic. However, when you study the history of war, the winning side is rarely the one that perfected the numbers game, its the one that had the best innovations, and that strikes me as FUN.

Sure, you could play completely with stock units, but to me, the greatest fun in any RTS has always been the moment when I build (or encounter) a completely new and untested unit. The thrill of the unknown. The fast adaptation. The curiosity (wtf is that?!?), followed by the rage (I hate that enemy unit soooo much!), the lust (oh, but if i just had one of my very own...) and so forth.

It could even add amusing metagame to PA. When you see an opposing player's experimental units, you don't get to know its stats in anyway, just get a "wtf" response before it rips you a new one or dies horribly. If you want to actually see its stats and parts in their entirety and try to build one yourself (instead of just guessing from some of the modules you see) then you have to do 2 things:
1. Capture one.
2. Win the mission.

Doing it this way makes a much larger metagame for one (you WANT to hunt down good players, risking your win-loss ratio, to acquire their experimental builds, people are less likely to ONLY play with one clan for the rest of their life or limit themselves to just pubs or just pros)

Secondly, the pre-build allows a much more complex modular system. It becomes less deciding "do I want wheels or tracks on this buggy" and more "what creative thing can I build?" (A little less the fast overly-simplistic modular system mentioned earlier, and more the modular systems of Kerbal Space Program or Spore's Creature Creator.)

Thirdly, this will, once implemented, make it easier for the creators to expand the game. Just add in a single new module will result in thousands of new units to be encounted and designed by players.

Example... assume no vtols in game. Instead of adding a single vtol vehicle, a vtol engine gets added. Sure it makes the unit much pricier, but people abuse the tech in various ways. Some people attach it to builders to get them making hidden bases easier, some add them to tanks to make jump tanks that get past defenses quickly, some add vtol engines in massive numbers to a single valuable building to make it mobile so it can dodge an otherwise inevitable destruction, others still add it to tiny shock troops and swarm with dodgey little units while others just use it to improve their resource management a bit. And also, that one weird guy attaches vtol engines to walls, just because he can, and ends up with a crazy weird area control device.

So one addition opens up a world of options instead of just one option, making for an RTS that literally has infinite possibility and replay value.

Side note: I could also see later on new building parts being released ala TF2 style... new parts (once released) can either be acquired in game via capture, random drop, or the cheaty-face "buy in market" method. There could be cost-free (in-game) cosmetic parts too that cost an arm and a leg (out of game) to fund further development (with a good portion going to original creator of the cosmetic part) which will allow REALLY customized units, so you KNOW who's attacking you. (Cosmetic parts may not come with capture, getting replaced by generic spheres or something.) I would waste sooo much time in the workshop for this game.

Making a new thread so can get a vote going on this version of the idea. (Specifically for pre-built modular units.)

I can confirm this coming as a/many mods...

Although doing it as mods brings the question of "can team matches dynamically download any mods the host is using"?

They say they will not only support a "modloader", but also a mod database. I believe the name they agreed upon is something like the "Uber Foundry" or something different.

Trophy, considering that we're on the same mod team, I think it'd be better if we compromised/consolidated on the way modular units work instead of making separate mods.

However, I'm not sure about how they're actually going to *benefit* gameplay. I think one thing that modular units could be increased in depth is adding an actual coding interface to them. So you could code an AI module beforehand - here's a really simple example:

Code:

class AIModuleBravo(AIMod):
    def onDamage(self, amount, attacker, attacktool):
        fleedir =  attacker.pos - self.pos
        self.moveorder(-fleedir, local=True)

You could code various AI Modules beforehand, and then drag-and-drop them onto various units to allow them to do awesome things not normally done in games. I.E., using a mass of engineers and transports as nanolathe gunships, artillery that runs away from targets as it fires, or subs that automatically submerge/surface depending on what is happening around them.

The obvious limitation here is the amount of strain these extra AI scripts put on the server, but a good way to deal with this would be to run the scripts on the client, and have the client tell the server what orders are being given to what units. But this dips into the whole bandwith issue, and gets complicated very quickly.

Wait wait wait wait wait.

YES! I am absolutely in favour of customisable AI scripts on different units

which language is that?

Python

WZ2100 holds a special place in my heart and I'm always on the lookout for games with this kind of customisation. The Scraps kickstarter launched today so I funded that, there's KSP as well, but not as far as I can tell any recent RTS titles with that system.

