Pgen technologies and possible diversifications

First of all, why?
Diversification is not only necessary with units, but also within the economy. There isn't just that one "fit them all" build queue for a balanced economy, but different requirements, depending on your desired playstyle / strategy / pace.

TA serves as a great example with solar panels, wind and tidal generators, geothermal and ultimately fusion reactors, each of them featuring unique mechanics.

But at the same time TA also failed in some way, since the pgens were strictly tiered in terms of build cost, and those which were not were highly depended on the map parameters, not the preferred play style. And ultimately every other generator was superseded by the fusion generator.

How can a pgen fit a certain strategy rather than the enviroment?
There are several factors which can be adjusted to fit a certain pace and play style:

• Planet coverage
  This is exceptionally important for solar generators are their energy output depends on the planets rotation, so it calls for wide spread or even redundant energy plants. But this also effects pgens with a bad build space efficiency, since this is also area which must be protected.
• Payoff time - how long will the generator stay efficient?
  This is something TA never featured, a generator just for the shortsighted ones. Cheap to build at first, but the operation is not free of charge. Burn metal for additional energy the good old 20th century style and hope that everything will be over before the charts turn red. 5, 10 or 20 minutes before it becomes inefficient, everything is possible.
• Technological barrier
  Hide the generator later in the techtree. While it is possible to do this, it is not recommended as the technology would need to make up with enhanced efficiency which invalidates earlier tech.
• Logistic barrier
  How much effort is required to build a working energy plant in the first place? Metal burning aggregates are surely the easiest one, and fusion plants in the orbit of gas giant the most difficult ones. But there is a huge span in between.
• Survivability
  Fusion technology is great. It's clean, it's effective - it's highly volatile. The SupCom Paragon resource generator is a great example (forget about the infinite mass production) of how an otherwise perfect resource generator can turn into a huge threat for your own base.
  But this also applies to other technologies, do you rather go for efficiency or durability?
• Efficiency
  Not every energy generator needs to have the same efficiency in terms of build cost to output, the higher the associated risk / challenge, the higher the rewards.

So what are possible techs?

• Solar panels
  Thats an easy one. Very cost efficient while facing the sun, but requires either redundancy or massive energy storages. Best suited for players who aim for full control of a planets surface, but not very durable. Yet most efficient in terms of build cost.
• Orbital fusion plant (gas giants)
  Ultimate energy source, but locked away behind a huge logistic and technological barrier. Worth it if you plan for aggressive interplanetary expansion.
• "Magic" energy generators
  Just what we have right now. No requirements, but efficiency pretty much sucks. What else would you expect when you tried to generate energy from thin air? And do we really need them at all?
  
• Metal burning turbines
  For the aggressive, early rush play style. Cheap to build, easy to defend, but efficiency will be superseded by every other generator after a few minutes of operation. But also helpful for other play styles if energy is scarce and none of the requirements for other techs are met.
• Fusion plant
  Nope, not the "magic" version. Essentially more an upgraded turbine, with increased payoff time, but also increased build cost and volatile effect. Most effective type of generator short of the orbital version, but not the most efficient one. Nuclear power is expensive!
• Enviroment specific generators
  Wind, tidal or geothermic. Will only work on certain planets and can be used to bias the economy on certain planet types. Efficiency can vary, but is above average when placed in efficient spots and subpar otherwise. Either locked to fixed positions or with terrain bonus and adjacency malus.

Other implications
Many of these technologies are quite situational and there might be huge fluctuations in energy output. The current energy storages are nowhere big enough to cover this.

Energy tends to become abundant in late game, so additional sinks are required. But not only factories are a possible sink, units are also.
With sufficient energy production, it would only make sense to use energy as a generic form of "fuel" for the operation of units, mostly for paying for attacks of high DPS units. It's not only about creating an army, it's also about managing the supplies. This does not only apply to bombers, but also to mobile artillery, and even tanks & heavy assault bots. Encourage the use of efficient, not only effective units.
There is nothing wrong if you had to abort an assault half way through the enemy base because your army burnt the energy reserves too fast. Steamrolling suddenly become far less effective if your steamroller can run out of fuel.

Burning metal on the most effective pgens is essential for creating the diversification. Strategy is all about making coherent decisions, and taking a certain path in the economy is a significant decision. Trading short-run benefits for long-term savings is a meaningful decision and creates significant disadvantages if things don't got according to the plan, either way.

I agree, great ideas there. I don't like the current magic generators that just generate energy from the nothingness, but the idea of generators that burn metal is.. odd, if you ask me. I would prefer the magic generators be scrapped for good and replaced with environment specific ones.

