Ye Ol' Reactor debate

[Cyborg]

I haven't been adding anything, mainly because so far the ideas seem awfully crazy. Not going to get worked up over something that's clearly out there.

Let me know in big bold letters if we start to take any of these revisions seriously.

[Toranth]

2) Reactor HP/armor should have nothing to do with their balance. Reactors are not a combat unit, players are going to build the most efficient one and just rebuild if they lose one to an attack.

I disagree with this wholeheartedly.  By this logic, nobody would ever use anything but the Economic Orbital Command.

Let's be honest, 90% of the time the Military Orbital Command isn't doing much of anything, but that 10% of the time that it is makes it worth building in a vast variety of situations. 

I do think that if we are to make individual types of reactors, there has to be a benefit and drawback for all of them.  In short, a simple efficiency vs. utility ratio:

Type 1. "The Sun"
-Produces a huge amount of energy.  Fragile and must be built a certain distance away from the Orbital Command Station on a planet.  When it explodes it takes your Orbital and Shields with it.  The concept behind this idea is that this reactor, though powerful, must be built on the edge of your forcefield, and when it dies it immediately gives the AI access to the next planet.

This is little different than a ZPG - Fragile, expensive, with a major drawback if lost - so build in secure zones.  Unless the AI is given a special raid priority against these, there'd be little worry about them.  If they're expensive enough, or have a large enough penalty, to need heavy protection, then they're unlikely to be worth the K needed for them.  And as mentioned, this competes directly with the ZPG.

Type 2. "The Runner"
-Produces a medium amount of energy, but is mobile.  Moves at an average speed.  Good for running before you lose your forcefield or kiting an enemy to take some pressure off of your Orbital Command.  Can't move through wormholes.

If it can't move through wormholes, it can't run.  All it can do is die tired.

Type 3. "The Phantom"
-Produces a small amount of energy but is cloaked, and is therefore unlikely to be detected by the enemy.

This one could be useful in border systems.  If the K cost were not too high, I'd consider this one.

Type 4. "The Trap!"
-Produces a medium amount of energy, but when it explodes it sets off a planet-wide EMP to give your units some time to respond.

First, unless it would be a long EMP, it wouldn't help much.  Second, if it's a AIP-free EMP, what's to stop it from being abused?  Finally, So many AI units are immune to EMP that using EMP for serious defense is usually a bad idea.  It's usually better to just scrap your units - at least you get something out of that.

Type 5.  "The Tank"
-Produces a miniscule amount of energy but has huge amounts of health and armor.  Used as a decoy since it is still a high-priority target for the AI, meaning they will focus it giving you time to respond.

Ok, this one I need to ask about:  Why?  The AI doesn't have a high targetting priority for reactors, so having a tough-to-kill reactor just means it takes the AI longer to clear the system after destroying your fleet, turrents, Command Center, specials, factories, etc.  If you force the AI to target these with a high priority (a game mechanic I'm interested in), it potentially breaks the game in that the player could manipulate the AI's in-system movement too much.  Can you imagine if an AI fleet arrived and was forced to move en-masse away from the command station to destroy a silly little reactor, all the while being bombarded by turret and such, before finally destroying the reactor and being permitted to turn around do something useful?

Type 6. "Special Beam Cannon!"
-Produces a low-medium amount of energy, is decently durable, and channels its energy into a powerful beam (think Heavy Beam Cannons) that annihilates foes.  Only loses 25% damage under forcefields.

Would there be a limit on the number of these you could produce?  If not, then you have an unlimited number of turrets that can be built.  If so, how is it different than just another turret + a reactor somewhere?

Type 7. "The Turtle"
-Produces a low amount of energy, creates a moderate shield around itself.

Again, would these be limited in number?  If not, FFs for everyone!  If so, how is it different?


Most of all, I see all of these ideas as mixing the idea of Power with other, unrelated, things.  That seems to me like an unneccesary complication that still doesn't do much to solve the problem.  Although the cloaked reactors may be worth considering, no matter what else may be done to the power system.