But, yeah, it's not within the scope of the base PA game.

I agree on consolidating code on how it's done. We could also work together. I have a pretty picky mod concept, I am not sure we can make that mod one and the same, but if you have an idea I can help you make that and vice versa.

We can never have too many mods. I will even forfeit my right to a multiplayer setup if you just let me do a campaign (which I actually have more interest in). We could work out setting up the mod, then making a campaign out of it.

PA is not going to change into a modular unit design model. However, I think that a modular system would work if the designs are compiled into a list that you set out of game. One of the greatest benefits of this model would be true player army customization. You just say the list has to include say 10 T1 units using T1 modules. The unit cost would be determined by the modules that are chosen. You would have to have perhaps 15 T2 units built from T2 modules. The buildings of the game would not need to change too much. The factory for each unit would be determined by the main chassis type of the unit.

I personally would find this style very gratifying. It would be kind of like building your deck in a card game like Magic the Gathering or Legend of the Five Rings or Pokemon.

I really think this would be a blast to play!

One Word for you:
Experimental / Megabot

Choose one

A system like this would be really nice to see, I'm not sure about some of the technical challenges mentioned by some people, but lets consider how it could play out if used properly

As with the Tool-belt vs Omni Tool argument, You cannot simply make all units modular and still expect to see the diversity and strategy employed by the tool-belt side of things. However considering the 'factions' we will see are essentially all the same race and have access to all the same units with a few varying commander models roaming around it makes sense to seek some other way for players to create something unique to their side.

In Supreme Commander the 4th tech tree was laballed experimental and basically had a few options you could choose from, again fitting the toolbelt sort of idea... a unit for a specific job. But the lab didn't feel like any kind of experiment, it always felt like an endgame strategy to work towards, not some ingenious experiment conducted on the battlefield which truly can turn the tides of war.

Consider in Planetary Annihilation an experimental mega bot lab, that lets you first build a chasis in the game world from varying types, be it bi-pedal, tracked, or airborne, then after the construction they can then choose how much firepower to include vs how much armour or speed. A finely balanced system meaning if you choose to have more firepower you will inevitably have less armour and/or speed and vice versa.

This kind of system will allow for the 'didn't see that coming effect', where your enemy has no idea what it is your concocting in the shrouds of your base . It would also be a great way to see players creating their own experimental based on what is needed in the simulation at that time. For example you might plan on building an experimental with 4 lots of guns, but only have the first 2 weapons completed, you could launch the unit as is if the situation has somehow changed for you to not need this extra weapons, Similar to the half baked feature of Supreme Commander 2.

Once the experimental has exited the lab it can undergo no more changes until it returns to the lab to be constructed upon.

Might as well throw in some kind of veteran bonus for long lived experimental s

While the one super unit experimental approach could be managed in game I definitely would prefer to have control over the build of my entire roster of units as opposed to just one super unit.

Additionally, the way the game plays already you would have to choose between the super unit or your nukes or getting off world as an endgame strategy. Yes it is an additional option for ones endgame but it really doesn't sound appealing to me like being able to modify the entire roster.

Now lets look at how this would balance.

I see balance as remarkably easy. All you have to do is factor the weight of the individual modules.

Speed would be a function of weight vs max speed of a unit. The formula could be a straight speed = max speed - weight or it could be more complex. The point is it would easily balance. As far as cost it would simply be the amount of metal needed to build the sum of all the components plus a premium for sophisticated features.

You only need 3 different component types.
Conveyance(Wheel, Tread, Biped, quadra-ped)
Chassis single or dual weapon slots(Light-Buggy, Medium-Small Tank, Heavy-Large Tank)
Weapon
Surface to Surface(Light, Medium, Heavy) with range, damage, fire rate, tracking, AOE modifiers, and projectile types.(Direct fire, Ballistic, Missile(guided), or Rocket(not guided).
Surface to Air.(same as above)
Electronic Countermeasures.(Small, Medium, and Large AOE)
Radar Jamming, Scrambling(shaky blips, ghost blips, no blips, radar spotting, etc.)

How many units are possible with what I have listed here?
4 Conveyance x 3 Chassis, x 6 weapon types, x 3 surface to surface weapons x 4 surface to surface projectiles x 3 surface to air weapons x 4 surface to air projectiles x 3 ECM x 3 ranges x 4 effects = 373,248 possible units and I haven't even factored in possible damage and range modifiers.

That would be AWESOME! In my humble opinion.