Land/Orbital - solar panels for base energy income, with those nasty night downtimes. Will work on all planets. Plus side for making day/night cycle become a core game aspect to consider, rather than just a nice graphical gimmick.
Wind - for planets with an orbit.
Tidal(?) - Oceanic worlds.
Geothermal - Lava worlds.
Fusion or something - Gas planets.
Ridiculous solar powwwwaaaah(?) - a very expensive and large solar panel orbiting a sun?

Without the magic generators or "burn metal" option energy creates problems that have to be countered with strategy. You are on a moon? Night times are extremely dangerous due to not being able to shoot (if energy/metal cost for unit shooting gets implemented) if you don't have sufficient energy storage. Lava worlds, altough dangerous to occupy, are extremely valuable for that mid/late game high energy income. At the current model I see no reason why I would like to go to a lava planet if I have a choice.

Metal spot hunting is an important gameplay factor right now, and even more so in the future. I see no reason why energy hunting can't be the same.

Yes please.

great ideas, i love the idea of rather than thinking oh thats where he has his power, thinking, oh he's using wind power - which means they have to be spread out and therefor an ideal target for hit and runs.

Burning metal is type of last resort. When you need the energy at all cost and you have failed to expand early on or if your play style doesn't allow for large energy farms.

It's by no means efficient to do it, and it will get you nowhere long term, but it can prolong the game a little longer and give a chance for a comeback.

Or in case of early rush: It makes it possible in the first place, but punishes the attempt with a penalty on the economy long term.

I tried various concepts, but burning metal was the only one which could provide a fair compensation for the lack of expansion since you would still need to expand to some degree to get the metal in the first place so you are forced to have vulnerable ground.

The orbital fusion plant has actually a slightly more complex role, it's not just a regular pgen, but more a way to say "I dominated the game and I'm now entitled to crush you" since it operates outside the regular balancing and allows you to bypass all energy based restrictions which would have limited steamrolling previously.

It will still be very difficult to balance metal burning in a way that players will not just use it early game because it's simpler than trying to prepare for night down cycles. Also I don't like the idea in general because to me it's just the energy operating cost that mexes have, in reverse. It also feels to me like a reverse mass converter, which I don't think this game needs.

I don't understand how this type of generator would punish you in the long term, since you can just delete the generators when you no longer need them. I understand the concept behind it but I don't think it would make a meaningful impact on the game unless it's so powerful that it overshadows other types of gens. If you don't have enough land to build other gens, you don't have enough mexes to burn metal.

I don't think that forcing the player to expand is a bad thing at all, it makes turtling impossible, focusing on the term defenses. If overall damage of units gets reduced, having a raid here and there while enemy main force is knocking at your front door still gives you time to react before all is lost.

It's not a bad thing to use different technologies in early game, but it's actually inefficient if you could as well build solar panels and energy storages early on - you are going to need them later anyway. It's wasted mass because the structure invalidates itself when it is used for too long, mass which could have been used for longterm efficient pgens.

As for the "inverse mass converter", remember why mass converters where an issue in the first place. They allowed for an exponential increase in the economy without sufficient ground control, past the limits of the map.
The inverse system does not suffer from this issue, as you are still tied to a limited number of metal spots. It does not cause exponential escalation, on the contrary actually. It slows down growth.

But you still have reserves
And it's not like you didn't have sufficient space to build other pgens, it's that every single one is highly situational. A metal burner can never fully replace any of the other pgen types since it will always weaken your economy long term.

I know it's difficult to comprehend, but every single pgen type in my proposal is situational only and has unique characteristics, advantages and disadvantages, every single one has a catch except for that one "magic" type. It is impossible to compare them to each other in a static manner.

I can rock with this.

Yes, theorycrafting can't ever be compared to actual gametesting. As I said, I think metal pgen would not work because I think it would be too hard to balance, but I have been wrong before. I would also love to see either of our proposals come into reality, but I fear that is highly unlikely. At least in the December release version.

This would also fit well with the goal of choosing spawn locations based off terrain and resource placements. It also might promote using different strategies base off what might or might not be available.

I like it!

I will say that if each power generator type is suited to a specific environment, then how is that any different to a single type of pgen that works anywhere?

It gives character and flavor each biome?

That does raise a good point though. Why not have a single T1 power generator which appears differently depending on the area it is in?

So if you build it on the coast it'll be a water/wind turbine (depending on if it's in the water or not), if it's near a volcanic area it'll be geothermal, and if it's anywhere else it'll be solar? It will automatically decide for you dependent on it's build location.