Personally, the best idea I've seen in the thread is the one about limiting the supply of power to fleets away from the controlled worlds.  It would actually provide a certain level of logistics to the game, which is purely strategic.  Imagine needing to protect a supply ship while raiding into AI territory.  Imagine all the ways that things could go wrong...
That said, I don't really have a problem with the way power works right now.  I just ignore it, like Armor.  Only on Golems Medium do I ever have to really think about how I'm going to supply my units.

[Wingflier]

Personally, the best idea I've seen in the thread is the one about limiting the supply of power to fleets away from the controlled worlds.  It would actually provide a certain level of logistics to the game, which is purely strategic.

As Keith said, the knowledge and logistics cost of this idea are a disaster.  AI War is already complicated enough for new players, you're adding an extremely complicated new mechanic.  What happens when your energy gets taken out while your fleet is at max distance?  If there's not enough energy to power them anymore, then your entire fleet is stranded and there's nothing you can do to help them.  Sure, it adds more strategy to the game, but at the cost of a huge burden of knowledge for the player and an entirely new (and frankly unnecessary mechanic).

The idea I'm presenting (which wasn't actually my idea to begin with) is actually hugely simplifying the game.  Instead of having to teach players about this confusing mechanic of 3 Reactors on each planet, some of which are more efficient than others, but only in terms of cost effectiveness instead of power, oh but which you can continue to mass produce anyway, even though you're not supposed to - I mean I've trained at least 5 players to play this game and it's still a very non-intuitive mechanic.

The mechanic I'm suggesting is simple - you get 1 energy reactor for each planet:  Choose wisely.  Done and done.  The benefits for each Reactor seem obvious, and if they aren't you tweak them.  Energy would actually be much easier to balance this way since there would be so much less nuance in terms of how much you actually could have if you spammed them in spite of the efficiency tax.

Ok, this one I need to ask about:  Why?  The AI doesn't have a high targetting priority for reactors, so having a tough-to-kill reactor just means it takes the AI longer to clear the system after destroying your fleet, turrents, Command Center, specials, factories, etc.  If you force the AI to target these with a high priority (a game mechanic I'm interested in), it potentially breaks the game in that the player could manipulate the AI's in-system movement too much.  Can you imagine if an AI fleet arrived and was forced to move en-masse away from the command station to destroy a silly little reactor, all the while being bombarded by turret and such, before finally destroying the reactor and being permitted to turn around do something useful?

If energy was a more finely-balanced, and therefore more important resource than it currently is (not saying it isn't important, but it's not a first priority for the AI), then it would make sense for them to target it more often. 

I'm not saying the AI would stop whatever they're doing to take it out, but if popping just a few of them could shut down your whole grid, it would be quite a priority wouldn't you say?  Even more than Orbitals are currently.  That doesn't mean they would ALWAYS go for them, but it would add some new options to the AI's repertoire.  Maybe you could even make a new AI type called "The Blackout" which specifically targets your Energy Reactors.

Maybe my ideas for Reactors could be better, but I think you're getting too focused on the details and not the main idea.  Simplifying how energy reactors work, while giving the player a lot of diversity and strategy in choosing one for the best situation could be just as, or more interesting, than the Orbital Command Series added just a couple years ago.

[keith.lamothe]

Let me know in big bold letters if we start to take any of these revisions seriously.

