Planetary Annihilation's Economy System

Here's the issue- as your economy grows in scale you are going to need more build power to spend it. More income and more factories and engineers. And, if your economy is global, then all your resources can be drawn anywhere, regardless of map size. This means that you can, with perfect efficiency, stack up enough build power to spend your entire economy in a single location. That increased quantity of build power isn't like your extractors- they can in fact be stacked up in a single place.

For relatively small scale maps like in TA, this doesn't matter. You have relatively few constructors, and the fact that your resources are available everywhere doesn't matter as much because there isn't as much area, period. But if you want to make a huge map in TA, you will only build one base. With the important exception of mexes, you have no reason to build in more than one location.


Consolidated Serial Production

Acquiring more resources gives you the ability to build projects serially at a faster rate. And serial production is more efficient than parallel production. Choosing between 1 unit in 1 second for each of the next 100 seconds and 100 units finishing simultaneously in 100 seconds (each building at 1/100th the rate) is not a choice. Unless you have a military, economic, or logistical need for parallel production, production in serial is simply better. And if players will always prefer serial production, then they will only ever build one real base, with a bigger economy simply serving to make that one base bigger/denser, rather than expand the player across the map.

The problem becomes more significant when multiple planets enter into the equation. Suppose extractors and power generators create resources which can be spent anywhere. It makes the most sense to consolidate energy generation, large-scale production, and other expensive infrastructure in a single base on a single planet. Outlying "colonies" might need minimal expenditure for local defenses, but players should consolidate in a single heavily industrialized and well-defended homeworld.


Economic Strategy

Players are practically forced to do what is most economically efficient. And certain types of assets or strategies go hand in hand with different economic styles. In TA on huge maps, a single massively consolidated industrial base used to produce hyper-mobile, infinitely stackable, versatile Hawk or Vamp stealth fighters was extremely dominant. Firstly because the economic system encouraged serial consolidated production, which made the distances between targets of significance huge. And that distance then created a huge advantage to the most mobile units in the game.

With more positions of interest, and with shorter distances between them, other types of less-mobile units become usable. For example, if both players had many bases across a planet's surface, then suddenly it makes sense to use ground units on all your front-line bases. The players can use a variety of types of units and weapons to attack and defend their bases and attack nearby enemy bases, and allow them to build in the territory vacated by destroyed bases.

Players should be filling territory, with conflict wherever their territory meets an enemy's territory. Players should make small attacks with local consequences, such as claiming a chunk of new land that used to be the enemy's, not just big moves direct at a single massively defended enemy "main" base.


Encouraging Distribution

I think the economy and logistics are the best methods which might be used by PA to encourage expansion and distribution through space on a planet, and across planets. These aren't the only possible methods- a carefully designed military system could make distribution across the map strategically advantageous. But as an RTS it is very important that players need to control large areas of land, build across large areas of land, and distribute units across large areas of land. Players should want to build multiple, even many bases on a single planet, not just a single one, and part of the player's strategy should be which bases should build what structures, units, movements, local tactics, and so on.

It is very important that PA be a strong RTS game even on a single planet. The multiple-planet idea is awesome and will be a part of PA's core gameplay, but PA should not lean on the fact that there are multiple planets at the expense of its RTS fundamentals. From a game design perspective, the PA devs should be trying to nail the desired gameplay on a single big planet before the interplanetary level even enters into the picture. This means giving players strong incentives to spread across a single large planet map. And once multiple planets do enter into the equation, strong incentives to build on many planets and not just heavily develop one which can be efficiently defended.


On Planet Killers

And because I just know the low-rent responses to this post are going to claim that planet-killer weapons will offensively encourage spreading, they are wrong. The only method that will work is making distribution more efficient for the player choosing to distribute. If centralization is still more efficient, but made risky due to centralization of loss from planet-killers, then that simply places a much higher emphasis on planet-killers on ending the game in a single shot, and detracts from flexible and reactive gameplay.

Players will still build to maximize efficiency, but if they get KEW'd, the game ends immediately. Deliberately building sub-optimally is asking to lose against someone who plays properly. And two players building optimally, knowing a KEW ends the game in a single shot, is terrible gameplay. Worst case scenario the game is a pacifist rush to a KEW, and the game ends.

You raise a lot of valid concerns, but i find that even as is you are encouraged to spread out across the map. defense against KEWs needs to be built so that you don't get your entire planet destroyed from the side your base isn't on, then you don't just lose that base you lose the entire planets worth of metal and energy production. Then you need to defend that base from someone coming to kill that defense before their KEW hits. So you will dot your landscape in firebases.

Oh but what if they land stealthily on your planet and use artillery to hit that KEW defense? you are going to need a standing army of some sort that has a response time to hit that artillery once it starts firing.

