OR  Zero-K Name:    Password:   

Musings about energy balance

30 posts, 836 views
Post comment
Filter:    Player:  
Page of 2 (30 records)

11 months ago
Here are two questions about energy.
  • Is spamming 500+ windgens in the lategame bad for performance, and how could we discourage such spam?
  • Should Singu be more vulnerable?

Wind generations outperforms Singu at around 1.95 average income. Some maps have large plateaus in the back where this average wind is easily exceeded, which makes windgen spam the best approach lategame. Discouraging this would involve lowering the maximum average wind to 1.8 or lower. This could be done by keeping the minimum win below a set value. A more radical idea is to make the maximum wind generation something like 0.5 - 3, in other words, to make height confer a small constant offset instead of a dramatic increase of the minimum.

Another question is whether Singu should be more scoutable and killable. When the economy is too safe players lose a vital way to interact with each other. The most difficult to scout or kill type of Singu is one underwater in a hole. Making Singu unable to be built underwater would make it more able to be interacted with and, as a nice side effect, Fusion would gain a unique strength.

With both these ideas it is important to keep in mind that Singu is not quite as powerful as its efficiency makes it look. If you have 4000 metal to spend on energy then it can be correct to make Fusions or Windgens instead of a Singu if the rate at which you can spend the 4000 metal is low enough. For example, it takes 5 minutes and 23 seconds before the energy produced by the Singu built at 20 BP exceeded that of four Fusions built at the same rate.
+8 / -0
I like the Singu idea.

my 2c, the mechanic of windgens gaining more income with height has always felt opaque - varies per map, and the effect of a given unit of elevation varies per map.

A simpler mechanic where eg. 4 height stratas of the map each confer a scaling fixed bonus (and this is visible when placing the windgen) would be more controllable.
+0 / -0
11 months ago
What was the reason windgens were bad for performance? Was it the animation? Could this be fixed in other ways (ex: disabling animation based on their number, changing from an animation to some graphical effect, etc.)?

Random idea: what about increasing digging terraform cost? I like terraforming, but sometimes it feels like it prevents interactions (ex: buried outlaws, caretakers, underwater singu, etc.). On a related note, is there any visual clue on how much terraform will cost?
+0 / -0

11 months ago
The wind equation is as follows.
groundMin, groundMax = max(0, Spring.GetGroundExtremes())
groundDiff = groundMax - groundMin
windSlope = 1/(1 + e^(4 - groundDiff/105))
windGenminIncome = 2.5 * clamp(windSlope *(windGenHeight - groundMin)/groundDiff , 0, 1)
As far as I know it is unchanged since 2007. The most esoteric bit is windSlope. Without the windSlope factor the peak of every map would produce a constant 2.5 energy income. Flater maps have a lower groundDiff so a lower windSlope, which prevents wind being absurdly good on the little bumps of Red Comet.

Wind generation is already shown in the placement tooltip. An overlay could be possible.

Wind is probably bad for performance because every unit has some overhead.
+2 / -0

11 months ago
the windgens in dune2 where just like tents, which catch the wind. just a suggestion of an non-animated model. current wind harvest could be indicated by colour.

do not touch terrorform. read http://zero-k.info/Forum/Thread/23681 and see how often terrorform is mentioned as an unique feature of zk.
+1 / -0
Wind and even solar is often superior to singus, especially in games where survivability matters (FFA, stretched out team games). I think it would be interesting to offer some higher risk-reward options for these games.

For example singus could boost nearby power production, making singu clusters even more effective. This automatically comes at the cost of having everything near the first singu be at an increased risk of being killed by the blast.

As singu clusters can be equivalent to super weapons in their strength, especially in FFA games, it might be worth experimenting with their visibility as well.
+0 / -0

11 months ago
do not touch terrorform. read http://zero-k.info/Forum/Thread/23681 and see how often terrorform is mentioned as an unique feature of zk.

