Ye Ol' Reactor debate

[Wanderer]

Yeah, not really my best friend in general, the concept of transmittable power through wormholes.

I still question the whole lore and mechanics of power in this game in general.  I realize there needs to be *some* form of cascading failure through your network if you push things too close to the edge on planets to ship ratios, and it's as a limiter for some of the golem balancing for Medium.

Microwave power pushed through the universe though just doesn't seem to make sense to me.  Locally perhaps, I'll give it soft SciFi for localized power distribution (turret networks, forcefields (which would block its own power)), etc.  But generically, not so much.  However, that doesn't mean I have a better idea, though I'd like to clarify what they DO do, from a game balance perspective.  This might help us come up with better ideas.

What do reactors do...
Reactors are a way to force defensive systems to start failing when the balance between defense and offense is pushed too high towards offense for the amount of defensive territory being protected.  This failure can then cascade to other defensive systems on any breach of a defended system.

This cascade either won't occur, because the remaining defensive territory can support the existing infrastructure, or if you're on the razor's edge will start a micro-management cascade limiting options that were 'just fine' before, causing a significant period of rebuilding and defensive work, if not actual game loss.

Reactors also enforce a continuous cost on otherwise 'build once and forget' free ships.  Ships particularly notorious for things of this nature would be Cloakers, Science vessels (which can't power down), non economic command stations, and long range turrets, amongst others.

This is the actual game design limiter that reactors and energy perform.  Big ticket items, like Golems on Medium, can cause significant reliance on overwhelmed systems, allowing for failed defenses to cascade that much faster.  This issue can be leveraged against multi-reactoring, which allows for a diminishing returns result from previous investments to allow a player to 'cover' a failure in another system, albeit at higher prices.

I'd like to open up a discussion to see if we can come up with any ideas that can help to balance these items that would fit better both into the lore and into opening up more interesting gameplay.  I'll start with my own brainstorms on each piece:

First, the obvious pieces:

Energy causes an artificial ship cap in the early game:
This could easy be solved by existing lore.  Instead of relying on power to enforce this, force the player to create Automation Stations allowing for a certain amount of ship power to be controlled.  This could be directly swapped for the currently energy required to power a ship.  By splitting it off from the rest, it allows for more independent tweaking.

Defensive Cascade Failure
I'd like to see this swapped out a little.  Instead of defensive failures causing the offense to destroy defensive measures, allow it to shut down the over-extended offense.  A little tweak to this mechanic can help deal with the whole 'deep strike threat' problem too, where a single raid starship causes as much threat as your entire fleet, particularly if you're in two systems at once, doubling the deep strike response.  Have your stations act as repeaters for the above mentioned Automation Stations.  The further away you are, the more 'resources' from the Automation Station the it requires to control the same ship, due to power decay through wormholes.

Make it reasonable, say 90% at 3 wormholes out, your average strike range from any friendly planet.  At 4 out it drops to 75%, and gets a whole lot worse from there.  This would prevent you (without inane levels of automation investment) from striking 10 worlds deep with your entire fleet but it could be reasonable to send a raiding party out that far.

Satellite worlds are no longer as 'big a deal' to simply neuter, but create a defensive choice as to holding it for repeater value or simply muscling past it, allowing for the concern of the loss of a repeater world meaning that you'll only be able to retreat your forces piecemeal.

Cost of operation
I don't believe that ships that aren't actively in use should continue to charge 'fees' for simply existing, particularly since by lore these things are human controlled either via dumb AIs or remote control.  I'm open to ideas and counter arguments though.  Turrets and Forcefields 'cost' by the fact that under most circumstances all but the snipers end up having to be rebuilt at some point.  I could see an ammo cost or something for them, and maybe the near-useless cleanup drone could be repurposed for something like that so they can't get completely out of hand (in particular the snipers).  It wouldn't be a hard ammo value, it would be more along the lines of something like Engine Damage.  When they reached 0 'ammo health' they'd need to be 're-ammo'd' or whatever.  Shorter range turrets and ffs, when they're actually used, are costing you in rebuild prices.   Let them get their re-ammo then.

For cloaker starships, science vessels, etc... have them simply cost 'staff' when they're actually doing something besides sitting under a FF hanging out at a friendly command station.  Otherwise take them out of the computation.  Heck, if it's not moving, shooting, or researching, take everything out of the computation.