And that's just with that one target. they could just as easily run through half your metal production if you haven't spread out properly?

The addition of multiple planets actually already increases the need to spread out because you need to concern yourself with the possibility of an opposing force landing anywhere you control.

I have an idea that eliminates the need for a "priority system" entirely, and is intuitive: Simply distribute power in discrete units, on a per-receiver basis. Allow me to explain:

You have 2 devices, 1 that requires 2 power and one that requires 10. You lose some power production and you now have 6 power production. Instead of distributing power proportionally to the requirement--1 to the 2 device and 5 to the 10 device--and have both devices suddenly shut-off or operate below-maximum, why not distribute the power according to device? That is, you have 6 power and 2 devices, 3 power per device, every device should get 3 power, then the devices that need less will simply free-up power for higher-drain devices. In the example, the 2 power devices would receive 2, and the 10 power devices would receive 4.

If you used a system like this, higher-drain devices would naturally shut off first in low-power conditions, and the lowest-drain devices would remain at 100% capacity despite the higher power devices.

Another example: You have 20 metal extractors, 5 factories, 3 defense guns and 1 radar. Each requires power at a rate of 2 for extractors, 20 for factories, 10 for defense guns, and 50 for radar. Thus, your entire power requirement to run everything at 100% (or better) is 220 power. You have exactly that much. But if you are attacked and half your power is destroyed, leaving you with 110, how should that be distributed? In supcom, all devices would receive half their requested amount, and operate and half efficiency (or shut off entirely). I propose this math: Give each devices it's share, weighting all devices equally. With this calculation in this example, you need 220 to run everything 100%, from 220-190, only the radar loses power, from 190-130, only the factories and radar lose power, from 130-58, only factories, guns, and radar lose power. Less than 58 power, and everything starts to be affected.

Is that clear to everyone?

Also, building upon the above, if you wanted to allow players to prioritize specific units or structures in the game to always receive power first, you could. That is, in the above example, if the player designated two of his factories and his radar to be priority receivers of power, his total power production would have to drop below their combined maximum (20 + 20 + 50 = 90) before any of them received less than 100% power... everything else would just shut off more quickly.

This, combined with the above post and the economy system you explained in the original post, would add much strategic and tactical decision making to economy management that is, simultaneously, intuitive and not needlessly complex.

KEWs.
Nukes.
Map control.
Energy nodes.
Artillery hiding spots.
High ground.
Choke points.
etc. etc.

Mex control is the #1 reason you would need to sprawl out and build multiple bases. It wouldn't be optional. Having domain over your resources is necessary to win.

This is the best reason why energy generation should not be a sector wide resource. It has been argued many times to great detail elsewhere, so I won't repeat it again here. Suffice to say, that forcing some kind of energy sprawl is just as important as having the standard metal sprawl. Developing new energy economies is restrictive enough to keep abusive expansion and strange strats in check.

IMO, transit is the best option. Building on the front should be easy, and moving units should be fairly difficult (but definitely possible). Therefore, the best option is to build a front line base, where units can be constructed directly on site and transit can work at full capacity.

Limiting radar against ground is good, if its can't do ground at all that's even better. More difficult vision means you gotta spread out more to see all the angles.

Base defense options should be adequate for preventing base duels, because the bigger base is always going to win a direct slugging match. That's one major reason I hate stationary TMLs, because they're base weapons designed to kill bases.

Base breakers should deal big AoEs (Like KEWs), terrible accuracy (like big berthas) or feature slow movement (like Supcom satellites) to favor acreage over density.

If the stuff in the backer's lounge is any hint, it looks like big bases will be unavoidable. All the factories and generators look like they're going to take up a lot of space, no matter how you use them. I like it.

Everyone is going to have a different idea of what deserves to be prioritized and what doesn't. The only fair option for everyone is to divvy it up evenly.

There are a handful of cases where the numbers don't lie- it makes perfect sense to have X running at 100%, to the exclusion of everything else. Similarly, there are things that should be running at 0%, unless there's excess resources to burn. Having these options available as global settings would be excellent. For everything else, it's simple enough to have players manage it however they see fit. Besides, a shortage simply means you need more energy.

1) Prioritizing energy is important?
Imagine your economy is energy starved but you do need that large gun to kill your enemies.
It would be very practical to be able to prioritize it above other stuff.


2) I wonder what the influence of assisting engineers would be on factories.
In FA it is very common to have say 200 t1 engineers assist an air factory.

An air factory uses relatively more power than other factories.
Imagine it would need 20 power to use one 1 mass

An engineer uses 10 power to operate to contribute 1 mass.
What if I would let a lot of engineers assist the air factory?
Would it change the ratio in which mass/energy is used to create units?
Or what should/could be done to avoid that?