As a counterpoint, I hate terraform.
+0 / -0
truth be told new ideas for energy inside zero-k are needed. there should be more balance in the game.right working on strider warships. like leviathan, ship with legs reptile resembling iguana, and strider frog. drawings on progressyes thinking the ship with legs be a frog like strider that's a ship but can leave water and pull out legs also.

two variants i'm designing but besides that energy is needed. anyone thought of adding new way of energy ? come on you can create new math adding a dam in a river and making terraform. you put a machine structure to collect the flow of water and u generate electrity and make math to create new forms of energy. thus new forms will help you equate energy.

really there is no need to balance, it's the same thing u just repeat and repeat equations balance it but people get bored. you need something new adding new energies will give you better balance.
+2 / -0

11 months ago
I've always disliked waterhole singus. This will be yet another thing that significantly changes FFA.
+2 / -0

11 months ago
on a sidenote:

1) double energy production.

2)make every unit cost energy upkeep (or else they are slowed or something?)

this would give energy a totally new spin. ofcourse this would totally impact the whole game but i think this is an interesting approach.
+1 / -0

11 months ago
terraform is unique and interesting, it´s just that it somehow has no application that is fun for both sides.
+0 / -0
* Is spamming wind laggy? You need 90 windgens to match a single singu... I have been in games where we literally teamkilled our own winds to fight the lag! The only reason I don't spam winds every game is because of the lag...

I like the idea of making the wind calculation less dramatic than it currently is, regardless of any other steps taken, by the way...

A few ideas to reduce wind numbers:

1. Make wind less desirable by nerfing in some way (More cost, less generation...)

2. Make winds a higher weight unit, to reduce their numbers. Double cost, Double Footprint.
a. Or even just a larger footprint. If they were the size of solars (Or had a "no build" but path-able area like factories)

3. What if winds degraded the power output of nearby winds? This allows them to continue their current early game role and balance, as well as "plop next to this mex" but makes large farms impractical and rewards taking empty land.
a. Have a wind check within some range; output_power = power/(1 + nearby_winds * .1)
b. Have distance between winds calculated smoothly, closer winds count much more against you than further away winds.

4. What if there were better options between wind and singu?
Wind 14m/e - 28m/e(averaged)
Solar 35m/e
Fusion 28m/e
Singu 18m/e

Wind... We have this weird situation where the cheapest unit is also the best unit. It's never worth it to make anything else until you are out of space.

Solar has a huge footprint, making them difficult to place (for their weight) and also all around suck for E production. (Cost -> 120, E -> 4??)

Fusion is in no way superior to wind unless you are out of space. Wind also gets built incrementally, making fusions never pay off VS winds - EVER. They should give some benefit for waiting to have it built, like singu does, by being more efficient than winds or solars. (E -> 42??)

Should singu be more vulnerable? Yes, but I would like to have other viable energy production options as well. One BB shouldn't trash my entire eco!
I am a fan of disallowing underwater singu's.
+0 / -0
another option is to make adv wind-gens that are effectively worth 40 wind-gens and yet only cost 20 to cover the gradual eco income you loose making expensive things.. that way small wind-gens are cheap connectors and fast start eco. but wind as a power option can be consolidated late game

the large wind-gen should have the same eco ratio of a normal wind-gen if you factor in the faster build-speed of the standard wind-gen. but this would reduce units numbers

could still be a small unit maybe 4 times bigger.. and still weak hit-points.. this way wind would not be nerfed over solars early start

.. wind needs to be better then solar for metal cost vs minimum power to keep base functioning at start to make interesting start build options and rush tactics.

like a few wind-gen and storage at start then troll com get some metal and spend the E your wind-gen saved up

is the fastest way to get to the middle of the map for a rush stinger
+2 / -0
Adv windgens are an elegant pinpoint solution for the unit-spam issue if a solution for only that is needed.