So, I'm not really excited about my brainstorm on the last component there, I realize SOME things need continuous costs... but there's GOT to be a better way.  Even if it's not as simple, we can figure out a way to make it more interesting, I'm sure.  Replaying DF has made me remember something about why I liked the design of this game.  Each emergent component, by itself, is simple enough.  The combinations is what makes the complexity interesting.  As long as whatever we can come up with can be swallowed in bite sized pieces (UNLIKE DF, where half the difficulty slope is the !@#!@!!!!!! interface and lack of manual) I believe it could be reasonably integrated... well, with the local AIW gods' blessings, anyway... they're the ones who'd have to code it!

[TechSY730]

I don't really have any ideas about the first two topics, but I thought of something for the continuous costs thing.

How about a cue from the Total Annihilation/Supreme Commander/Spring RTS model of energy consumption for military units and buildings?
Each military unit has a base energy cost, but it isn't a huge one. However, firing their weapon makes their energy spike quite a bit. This way, units that aren't doing anything don't take up too much power, but you start to feel the pressure when those units are in active combat. (Engineer, support, and other non-military units are a bit trickier, I can't remember how their energy is handled off hand)
I think one of the things that made this system work in those games is that energy was a flow based resource, instead of the "total produced - total used" type resource energy is currently. They handled energy similar to how metal and crystal are handled now.
May take too much tweaking and balancing to make this work for AI War, and it might not even fit this model of the game, but it might be worth considering.

[TechSY730]

Oh,and for some previous discussion on these types of issues, Defense Supply Mechanic. Although the OP of that thread is about something tangentially related, many of the things mentioned in that thread cover many of the points in this thread's topic.

[Drjones013]

Just reimagine reactors as human-controlled processors. It directly translates: diminishing returns on excessive amounts of processors due to bottlenecking (the command center can only handle so much), ships have their own reactors (engines) for power, catastrophic failure of defense systems because the computer just destroyed/hacked your HPC....

Could lead to an entire new command center type. And a new AI ship that hacks your systems. Does the future have good anti-virus programs?

[keith.lamothe]

The energy system is one that could stand some thinking outside the box   The basic model has had many variations tried early on in AIW development and a bunch of alterations considered since then, but none that seem demonstrably better (or even as-good) in terms of avoiding worse problems.

I like the idea of a per-planet resource, potentially with a "transmittal" to neighboring planets that isn't 0 (like the defense-supply/fire-control mechanic we discussed in recent months) and isn't infinite (like the current system), but that does add a substantial amount of complexity to the game.  Particularly if mobile ships could just suddenly shut down because they moved through a wormhole (or be denied traversal) because there's not enough special sauce available on the target planet.  That rule would probably be a significant pain code-wise and nearly impossible to make work in a fun way for the player.

But I think there's some good stuff here, so I'm listening

[Hearteater]

I don't feel energy limits early fleets at all.  In fact, unless you are playing 9+, it hardly limits much at any point in the game.  Taking 5 planets in the first 1-2 hours isn't unusually in average difficulty games, at which point you can take 1-2 ARS and operate at full ship cap without problems just by double-reactoring (and you don't even need to do that if you keep your turret count down).  I see the main problems with energy as:

1) It is tied to other resource system (it reduces m+c income) making it hard to balance independently.  It is effectively a derived resource so it can't ever be a separate balancing lever.

2) Energy is only a problem when your fleet gets larger.  To get a larger fleet you need to capture an ARS or research a higher Mark of an existing ship type.  Research costs Knowledge.  To get Knowledge you need to claim a new system.  So nearly everything that that increases your need for energy also provides you more energy.  As a result, energy is only a problem when you start losing system, it is never really a general concern.

3) Units can be powered down, effectively making the danger of running out of energy a hollow threat.  With active pause you can always pause on a brown-out and power down the exactly correct units to keep your shields operating.  This is an even greater issue with reactors being powered down because it gives you the ability to have nearly unlimited emergency power.

To improve the energy system, I would consider the following:

a) Remove the m+c cost of reactors completely and cap the number of reactors in a system.

b) Re-design Reactors to work like Command Stations.  There are multiple types of Reactors, each with pros and cons.

c) Powering down ships no longer reduces energy consumption.  It still has all the other effects (such as dropping shields and holding fire).

d) Re-balance energy costs, in particular for turrets.  It may be ideal to pull energy costs from turrets entirely, or maybe to just 10% of their current values.  Push the offensive unit energy costs up to compensate, making low energy ship types an interesting strategic choice.