I'll get to the other parts later...

It's not always better to do just serialized building. There's going to be some block of time spent in overhead for each unit build. A factory spin up time and a unit drive-off time. By going solely serial you will have to pay this time cost every time. Going somewhat parallel you can get better efficiency out of the entire system simply because some factory will be doing work during this time overhead.

On the second half, putting all your resources in one place may sound like a good idea, but if I'm scouting properly I know exactly where I'm going to hit your first.

If I may ask, how long has this idea been in the works? Just curious.

But I also know where you are going to hit- because there is nowhere else to hit. This is the entire meat of the issue.

If all your infrastructure is consolidated in one location, I have only one strategically significant location to attack. And if all my infrastructure is consolidated, you also only have one location to attack. If either of our bases is lost, the game is over, justifying almost unlimited troop commitment and/or defenses at that location, and requiring a similar commitment in order for an attack to succeed.

I would like to see both players with many bases on a single planet. Spread over enough space, it becomes mathematically impossible to construct enough defenses to ensure every location cannot be defeated by a suitable commitment of enemy forces. The game becomes a question of how many forces you want defending each location, and how many forces attacking each enemy location. You don't grab every unit you have and send them at just one enemy base- that's a huge overcommitment when a smaller force could do the job, when you could attack in multiple locations or beef up your defenses elsewhere. Scouting enemy troop movements lets you skimp on defensive deployments and more efficiently calibrate your attack forces to efficiently capture enemy bases.

My point is that even if there is an efficiency ceiling for production efficiency from one factory due to roll-off time, etc., it still makes strategic sense to have a single large main base with a universal economy, even if there will be multiple factories. Consolidated energy can be sent anywhere, and buildpower can use resources from anywhere, with no cost from distance from infrastructure, even though distance from the enemy is the best defense. The defensive advantages categorically outweigh the zero economic efficiency cost of consolidation.

Even if you wanted to build elsewhere later, such as building a forward base while invading an enemy world, having a single large industrial center to quickly build the engineers and transports, and then provide essentially unlimited energy from a safe location is still ideal.

"So in war, the way is to avoid what is strong and to strike at what is weak." as it stands in Sun Tzu's Art of War
The point is: You are both doing it wrong. Why would you run your army into the giant well defended base, if there is a plethora of vulnerable economy everywhere else? You can't defend everything with one giant base. Ergo: Don't make the error of building one.

Edit: Actually, Scathis can get out of this by saying that he meant to hit everything but the base first. Sneaky use of words there

Yeah, I'm pretty sure that if you build one huge industrial complex, I'm just going to rove around and remove all your extractors, building small bases and defenses at each cluster. Then when your huge army drives out to strike one base, all the other bases will collect their troops and strike your now mostly undefended base.

At worst, I'll lose a small base, but if you move out with everything, you'll lose your only base. If you move out with a fraction, a small base should be able to hold, meaning I have more territory but less vulnerability.

As long as the map is big enough and even the highly mobile units aren't fast enough, players will automatically spread out.

Of course, if you have multiple bases, your opponent has one, and he attacks one of your many bases with a full army and you try to stop him from doing that... you're doing it wrong. Strike the most vulnerable and high-reward target, his now undefended primary base.

Especially if it contains some volatile structures, you don't even need to win the game right there, you just need to bring down one or two volatile structures and have them blow a few holes in the base. With everything packed so closely together, a few well placed shots can bring someone's entire production line down, and you'll only be down a single base.

The only reason playing scattered like that wouldn't work is if travel times are negligible, which basically means the maps are too small or some of the units are too fast for their efficiency (or too efficient for their speed). But that's mostly because the only thing that seperates two locations is travel time between them; if there is no travel time, then the whole map automatically becomes a single location.

I don't get how serial production is more "Efficient" than parallel, are you saying that players should only ever use one base to build troops?

Also you seem to be saying that players only need to build defenses on other planets, and will just make troops in their main stronghold and send them to reinforce where they're needed? If so consider that even in SC, having a factory or two near an enemy base is enough to give you a massive advantage simply because your troops don't have to cross them map. When the troops don't have to cross a solar system this advantage would be compounded =P

A few pages back there were a series of arguments relating to dealing with issues that arise with huge scale. Please have some common sense and stop assuming PA will be bigger scale than FA or TA.

Would you rather receive $100 a day every day for 100 days (totaling $10000) or wait 100 days to get $10000? Obvious answer right?

So if you want your $10000 incrementally, eg - serial production, you need to concentrate all your build power on one factory or on one construction at a time. Since all your buildpower will end up in this one base, this one base is the only base you will have since you need buildpower to build bases right?