What I am considering seriously at the moment, as a major step but not necessarily the last, is:


1) Add new "Energy Collector" building:
- built from same menu and in same way as existing reactors
- cap 1 per planet (scales up with multi-HW games; edit: might actually have it scale the output, instead of the cap, to avoid clutter)
- requires friendly planet
- produces the equivalent of a current EnergyI + EnergyII + EnergyIII
- costs nothing to run (no m+c, etc)
- no stacking efficiency penalty (not that there would be, with a cap of 1)
- cannot be put in low power

2) Add new "Matter Converter" building:
- build from same menu and in same way as existing reactors
- no cap
- requires friendly planet
- produces the equivalent of a current EnergyI + EnergyII + EnergyIII (I'm very open to feedback on this number, just seems like a good starting point)
- costs m+c to run such that its e/(m+c) is near that of a current EnergyII running at minimum efficiency (so like the 10th energyII on a planet right now; the inefficiency malus has capped out long before that point iirc) (again, very open to feedback, the idea is that it's inefficient, but limitless)
- no stacking efficiency penalty
- cannot be put in low-power; you can scrap them if you no longer need the excess capacity, but no quick twiddling back and forth (we've found the twiddling to not be much fun)

3) Remove the existing EnergyI, II, and III from the game (probably converting them to Matter Converters when converting old saves; the player can then build their Collectors and scrap whatever Converters they don't want; open to feedback on a kinder way to handle the transition)

4) Make low-power mode no longer reduce energy expenditure on anything (I'm open to feedback on whether golems should still take less e on low power; I can see both sides of that)

5) Something of an aside: rebalance the ZPG to be competitive in the new model (probably take its token (m+c)/second cost off, and certainly boost the output)

Done   After that, if we want to have a cloaked variant of the collector for hard-to-defend planets, or whatever, we can figure it out from there.  I just think that the above is the core step towards a more sane energy model, and would be a very positive thing even if that's all we did in the near-term.


All that said, if your reaction is "GAHH!" then I'll just move on to something else and see if another major contender comes out of this discussion

[Cyborg]

Let me know in big bold letters if we start to take any of these revisions seriously.

What I am considering seriously at the moment, as a major step but not necessarily the last, is:

All that said, if your reaction is "GAHH!" then I'll just move on to something else and see if another major contender comes out of this discussion

I'm thinking. What you wrote isn't that offensive, really. It could work. I just need to turn it over for a day or so.

[keith.lamothe]

What you wrote isn't that offensive, really.

Kind of like "well, that smell isn't quite as bad as when they're cleaning out the chicken sheds a mile down the road..."

[Varone]

2) Add new "Matter Converter" building:
- build from same menu and in same way as existing reactors
- no cap
- requires friendly planet
- produces the equivalent of a current EnergyI + EnergyII + EnergyIII (I'm very open to feedback on this number, just seems like a good starting point)
- costs m+c to run such that its e/(m+c) is near that of a current EnergyII running at minimum efficiency (so like the 10th energyII on a planet right now; the inefficiency malus has capped out long before that point iirc) (again, very open to feedback, the idea is that it's inefficient, but limitless)
- no stacking efficiency penalty
- cannot be put in low-power; you can scrap them if you no longer need the excess capacity, but no quick twiddling back and forth (we've found the twiddling to not be much fun)

If you were to do this couldn't we have matter converters that just convert metal and another type that does just crystal, similar to the manufacturers we have currently.

[Minotaar]

I like Keith's proposal as it very clean and simple while leaving lots of space for improvement.

One question I immediately have though is the cost of the Collector. Is this something we intend to be on every human-controlled planet or is there supposed to be a choice between having or not having it? Pretty sure it's supposed to be the former, as having time be the only cost associated with something proved to not work out very well in the past. So we might as well make it cost 500/500 or something and assume we get some 100k free energy per planet when we balance the Converter.

The Converter should probably also scale its production with the number of homeworlds so as to not make people just build 8x of them.

