Factorio Forums

Well, I can not say I disagree with you on all those points, but I don't agree with you on them either.

As I said, I don't think prod 1/2 and speed 1/2 are balanced. They are just way worse than their T3 counterparts. They might be a lot cheaper to build too, but they feel just not worthwhile.

474% Productivity is a much harder extreme than speed modules. Speed modules only give you more throughput for more energy. You could just as well build more assemblers - which would be much more energy efficient. If you then use efficiency modules you are way way way more efficient than using speed modules alone.
Productivity on the other hand gives you FREE items. So even if the throughput of them isn't as great as speed (prod/speed beats pure speed only at 11-13 beacons affecting a single machine in throughput!) it is so powerful because you get more out of less.

474% Productivity means you get nearly 5 times the stuff out of the same ingredients. THAT is the true power of productivity, not the throughput. The throughput when combined with speed modules is only the cherry on top, making productivity energy efficient and time efficient.

Efficiency is good enough for just the energy savings. Pasting solar panels is annoying. It would be a lot better if there was another mid-game energy source like nuclear and solar panels were high tech (=expensive!) end game content. And you couldn't just spam them. Pollution is kind of a weak mechanic at the moment anyway.

I know many that feel like Efficiency 1 is really overpowered (you get easy 500% energy efficiency on nearly everything). Problem is that Tier 3 is so bad because unlike Prod/Speed 3 it scales linearly and because of the 80% hardcap.

Some tests I did to satisfy my curiosity:
http://imgur.com/a/zAx71

Efficiency modules would mostly make the no module version 5x more energy efficient (test would've taken 5x longer too xD).

Also I had a really evil idea to make efficiency modules stay relevant till the end game: Infinite Evolution

Obviously would need to be paired with infinite research for all weapons/defenses. The evolution factor would also need to affect the stats of biters, rather than just the types of biters. Maybe increase their HP based on evo factor. Imagine what a 50000 HP behemoth biter would be like?

I would make it so efficiency modules would effect POLLUTION as a primary goal.
Currently the main downside from using heavily beaconed/moduled setups is the insane pollution it produces while power is not that of an issue really.
If they had these numbers effecting pollution it would actually be useful as it would be multiplicative with power consumption.
I don't care if an assembler consumes 800% power it it is actually power-efficient per item produced. But if you could replace say 1 of four beacons with say -80% pollution I would certainly go for it.
This would also open up a scenario for heavily moduled mining outposts as currently this would make the outpost a tasty bait for the biters.

Or as an option energy efficiency and consumption % could be separate values being multiplicative. This would also be an interesting stuff to check out.

I will also drop those here as related:

viewtopic.php?f=16&t=21052

viewtopic.php?f=16&t=33863

Efficiency modules reduce energy consumption AND pollution, by the same amount.

Yoyobuae wrote:Efficiency modules reduce energy consumption AND pollution, by the same amount.

They do reduce pollution as a side effect now as that one is now directly bound to energy consumption. Thats why its bonus is additive to Speed module energy consumption and has near to no effect in beacon heavy setups.
If they would have a separate pollution decrease bonus it would be multiplicative with energy consumption and could really decrease pollution by 80% no matter how many speed modules are currently effecting the building.

Haven't read the whole conversation yet, but I wanted to point out a rather nice use of T3 Efficiency modules: In assem3s you can run 3xEff3 and 1xSp3, and get 150% speed and +70%-150%=20% energy cost, which lowers energy cost per item even further than w/ 2x T2 Eff, to 13.(3)% of base and you get 1.5x rate per assembler while doing it. Nice for low energy cost high product volume assembly setups (also a beacon-free setup so again, energy is saved). So if you want to truly minimize energy expenditures for given production volume, T3 Eff have a use.

Frightning wrote:Haven't read the whole conversation yet, but I wanted to point out a rather nice use of T3 Efficiency modules: In assem3s you can run 3xEff3 and 1xSp3, and get 150% speed and +70%-150%=20% energy cost, which lowers energy cost per item even further than w/ 2x T2 Eff, to 13.(3)% of base and you get 1.5x rate per assembler while doing it. Nice for low energy cost high product volume assembly setups (also a beacon-free setup so again, energy is saved). So if you want to truly minimize energy expenditures for given production volume, T3 Eff have a use.

I did make use of that setup when I wrote this thread. The reason I didn't mention it is because it's blatantly obvious why it's not actually effective. If you build 2 assemblers with this setup you could just build 3 assemblers with 3 efficiency 1 modules, which is much much cheaper. Sure, the energy efficiency isn't quite as good, but it's very close and the price you pay for 4 T3 modules is just not comparable to what 3 T1 modules and an extra assembler cost.