To expand on a Reactor re-design:

Their m+c cost to build would be high as would their construction time, making losing a reactor very noticeable.  Something around 50-75% of a Mark I Starship in terms of both resources and time.  They are limited to 3 per system.  Ideally they should have a restriction on being placed too close to each other, with the goal being 3 could just barely fit in a triangle under a Mark I Force Field, but that really just makes defending them more interesting.  Reactor types (for starters) with a target energy production values of 180k/system:

Basic Reactor - Reasonable health, no armor.  Generates 60k energy.  This is the balanced middle-ground reactor.
Armored Reactor - High health and armor to survive repeated attacks.  Generates 48k energy.
High Energy Reactor - Very low health, no armor.  Generates 72k energy.
Stealth Reactor - Low health and cloaking.  Generates 40k energy.
Home Reactor - High health, no armor.  Generates 72k energy.  Cannot be built, player starts with three in their home system.

[Varone]

Couldn't we have a supply resource similar to M&C.

Rename reactors to supply generators and have the same mechanic but the supply is generated over time and all ships use up supply.

More supply is needed the further out an offensive ship is so deep strikes are fine at first but when your supply runs out ships are going to start being forced into low power mode. perhaps even a CTRL so that available supply goes to offensive first then defensive or visa versa.

Have the supply generators store a certain amount of supply so the more generators you have the higher your max supply but more than one supply generator on a planet has diminishing returns and eats M&C just like reactors.

From a story point of view you could say that supplying ships further out is more expensive because of lost supply ships going through AI territory and the length of time needed to get there.

[keith.lamothe]

Couldn't we have a supply resource similar to M&C.

Rename reactors to supply generators and have the same mechanic but the supply is generated over time and all ships use up supply.

More supply is needed the further out an offensive ship is so deep strikes are fine at first but when your supply runs out ships are going to start being forced into low power mode. perhaps even a CTRL so that available supply goes to offensive first then defensive or visa versa.

Have the supply generators store a certain amount of supply so the more generators you have the higher your max supply but more than one supply generator on a planet has diminishing returns and eats M&C just like reactors.

From a story point of view you could say that supplying ships further out is more expensive because of lost supply ships going through AI territory and the length of time needed to get there.

How would that be different than simply having all ships have an m+c upkeep-per-second (that increases as they go further out, perhaps)?

Not saying it wouldn't be, just a bit low on mental energy right now and seeing if I missed something

[Varone]

How would that be different than simply having all ships have an m+c upkeep-per-second (that increases as they go further out, perhaps)?

Not saying it wouldn't be, just a bit low on mental energy right now and seeing if I missed something

For the same reason that having the AI trash your systems can lead to a defensive cascade failure. If the AI hits your supply then your going to quickly run out of supply for shields and offensive ships. Plus having m+c upkeep would encourage people to put ships on low power all the time when not using them and would be a headache if you want to speed build something.

[doctorfrog]

This talk is pretty fascinating, but I have to admit that it would make my head swim in-game. I am glad that the OP started with the question, "what is the function of the game mechanic of energy requirements? How can we make it more trouble, yet more fun?"

Rather than, "the energy system in this science fiction game is unrealistic! Here's how to make it more realistic according to me and me alone, and dramatically reduce the amount of fun in the game!"

So that's good. Obviously we'd have to be very careful about how to employ more complex energy management, to keep it simple to understand and fun enough to encounter, rather than a complicated chore that has you clawing at your scalp during an energy crisis.

My contribution is not a suggestion, just another energy mechanic I saw in a game: Startopia.

It wasn't very intuitive at first, but I think the way it worked was that a single "Energy Collector" was able to provide 100,000e. The output was continuous, in that it could power buildings with a drain of 100,000e or less.

You could buy more Energy Collectors, but they were prohibitively expensive and were basically just energy banks; they didn't come with 100ke, you had to fill them up with surplus energy, to a maximum capacity of 100ke.

That 100ke could be expanded, however, I think by increments of 25ke, by adding "Energy Boosters." These would give you free e for the purpose of powering buildings only. What kept you in check, though, was an ongoing cost of 25e per 30 seconds as these things ran. You couldn't just put them all over the place, because though they expanded your network, they were themselves net losers.

So, you could actually have much less power than needed in the "bank," and still have things running along quite well, as long as you could count on a growing surplus over time.

The Energy Boosters also acted as pylons. If there were issues with the power grid, the nearest buildings to the Energy Collector or an Energy Booster were given priority power, with peripheral buildings losing power sooner.

The rub was that e was also the in-game currency, so you could have an economic crash that also affected your energy grid. Startopia is essentially a management game, so it kind of worked. You invested in money-making and maintenance buildings with e, then hoped for a net gain.