This would then simplify base construction because you'd just choose 'pgen'. One choice instead of 3 or 4. T2 would have the fusion reactor, and orbital would have the T1 for any orbit, and T2 for gas giants. Simple and easy!

Well if each kind had different power outputs, with the least common terrain variety having the most output, it would open up different strategies based on where you chose your spawn. Or where you might want to expand to next.

I'd think having different advantages uses and stats would add more to gameplay....but them having models being just flavored by terrain type instead, I would kinda really enjoy that too.

Argument against multiple T1 pgen build options:
— If all T1 pgens have the same stats their appearance does not matter
— If their appearance does not matter, being able to choose which T1 pgen you want to build in the menu is adding unnecessary bloat and complication to the game (X options where there should only be 1)
— Therefore, if we want there to be multiple appearances of pgens (solar, wind, tidal, etc...) having only one constructible pgen available whose appearance is determined automatically by its build location is the best option.

But what if we want the different appearances to have different stats? How do we give the player control without adding bloat to the game?


Implementing different T1 pgen types without bloat:
As I like the idea of there being different stats for T1 pgens, as OP suggested, I came up with this suggestion to reduce bloat:

• Create a color-gradient system for constructing T1 pgens, this would toggle as soon as "T1 Pgen" was selected from build menu
  • Gradient would go: green —> yellow —> red, with green being optimal (or something color-blind friendly... any gradient will do)
  • The system would display the color for the optimal pgen placement (optimal: total energy/day?)
    • e.g., if solar would display yellow, and geothermal would display green, the system would display green
    • As you move your cursor across the gradient, the 'ghost structure' would change to the pgen type that would be constructed at that given location, informing the player of the change
  • If two pgens were exactly equal in value the player would be given the option to choose which one after they placed the structure
    • This could be indicated with a symbol on the gradient, informing the player that constructing a pgen there will create a dialogue box in which the final pgen is selected
    • There could also be an 'order of preferences' section within the game menu. Here the player could sort their preference of pgen types, the computer would then break ties via this preference list
  • Perhaps holding a key would bring up an 'advanced menu' which would allow players to choose the pgen they wanted, even though it was sub-optimal?

This idea allows for different pgen stats without adding the bloat of multiple pgens into the build menu. It also removes the necessity for players to understand and memorize the pgen system in order to be competitive in gameplay. It also accounts for those who will decry the lack of player control, as it includes manual overrides.


Different production styles for pgen types:
This may be difficult to balance, and thereby may be unviable. Also, I may repeat some ideas that have been said by others in this section but I believe I am adding to the mix.
Varying stats for pgens could be as simple altering the rate at which they produce energy with respect to the time of day (or other factors).

• Solar — Peak production: Midday
  • Solar Pgens would have a bell-curve energy production. 0 at midnight, increase as the sun rises, reach their peak when the sun is directly overhead, and drop back down to 0.
  • Give solar pgens the ability to tilt their panels toward the sun and you may end up with more of a 'plateau' curve, a longer period of time at max production
• Tidal — Peak production: Mid-tide (X times per day)
  • Tidal height (position) can be charted roughly as a sine-wave. Therefore the tidal flow is greatest in between tides (as the tide is flowing in or out) and least at the exact moment of high and low tide (as the tide has finished flowing in or out). Taking the derivative of the tidal position gives us the velocity of tidal flow (a cosine-wave). The frequency of the graph will be determined by how many tides occur per day (perhaps determined on a planet-to-planet basis by moon characteristics?)
  • Tidal generators will produce the most energy when the most amount of water is passing them. Therefore (in this approximation) they will produce energy on a cosine-wave.
  • Therefore tidal generators should generate power with a consistent rate of fluctuation (we're not dealing with seasons, right?). This rate of fluctuation should be fairly slow (We get 3-4 tides per day-ish?)
  • This will create a fluctuating power source, but one that will still generate power at night (unlike solar)
• Geothermal — Moderate production at all times
  • I'm not aware of fluctuations in geothermal temperatures. However there may be, and feel free to enlighten me if there are
• Wind — ?
  • I don't know if wind will be implemented into the game, or how it would be modeled, therefore I have no suggestions for how this type of pgen would fluctuate. I'm just going to post some random ideas.
  • Perhaps wind has a low-production that remains constant? Players may choose consistency over energy output?
  • Perhaps wind can be built within canyons (which behave like wind tunnels?) a location that receives too much shade for solar?