If you're wondering why the "build more assemblers instead" argument doesn't qualify for the competition to efficiency modules (speed/productivity) it's because productivity creates free items. Having equal production from normal assemblers and ones using productivity means using less raw materials in the latter - that's why an argument can be made for them.

If you're swimming in T3 modules and want to save space or just want to produce something (e.g. T3 modules) then of course the setup is fine to use. But generally speaking it's not really a competitive alternative, which is what I want T3 efficiency to be.

Syrchalis wrote:

Frightning wrote:Haven't read the whole conversation yet, but I wanted to point out a rather nice use of T3 Efficiency modules: In assem3s you can run 3xEff3 and 1xSp3, and get 150% speed and +70%-150%=20% energy cost, which lowers energy cost per item even further than w/ 2x T2 Eff, to 13.(3)% of base and you get 1.5x rate per assembler while doing it. Nice for low energy cost high product volume assembly setups (also a beacon-free setup so again, energy is saved). So if you want to truly minimize energy expenditures for given production volume, T3 Eff have a use.

I did make use of that setup when I wrote this thread. The reason I didn't mention it is because it's blatantly obvious why it's not actually effective. If you build 2 assemblers with this setup you could just build 3 assemblers with 3 efficiency 1 modules, which is much much cheaper. Sure, the energy efficiency isn't quite as good, but it's very close and the price you pay for 4 T3 modules is just not comparable to what 3 T1 modules and an extra assembler cost.

If you're wondering why the "build more assemblers instead" argument doesn't qualify for the competition to efficiency modules (speed/productivity) it's because productivity creates free items. Having equal production from normal assemblers and ones using productivity means using less raw materials in the latter - that's why an argument can be made for them.

If you're swimming in T3 modules and want to save space or just want to produce something (e.g. T3 modules) then of course the setup is fine to use. But generally speaking it's not really a competitive alternative, which is what I want T3 efficiency to be.

It's cheaper to build T1 Eff and more assemblers, but the items produced will still cost 20% energy instead of 13.(3)% energy. (Not to mention pollution per item produced is likewise affected by the energy use). So while you're right that the fixed costs are much higher, there is a meaningful savings to be had (if energy and pollution conservation are important to you at least). Productivity is actually only really valuable when a resource is scarce, though it's also handy for making lots/expensive high-end products because you can exploit productivity multiplier stacking to get an impressive increase in end products (and synergy with speed beacons amplifies the efficiency of doing this). When resources are actually plentiful, it's far cheaper, in terms of fixed costs, to just build more lower level infrastructure and mine more raw resources (not to mention the savings from not having to make the modules and supporting infrastructure for energy costs).

The biggest problem I see with modules is that energy consumption is linked with pollution and there is no antipollution module.

Nasabot wrote:The biggest problem I see with modules is that energy consumption is linked with pollution and there is no antipollution module.

Efficiency modules basically are the anti-pollution modules. Because lowered energy consumption=less pollution.

Eff1 modules are super effective, making them is pretty much a no brainer. I would even say they're overpowered (compare cost, effect and time investment in efficiency module vs equivalent solar farm), but late game strength of prod/speed combo and limited module slots reduces that, making them extremely good only for limited time. Overall - pretty balanced.

Eff2 usage outlined in the OP and other posts seems like a joke to me - that much extra cost for additional 20% of energy & pollution shaved off some special buildings? Which will eventually be either replaced, or their modules will be? Ehhh . . .
Eff3 are same as eff2 exept being even more expensive . . .

Considering devs push the logistic / scaling aspects of the game instead of fighting ones (and let's be clear - community wants it that way), there is fundamentally no place for current mold of eff2/3 modules in the game. Eff1 makes it in only because it matches/outperforms the solar panel on cost & benefits & time investment. Only mentioned way to make them really useful would be in the "beacon efficiency boost", but that would cause another balancing mess. I'm for trying that, but I doubt devs have time for it.

Tev wrote:Eff1 modules are super effective, making them is pretty much a no brainer. I would even say they're overpowered (compare cost, effect and time investment in efficiency module vs equivalent solar farm), but late game strength of prod/speed combo and limited module slots reduces that, making them extremely good only for limited time. Overall - pretty balanced.