This allowed you to really jack yourself up by overinvesting in buildings that didn't make money back quick enough, or really suffer if you came under attack and didn't have enough security (a net loser that requires a substantial up-front investment). Then you'd start seeing brownouts all over the place. Much of the game tension was in making wise purchases and setting them up in the perfect spots to attract tourist money, while also protecting those investments from disaster.

Obviously, not all of this translates to AI War, but it's a concept of non-trivial management of constantly flowing energy.

edit: see also http://www.rakrent.com/rtsc/html/star04.htm

[Wanderer]

The energy system is one that could stand some thinking outside the box   The basic model has had many variations tried early on in AIW development and a bunch of alterations considered since then, but none that seem demonstrably better (or even as-good) in terms of avoiding worse problems.

Hey Keith, what's the core reason for the energy mechanics?  I know what they've ended up doing, but what is the original design intent behind it?  With that information to hinge from we could probably come up with some approaches to the core concerns that would end up more acceptable to you gents.

[keith.lamothe]

Hey Keith, what's the core reason for the energy mechanics?  I know what they've ended up doing, but what is the original design intent behind it?  With that information to hinge from we could probably come up with some approaches to the core concerns that would end up more acceptable to you gents.

I played a bit back in the 1.x days but didn't actually work here until the 3.x days, so Chris did all the major iterations of the energy system.  So really it is more a matter of what it's doing now.  The main things we want to avoid (off the top of my head, may be missing something) :

1) Having energy both simulate and play itself; abstraction is preferable to automation there.
2) Adding a lot of extra complexity; going from a galaxy-wide resource to a per-planet resource is almost inevitably a pretty big increase in complexity.  A per-planet resource that partially transmits to neighboring planets is even moreso.
- More complexity isn't necessarily a no-go, but the "fun payoff" will have to be pretty high to justify it.
3) Lose the global-population-cap aspect of the mechanic.
4) Otherwise making things less fun.

On the scientific realism thing... yea, cross-galaxy power seems weird.  But so does cross-galaxy m+c.  And from the other side I think energy is a pretty good word for "what makes it go" for just about everything with an energy cost (and things without it can be thought of as having onboard reactors, etc).  Whereas "control" or whatever doesn't quite fit for all of them.  Ultimately, while I'm happy to move towards better immersion where it's an obvious benefit all around, it's heavily secondary to the gameplay itself.

I like Hearteater's idea earlier in this thread; not sure about all of it but I think it's something that could work and be more interesting.  Out of curiosity, are there any strong objections to it?  I imagine the hard-cap on possible energy production would be a problem for some.

[Wanderer]

I don't feel energy limits early fleets at all.  In fact, unless you are playing 9+, it hardly limits much at any point in the game.

I know I have a VERY biased view on the game, but my opinion is if you are playing at 'edge' difficulty (Edge being wherever any particular player finds the difficulty balanced against them at) that you will eventually need more fleet than you have energy for.

Even if you simply consider first world builds, crank on a few starships and you'll hit base limit.  Don't K up in manufacturies and you'll stall out your economy pretty heavily with heavy reactor builds, which is pretty easy to do if you start MK II'ing Bombers, Fighters, and special ship right out of the gate.  It definately controls choices, since you can't feed the beast, per se, in the golden hours.  At least not until you've revved up the AIP.  After that I agree, it's golden hours only for that limitation unless you're playing low AIP (or medium golem) games... which is a different discussion.

1) It is tied to other resource system (it reduces m+c income) making it hard to balance independently.  It is effectively a derived resource so it can't ever be a separate balancing lever.

This I agree with.  One of the reasons I would like to, eventually, unhook Energy from M+C is I'd like to discuss a series of tweaks to the M+C system eventually.  Right now it's a feast or famine kind of thing.  I have a few core ideas on how that might be able to be better balanced but it needs to be a unique component centered around construction only.  But well, that's a long term discussion.

2) Energy is only a problem when your fleet gets larger.  To get a larger fleet you need to capture an ARS or research a higher Mark of an existing ship type.  Research costs Knowledge.  To get Knowledge you need to claim a new system. 

Actually, you don't.  K-Raids and spire Archives.  But I know I'm splitting hairs on your point with that...

So nearly everything that that increases your need for energy also provides you more energy.  As a result, energy is only a problem when you start losing system, it is never really a general concern.

I semi-disagree.  Special operations units (Mob of Science IIs, Transport Herd, Cloak brigades, Fortresses) all outweigh their energy concerns compared to the affect they have.  Fortresses are a good example.  At 90k in Energy, and 3000k in research, for your one planet worth of gain (125k in energy, 225k at double reactor and 20% loss of conversion) you're using up 2-2.5 planets at double reactor worth of power.  Throw in a stack of transports simply so you can get your scouts out that much further and you're looking at a massive energy spike with no compensation.  Investing a K worth of knowledge into a few turret upgrades can have a similar effect.