Making low-power mode not reduce energy consumption is kind of dangerous, I think. If you had just enough energy and you lose a reactor (either one), your options are to either scrap some things or build a Converter somewhere. That doesn't sound too bad if the Converter doesn't cost too much up front. But the real problem here, I think, is that being in a state of brownout is terribly dangerous to the point that there has to be a quick way out of it. The energy mechanic is basically supply from any other RTS, and I haven't seen an RTS that punishes you for losing supply (except not being able to produce any more units). So what if instead of completely shutting down all of the defenses, having negative energy would simply reduce their power (smth like decreasing the fire rate of turrets) and make forcefields flicker periodically (for dramatic effect  as well as still providing some protection, even if unreliable)? That would make you slightly more vulnerable during an extended period of time required to bring power back up, which seems more interesting than the current system, which makes you extremely vulnerable for a very short period of time.

On the other points, not sure how to convert old saves until we figure out the stats exactly (namely, the Collector question: if it's free, we should just put one everywhere and then plop down some Converters to make an amount of energy close to what the player had previously). For the golems, I have a different proposal: make them still reduce energy consumption severely (by 75-80%) when powered down, but require time (upwards of 5 minutes) to bring back online. That would solve the twiddle problem (which is very much present with golems), as well as require players to use golems more wisely, as having them always ready for action would entail paying a serious upkeep for repairing and the needed energy. It is very fitting to their superweapon nature as well.

Last, but by no means least, if we are to have a structure called the Matter Converter, its energy output cannot be anything else than E=M*C^2 

[Hearteater]

My only concern with unlimited Matter Converters is they can all be put on your most defendable planet.  Can they be powered down?  If so, I can build 30 on my homeworld and never run out of power regardless of what a wave might do to me just by turning on the necessary MCs.  I presume the existing hotkeys for power management would toggle MCs on and off (assuming they can be powered on and off)?

A somewhat related concern is without a per-system limit, someone can find a way to cheese the current energy limit while running with very low system counts.  Basically you have unlimited energy even if you have just your home system, you just need a way to get enough m+c to power your ships.  I can't think of how someone could do it at the moment, but I'm sure someone will figure out how to run two golems on only three systems.  Is there any reason a cap of say 3-5 wouldn't work?  Do we really need to be able to drop 20 MCs on one planet?

[Mánagarmr]

- costs m+c to run such that its e/(m+c) is near that of a current EnergyII running at minimum efficiency

That's awfully close to Einstein...

-All that said, if your reaction is "GAHH!" then I'll just move on to something else and see if another major contender comes out of this discussion

It actually sounds like a great idea! I don't really have much to add to it.

[keith.lamothe]

If you were to do this couldn't we have matter converters that just convert metal and another type that does just crystal, similar to the manufacturers we have currently.

Sure, we could do that at some point.  I'd prefer to start with just the m+c=>e one to keep this as simple as possible: I really don't know if this whole thing is going to "work" in the "is this fun?" sense, so I'd rather have the initial core of it hit playtesting with as little extraneous (and potentially tester-distracting) issues as possible.  That's also the reason I'm not talking about variants of the collector (high energy, dangerous; low energy, cloaked, etc) right now: not that I'm opposed to the idea in principle, just that it's an unnecessary layer of complexity if the immediate goal is simply to make the energy model no longer encourage so much micro (while not trivializing it completely).

One question I immediately have though is the cost of the Collector. Is this something we intend to be on every human-controlled planet or is there supposed to be a choice between having or not having it? Pretty sure it's supposed to be the former, as having time be the only cost associated with something proved to not work out very well in the past. So we might as well make it cost 500/500 or something and assume we get some 100k free energy per planet when we balance the Converter.

A friendly planet without a collector with be an aberration, yes.  But:
- I expect that there will be different variants of the collector you can put down later, so there'll be choice there.
- I don't want these to be easily insta-rebuildable so that if the AI takes one out it isn't just a matter of pause;place;engies;unpause;back-to-normal

The Converter should probably also scale its production with the number of homeworlds so as to not make people just build 8x of them.

I'm not too worried about that.

Making low-power mode not reduce energy consumption is kind of dangerous, I think. If you had just enough energy and you lose a reactor (either one), your options are to either scrap some things or build a Converter somewhere.