Moreover, the post you quoted was heavily suggesting that we find ways to stop players from only using one base to build troops. Where did you get the idea that he was suggesting we should only ever use one base to build troops?

If there's a gigantic, impenetrable, singular base, you KEW and orbital laser it to oblivion. Bam. The planet is now yours, and not much harm was done to the surface.

The problem with singular bases has more to do with a player's ability to manage their assets and rally points, rather than any deliberate thing. Starcraft 2 shows that players will have multiple base locations if the management is easy (warp gate), if they run out of room (Terran), of if there's no other choice (hatchery).

This is one of the game's core selling points. All the work going into the engine is so that games CAN be retardedly huge. How many other RTS titles let 40 players in a single match?

You are assuming two things:
1. PA will not be bigger scale than FA or TA (that's certainly false)
2. That time "costs" nothing. If it takes 3 seconds for a unit to exit a factory, than more factories = less total time to construct large amounts of units, as there is less delay time.

Having 1 factory is also a severe case of "all eggs in one basket", which is pretty dangerous when KEW's are involved, not to mention the inevitable economic disadvantage of not expanding.

Seriously? You're saying that PA isn't going to be a much larger scale than any other game you've encountered? Let alone TA?

Assuming that the world we saw on the first video is an average-sized planet... it still took some time to move from one end of the planet to the other. It still had two bases on it plus a lot of extra room. Games will most likely include dozens of worlds with similar characteristics to this world, plus additional base opportunities such as asteroids, moons, and gas giants.


And the current discussion that is going on... I understand the perspective of having a consolidated base with many outlying resource bases (which is what I was trying to get at earlier). Using SC as an example again... if any player is able to pull off having several bases, losing a couple of them doesn't even phase that player. Zerg for example can expand so readily then just swarm you from one location very effectively. I can see PA taking a similar direction.

I'm really looking forward to trying this economy system.
It seems familiar for us TA/SC fans, yet better for newcomers and for everyone, as it will be easier to read how much something will cost to build and to balance the economy.

One question though, any thoughts on how this will work in the UI? :?:
If Metal is a direct cost for building stuff and Energy is a indirect cost (for running the factory/eng.) and the build time is just a function of the builders efficiency and the economy. Will there still be indicators in the building menu with preliminary information about much time and energy that will be spent?

Something like this in the building menu should be very clear for players:

• Build cost: 640 metal.
  Build time: 5m20s. (2 metal/s).
  Energy demand: 32 GW (0.1 GW/s).

"Demand" is a good word for the Energy, as it differentiates it from an actual direct cost.

In other words, the economy system is "changed" on a lower level, but will the players on UI-level still be able to easily see how much time and energy that is demanded for building stuff?

It's very easy to calculate build time under this system. The trickier issue is making sure you take into account all of the assisters to get the correct time.

An issue I forsee with making builders consume 100% energy even when they're starved for metal is it's going to become a thing in competitive play to manually stop every builder you don't need just to save on all that wasted energy. Not only do I forsee leaving all your builders on being a greater newbie trap than anything you're attempting to solve, I imagine this will be a really frustrating part of the game that no one likes, yet has to be done.

Also, I resent the notion that simple resource priority toggles for units reduces fun and makes the game about min-maxing. Its job in games like Zero-K is to streamline actions players want to communicate to the game. Some constructors may operate on the front line, and it may be urgent for that defence tower to get all the resources to be built as fast as possible. Other constructors set to build nothing but energy, or extra T1 factories I would want to set on the bakburner, expending all my excess. Remembering to take builders off of projects half a planet away then reassign them again, when those builders wouldn't be useful anywhere else anyway, is not fun. And Total Annihilation already had a good visualisation of how much resources nanolathes were using; by how much they were stuttering every tick.

And just another thing. I'm still curious to know if metal makers(assuming Planetary Annihilation will have metal makers) will be automated or not. I only ever need them to convert energy at the same rate I'm excessing it.

If it sucks we'll change it. Builders only use energy when building something btw so turning them off is just not using them... Or put them on patrol and they will reclaim when you are low on resources.

Why do you resent it? The idea that we should automate as much stuff as possible simply isn't true. Where the line is drawn is important and it's not always going to be on the side of automate as much as possible Using your argument we should just tell the AI to build us the most efficient base. Where do you draw the line on automation?

We will have similar. This is my whole point, I want as much information in the game world as possible instead of hidden away in pop up menus.

First off I'm not sure we'll have metal makers, it's up in the air at the moment.

They may or may not be automated. I think you are losing sight of the fact that no decision is final until we've all (including you) had a chance to playtest this stuff. Please don't go off getting mad at decisions you disagree with until they are final.