Strategic depth created through multiple pgen types:
Solar power (being constructible nearly anywhere) will likely be the dominant pgen type in most games. Therefore there will be less energy available to use during the night. If (as was suggested by OP) unit operation is given an added energy cost, nighttime operations could become a strategic method. Assaulting an enemy during the night if you have a stronger energy economy/storehouse could result in their energy-hungry defenses going offline. You also risk running down your energy stores much more rapidly than you're replenishing them, and it therefore becomes a calculated risk.
If your economy is struggling you may then begin to fear evenings and the threat they bring.

This idea which creates a non-constant energy economy creates reasons for players to seek different energy sources to maintain a more near-constant energy curve. This could help fuel conflict over certain energy sites. If night battles turn into an opportune time to attack then energy sites that provide power during the night will be in high demand (orbital solar arrays, tidal, geothermal, fusion, and gas-giant orbitals). This will create reasons to fight over resources, and losing enough of your power-sources and power-storages could be a deadly turn of events.

(I'm sure there's more strategic depth that is created with this idea, but it's much too late for me to continue working on this post.)

Side topic: Orbital and T2 Pgen
I believe this has all been said before, but I'm just going to throw my vote in too.

Orbital Pgen:
T1 — Solar (Can always face the sun depending on orbital implementation?)
T2 — Highly efficient but restricted to gas giant orbits

Land:
T2 — I liked the idea of the T2 reactor being very efficient while also essentially being a bomb sitting in your base.

Wind is dependant on many things, but if you want a quick and dirty relationship that you can tie to an in game mechanic already:

On anything more than a just the local scale, major wind movement is dependant on three things. Having an atmosphere is one (obviously) and two other effects. Thermal energy from the sun and the coriolis force generated by the planet as it spins. Most planets that are close to the sun have far more 'wind' than those further away. Similarly planets that rotate faster, have stronger winds.

It might be a bit overkill, but a thermal radiation map could be used to determine wind speed. You have a map of arbitrary polygons, similar to Cost Maps, spread across the planet. Each "part" of the radiation map has a certain heat capacity and heat retention, as well as natural heat as a result of the planet's geothermal activity. Parts of the radiation map are heated when exposed to the star, and cool during night time.

For example, a mountainous area would take a long time to heat up, but would retain that heat overnight; a sandy area would heat up quickly, but would loose that heat extremely quickly at night; water would have very gradual, almost arbitrary changes in temperature over the course of several in-game months, due to water's high heat capacity and resistance to heating.

The differences in heat capacity and thermal conductivity would result in temperature differentials, and because hot air moves towards cold air, rough winds could be calculated from the radiation map.

I want to make a mod for this now. Uber, mod tools, kthnxbai.

I don't see much of a reason for wind to be too realistic. Most of the time you won't be paying any attention to it.

Why not have randomly change between two set values created with the planet. Explain to my why it would be better to have a more complex system when the mean wind power generation would roughly stay the same. Should players get an advantage because there area is more windy, or that for the first few minutes one player had lots of wind while the other player has nothing.

Miss-assumption. As indicated in the OP, there is not only one possible implementation for T1 pgens. All the pgens you described are in the same "class" of pgens, being the "environment specific" ones. These are actually completely interchangeable, this goes even as far as only swapping out the model, depending on where the pgen was placed and having the same stats otherwise. That's the same as with the different models used when AA towers and mexes are placed on water, despite sharing the same button. There is IMHO no need to overcomplicate pgens inside this class, the differences are purely cosmetic.

But you also have two other possible T1 pgen technologies available, solar and the metal burner. Both behave completely different from the enviroment specific ones as they feature unique mechanics and suit a unique play style.

Who said it would be efficient?
It's actually less efficient than to go solar, long term. But it eases the need for planet spanning land control since it doesn't suffer from the fluctuations solar pgens have due to planet rotation and it is not affected by the diminishing return mechanics (or limited build spots) the environmental ones suffer from.

Actually I've heard concerns of both sides, the one group of players consideres metal burners a bad choice because it will NEVER pay out in long term games, but others consider it OP at the same time because it would be able to give a great boost at low cost. I say the truth is somewhere in between, there is no situation in where this tech could be considered efficient, but a lot of situations where it is effective.

Indeed, that was one of the drawbacks of wind pgens in TA, they were not reliable and even worse, it felt random. But TA did feature a great mechanic for the enviromental class of pgens, apart from the random fluctuation: Diminishing returns.
Even if the pgen wasn't bound to a specific location, it would still produce less energy when placed to close to other pgens of the same type. So there was e.g. no point in covering a small area in wind pgens since they would hinder each other. Same is also applicable to geothermal or tidal, even without dedicated build spots.