Yep   We want the loss of energy production from AI attack to be potentially disastrous   That's one of the requirements that's made it take so long for us to make progress on the question of how to improve the model.

More importantly: as long as low-powering stuff can affect current energy budget, there will be motivation to micro what's on and what's off.  That has had a simply hateful impact on the amount of micro required for optimal play, and I want it gone

My only concern with unlimited Matter Converters is they can all be put on your most defendable planet.  Can they be powered down?

No, as I said, these could not be powered down: they would run until scrapped (again, getting rid of the twiddling micro).  Scrapping could be done instantaneously, but rebuilding would take some m+c and time.  That time window is the danger for the player.

But yea, they'll presumably all be dumped in a defensible position like the existing manufactories; that's fine.  My hope is that people won't be leaning to heavily on converters except when they've suffered catastrophic collector loss, at which point I'm not wanting to kick-them-while-they're-down by making it necessary to put some converters in exposed positions.

A somewhat related concern is without a per-system limit, someone can find a way to cheese the current energy limit while running with very low system counts.  Basically you have unlimited energy even if you have just your home system, you just need a way to get enough m+c to power your ships.

This is already possible: the stacking inefficiency penalty caps out and additional reactors of a particular mark on that planet all produce the same minimal e for the same m+c.  As I said, my intended efficiency for the converter is that of the current EnergyII at that "minimum efficiency" level.  I imagine it will probably change from that point to be a little less punishing, but the idea is that these won't be any more cheesable than the current system.

That cheese is mildly annoying from a design standpoint, but ultimately either there needs to be a way to get unlimited energy from a single planet (at exorbitant cost, as it is currently) or a lot of things about the game will likely need to be changed.  I'm content to leave that particular facet of the energy model as-is.

[Fruny]

I'm not going to much help with the mechanics as I am a complete neophyte, but here's a not-so-random flavor idea anyways which, reskinning aside, might suggest mechanics : what if power really is computational power. I believe something similar was suggested upthread.

Could the AI hack you back or spread viruses, taking over some of your fleet, turrets, by subverting your computer systems (reactors) even though you still have power to spare ? After all, you do raid its computer infrastructure yourself, for various benefits. That may also explain why coprocessors interfere with supply.
What happens if you have huge amounts of compute power lying around ? Could you (caveat computor) end up ultimately unleashing another AI ? Isn't that how we got into this mess in the first place ? Without going for anything so drastic, you might unleash some nastiness based on unused power (in particular if you disable turning "reactors" off), or on excessive reactor concentration (inefficiency penalty could contribute to effective AIP if that wouldn't make it too volatile.)

There has to be something to consider beyond just balancing m+c vs. energy when deciding to build that additional reactor or to capture that ZPG.

[chemical_art]

I like the idea, but what is the goal?

If the idea is to allow more variety then it succeeds (this allows new energy reactor types)

if the idea is to reduce micro it fails miserably for you need to take more activity to reduce power. Except to reduce power now instead of a flick of a keyboard button you pause the game to remove units. Only machoists wouldn't pause the game rather then tough it out due to how critical energy is for defense, and low power only occurs while on defense the vast, vast majority of the time.

[Cyborg]

Given this much thought, and while I do see some abusable corners, I think it's better than what we have. I vote yea on Keith's revision. I think it's going to need some fine-tuning after implementation, so make sure that you do this when you have the time to adjust the numbers afterwards given player feedback.

[PokerChen]

The new base mechanic will still increase micro for me - I'm used to attaching the hamster to modified ~-keys for quick switching. If manufactories could be manually powered with keys I'd do that too. Removing the twiddling doesn't actually solve micro issues - your currently proposed mechanic simply imposes a greater micro penalty for playing above the energy collector threshold and nothing below.

Granted, that's a lot of power until one is using golems and spirecraft. It might be easier just to switch the twiddling to the new converters - besides, you ARE talking a rather low efficiency as a starting test value, correct?