40 for the price of 20 sounds a bit extreme though.
+2 / -0
its difficult to say iv run a few numbers but yeah lots is involved under that hood.

gradual feedback eco when you have no metal means that smaller is the better option mostly.. ie its better to get something and pay off the eco in installments that help pay it off faster.. so to make a heavy eco option viable you must make it produce more energy.. in this case make it cheaper so that your not stalling while you invest.

imagine 30 cost vs 40 equivalent

why would i make that when im stalling metal .. if its a 50% value gain it will turn my head like it should

never forget a good player will always have no metal
+0 / -0
DErankAdminmojjj's thought reminded me that a while back I was thinking about a supply line mechanic where units only had a certain amount of ammunition and would need to return to cons or special arming units to retool (which would of course be vulnerable targets to intercept), but you could achieve a similar outcome in a much less fiddly way by having units costing energy to maintain.

Never mind slowing units down - have them stop firing if you run out of juice. Maybe armoured mode can also cost energy for those units which have it (likewise ALL cloaks).

A lot of balancing work would need to be done on the right e-cost for any given unit but it would certainly shake the game up in a big way. Energy buildings would need to generate more energy though. To keep the numbers as intuitive as possible, I'd suggest whole integers for the e cost (e.g. if a glaive costs 1e I know that if I have 10e, I can sustain 10 glaives). What to do with fleas and darts would be interesting though. On the other hand, e-need of artillery and berthas could be particularly high to offset the fact that they otherwise have the ability to pummel at will. Constant cost rather than per shot cost means all the time they're not shooting they're burning a lot of energy to no purpose. This might change behaviours.

My thoughts on the consequences of doing this:
The early game would be slower burning which might tilt the game too much towards porc creep vs raiders, but that relationship can be tailored by increased energy consumption from porc. The slower start might be less intimidating to newbies as well (I know I've had them rage when I've been raiding their base within a few minutes of game start).

Rushes might be discouraged as there might not be e to sustain the big boy (or if you do build the energy in advance, you increase the chance of being spotted).

Hitting eco would now directly, rather than indirectly affect your opponent's front line capabilities. If you destroy enough energy, their front line forces might be shut down entirely. Early game raiding might be a bit more awkward - but late game raiding could have a much bigger impact. This would also incentivise the use of defensible point energy sources like fusions and singus vs space inefficient wind (especially) and solar.

The mechanic could be developed such that the e cost is only taken when units are fighting (not on a per shot basis, but on whether they are engaged or not). If this were done, depending on the numbers, one might not be able to freely throw units at the enemy and have them fight at full effectiveness. If you have more troops than your economy can sustain, then you can't use them. On the other hand, storage could be weaponised as a strategic energy reserve. At a critical moment, you could use your stored energy to push reserve forces into the fray, powered for a while by stored energy to turn the tide. I think this could be a very interesting change to the flow of the game.
+0 / -0

11 months ago
There seems to be general agreement about making Signu non-amphibious.

On wind, I like how it is cheap and fragile. I don't want to add another energy structure or to make Wind more complicated. If anything, this could be an opportunity to make the wind calculation simpler. I like how high positions on the map are good spots for wind as they spice up energy placement. Basically I like a lot about how wind behaves at the moment. This means that wind would need a big nerf at high altitudes on some maps. Is such a nerf a good idea?

For simplicity, what if the best wind income was 1.0 - 2.5? The tallest hills on maps with particularly tall cliffs tend to have an altitude of around 500 to 700 elmos. There would be a few simple ways to unify the minimum wind increase.
  • Tie generation directly to altitude. Perhaps have 0 at sea level and 1 at 600 elmos. The simple option.
  • Take the minimum and maximum heights. If maxAlt - minAlt < 600 then set zero wind at minAlt. If maxAlt - minAlt >= 600 then set full wind at maxAlt. Vary the wind by 0.1 energy per 60 elmos.
  • Instead of taking minAlt, take the minimum mex altitude. Increase wind at 0.1 per 60 elmos from there. The PTrankraaar option.
+2 / -0

11 months ago
Oh I thought the idea was to buff singu to make it better than wind farms :|
+1 / -0

11 months ago
what about the idea of having static windgen-models?
+1 / -0

11 months ago
I would want to see data about the performance impact of rotating pieces. A static windmill does not sound worth the kind of performance improvement that I expect.
+2 / -0
Page of 2 (30 records)