However, once again it all comes down to system loss causing the issue, because the simple fact is you can't build what you can't power.  Thus, Energy takes on a 'defensive failure' mode because FFs are the first thing to go.  Now, you can micro around and start de-activating your fleet to bring your brownout back online, but the end result is the offensive 'push' of the forces (that you're maxxing energy out creating ships) is causing the defensive cascade. 

I personally would prefer to see a system that avoided the resultant micro which would have the same effect as the end of the micro: Half your offense is left powered down and on automatic pilot for a home system because you needed to bring your defenses back online.  Thus that whole control mechanic theory.  Now, I agree, the complexity of it is probably a bit much.  I'd hate to introduce another black box mechanic like firepower, but it's a brainstorm only. 

To improve the energy system, I would consider the following:

a) Remove the m+c cost of reactors completely and cap the number of reactors in a system.
b) Re-design Reactors to work like Command Stations.  There are multiple types of Reactors, each with pros and cons.
c) Powering down ships no longer reduces energy consumption.  It still has all the other effects (such as dropping shields and holding fire).
d) Re-balance energy costs, in particular for turrets.  It may be ideal to pull energy costs from turrets entirely, or maybe to just 10% of their current values.  Push the offensive unit energy costs up to compensate, making low energy ship types an interesting strategic choice.

I personally like options A and D as a starting point.  I'm not particularly fond of C simply because of golems and their need to repair.  It's nice when you can 'swap them out' while one repairs and another heads back out with a rebuilt fleet. 

Ideally they should have a restriction on being placed too close to each other, with the goal being 3 could just barely fit in a triangle under a Mark I Force Field, but that really just makes defending them more interesting.

The problem with that is because of the spacial concerns, nothing else would fit under the FF.  You'd basically just always end up with a multi-FF system on any planet that had to defend its reactors.

Basic Reactor - Reasonable health, no armor.  Generates 60k energy.  This is the balanced middle-ground reactor.
Armored Reactor - High health and armor to survive repeated attacks.  Generates 48k energy.
High Energy Reactor - Very low health, no armor.  Generates 72k energy.
Stealth Reactor - Low health and cloaking.  Generates 40k energy.
Home Reactor - High health, no armor.  Generates 72k energy.  Cannot be built, player starts with three in their home system.

Hrm, that's definately interesting.  It doesn't necessarily break the energy mechanic off in general though.  Here's one of the things that bugs me: If I'm building a Military station, why is it soaking energy off the reactor?  Doesn't it HAVE one?  What about the ships?  Don't they have engines?

Now, in theory we could simply call Energy something else, Manpower or Communications, and it'd be the equivalent effect, but these should, to me, all start at the command station, not attachments to them.  Hrm, I'll have to ponder that a bit.

EDIT: I should mention one of the reasons I like the distance decay mechanics is because it brings raids back into the game not because you're forced to, but as an option to mass-walking your fleet to a target and taking supply points along the way.  I'd really like to see a way to bring some guerilla tactics back into the game.  Right now the fleet-ball of doom with minor-micro is optimal to a huge degree.

[Wanderer]

For the same reason that having the AI trash your systems can lead to a defensive cascade failure. If the AI hits your supply then your going to quickly run out of supply for shields and offensive ships. Plus having m+c upkeep would encourage people to put ships on low power all the time when not using them and would be a headache if you want to speed build something.

Anything I'm going to come up with is going to try to REDUCE micro, not increase it.  I don't want Civ in my AIWar.

[Wanderer]

This talk is pretty fascinating, but I have to admit that it would make my head swim in-game. I am glad that the OP started with the question, "what is the function of the game mechanic of energy requirements? How can we make it more trouble, yet more fun?"

Rather than, "the energy system in this science fiction game is unrealistic! Here's how to make it more realistic according to me and me alone, and dramatically reduce the amount of fun in the game!"

Yeah.  I can ignore the component (or mentally rename it, like I have with the concept of humans using AIs again) but what I'd really like to do is tear the component down to its roots and see if there's a better way to implement the different game limitations it currently imposes.  The only way to do that is to describe the affects it has, rather than what it is.

That and if I had all the good ideas I'd have made the game in the first place. 

The Startopia model that you've described (I've not played it) sounds like it would be easier to implement as more of a pylon kind of thing in each system.  I'm not entirely sure that's a lot different than the turret control system that was discussed way back and went down in a flaming pile of wreckage.