A realistic system to simulate pgen operation:
To clarify, I know all of the concepts you both just posted, but it was my understanding that Über has no plans of introducing weather to the game so I didn't think to include those as factors towards T1 pgen simulation.

However it would be interesting if they did simulate solar radiation behavior to determine energy production on a given planet.
Using the inverse-square law to model the decrease of solar radiation that reaches distant planets, as well as the shape and tilt of the planet to model solar spread they could create a system naturally balances the different T1 pgen systems. Solar would be most effective at the equator of a planet nearest the sun. The locations that the solar radiation impacts (and how long any given location is heated) that is modeled by this system could be used to create MushrooMars' heatmap system, and thereby drive wind currents.
These wind currents would then be affected by the atmosphere density of the given planet, and the Coriolis effect of it as well, and these would drive wind turbines.
Tidal turbines would be affected by the tidal system I described earlier, as well as the amount of solar radiation imparted to the seas and the Coriolis effect. Heck, while we're at it we can throw in thermohaline circulation to drive currents as well, maybe even help ships move faster or slower while they're on the suprface currents.

With this system we could even model nuclear winter and 'mass extinction' events (well, I guess everything is already extinct in this game...) caused by nuclear bomb explosions or asteroid impacts filling the atmosphere with debris. This would force players to zoom in to see under the clogged atmosphere and play in simulated night and pea-soup conditions as the planet slowly ices over.

Personally, all of this sounds awesome to me, but as MushrooMars said, this is where mods come in. (We're not playing "Planetary Environmentalist", after all.) So I'm still at a loss on how to model wind-turbine generation rates without making it too confusing for the player. How do you represent the atmosphere of the planet in a clear way? (I guess if we follow my auto-build/gradient system, it does not matter because the game will determine what to build). But still, with what rate does it produce energy at to complement the bell curve of solar, and sine wave of tidal? (As per my earlier suggestion). Or we can completely forgo giving environmental pgens any difference in the name of simplifying the game and let mods figure it out (likely the path the devs will take?)

I wouldn't call that a misassumption, from what it sounds like we had a terminology misunderstanding. My entire post did revolve around 'environment specific' pgens, I just didn't include that phrase. Additionally you may have noticed that I did not mention the 'metal-burner' pgen idea, and stayed wholly in the realm of environmental pgens. Sorry for the misunderstanding, but I do not see it as a misassumption. Also, you came down in agreement with everything I said in that section:
"These are completely interchangeable" and "There is IMHO no need to overcomplicate pgens inside this class, the differences are purely cosmetic"
(I did not know that AA and mex already change their appearance depending on where they are built (I'm not in alpha). It looks like we're headed in the right direction.)

However, exterminans, I fail to see how solar does not fall under "environmental pgen" under your classification system. Solar power is affected by (as mentioned earlier) the tilt of the planet, and the distance from the sun, environmental circumstances. They also do not provide power during the night.
Unless, in your vision of the system, 'environmental' pgens produce power at a constant rate that is equal to all other environmental pgens. Thereby making them "completely interchangeable" while differentiating them from solar. This would then remove the need to give 'environmental' pgens different generation rates to make them unique. I would, however, not use the term 'environmental' to describe your vision of these pgens, because 'solar' and 'environment' go hand-in-hand.

So, in summary, I think we're in agreement. T1 'environmental-sans-solar' pgens are equivalent. Therefore it doesn't matter how they look because they all perform the same. I still do like the idea of having environmental factors influencing 'environmental' pgens, and I believe I outlined a decent starting point for a system to model that. But, as I said earlier, that would work better as a mod due to the complicated nature of the system and the unnecessary depth the player would then have to learn to properly manage their power).

Regarding the metal-burner? I'm not going to touch that subject. As I'm not in alpha I have no idea what the balance(1) looks like, or how that would alter it. Also, I've seen a lot of dislike towards the metal-maker (I know it's the opposite, but it's a similar idea), and I'm not going to get caught up in that mix.
(1): I'm aware there is no balance, but being in-game would allow one to have a general idea of what to shoot for

Why would it be less efficient than going solar? As far as energy/cost? It doesn't cost anything to run, does it? It would seem to me to have an expensive up-front cost, but then give a large energy creation, whereas solar would take longer to get you energy, but would be very cheap as far as up-front investment. Therefore the reactor would *eventually* be more efficient than solar. (Kinda like buying a new hybrid car to save on gas... in 10 years)