Eff2 usage outlined in the OP and other posts seems like a joke to me - that much extra cost for additional 20% of energy & pollution shaved off some special buildings? Which will eventually be either replaced, or their modules will be? Ehhh . . .
Eff3 are same as eff2 exept being even more expensive . . .

Considering devs push the logistic / scaling aspects of the game instead of fighting ones (and let's be clear - community wants it that way), there is fundamentally no place for current mold of eff2/3 modules in the game. Eff1 makes it in only because it matches/outperforms the solar panel on cost & benefits & time investment. Only mentioned way to make them really useful would be in the "beacon efficiency boost", but that would cause another balancing mess. I'm for trying that, but I doubt devs have time for it.

Beware of mis-interpreting the numbers here. Upgrading from Eff1 to Eff2 cuts energy usage in the relevant machines in half (from 40% to 20%), and likewise affects pollution the same way. That's a very significant impact. As for Eff3, the best use I know of by far is in Assembly machine 3s where you can run 3xEff3 with 1xSpd3 and get 150% speed at 20% energy cost (so each item made costs just 13.(3)% as much as without modules and likewise for pollution generated). This is a further reduction in energy usage for given amount of goods produced, compared to 2xEff2, of 33.(3)%.

Beware of mis-interpreting the numbers here. Upgrading from Eff1 to Eff2 cuts energy usage in the relevant machines in half (from 40% to 20%), and likewise affects pollution the same way. That's a very significant impact. As for Eff3, the best use I know of by far is in Assembly machine 3s where you can run 3xEff3 with 1xSpd3 and get 150% speed at 20% energy cost (so each item made costs just 13.(3)% as much as without modules and likewise for pollution generated). This is a further reduction in energy usage for given amount of goods produced, compared to 2xEff2, of 33.(3)%.

(emphasis mine)
Cutting energy in half on those levels is so low energy gain that when you compare it to the cost of that gain it just doesn't seem even close to any electricity generation cost in the game.

The ratio of direct cost/energy saved might be theoretically better for Eff3 modules in combinations with prod/speed but at that point you're also losing resources you'd gain from more prod modules. (or saved resources for prod modules by using more speed modules) And generally don't care about energy anymore that much. Eff1 still makes sense to use where prod/speed doesn't make sense (or is just plain too expensive), but once again - only because it compares well to solar panels.

Don't be too obsessed with relative gains, they're meaningless when the absolute value is too low to matter.



EDIT: As for numbers:

Price of Eff1 is comparable to solar panel. Based on your preferred weighting it is between 0.8-1.5x the cost.
Solar panel gives at max 60kW.
Eff1 saves 45kW in Assembler2, 61kW in Assembler3, 27kW in Mining Drill. Which means that, based on your factory/mining utilization, it saves pretty much same amount of energy that would solar panel generate.

(notice I am not even trying to count the cost of Accumulators, as that would complicate calculations based on your backup-power options and very likely not improve the relative strength of solar panels)

Considering solar farms are big, pain in the ass to build, and efficiency modules can now be inserted via robots, it's clear Eff1 is at least comparable and based on your preferences it might be many times better.

Now Eff2 - More than 5 times the cost of Eff1. Energy savings are just 33% better. Easy math.
Eff3 - More than 5 times the Eff2 cost. Hahaha.
Even with usage in setups that could offer order of magnitude bigger energy savings than std Eff1 in std situations, it is hard to see the setup where cost of Eff2/Eff3 + opportunity cost compares favorably to those of more solar farms (or just Eff1s). If you are in a death world and desperately want to reduce pollution and at the same time don't have much (rather zero) space then MAYBE Eff2 can be cost-effective, but additional tweaking of setups for that doesn't seem to be worth the trouble for me.

Eff3 is just pure garbage.



EDIT2: formatting

^ You're missing the point, it's more about being able to make stuff for very little energy, and, more importantly, correspondingly low pollution. Pollution is hugely important on death worlds for instance. Yes, you pay for it considerably in terms of fixed costs, but the same is true for the speed and productivity gains of the higher tier spd and prod modules as well. Not to mention that prod+speed beacons is incredibly energy intensive and thus extremely pollutive (which is fine if you have the infrastructure to handle the heavy duty defense you will need or the biters are set to passive, but if not, then it's quite prohibitive to even attempt). Also, Efficiency module based setups make more sense for smaller scale setups where you don't need extremely high throughput (especially since beacon based layouts don't become terribly efficient unless they are quite sizable due to perimeter builds getting/providing less than interior buildings in otherwise efficient layouts).

Frightning wrote:^ You're missing the point, it's more about being able to make stuff for very little energy, and, more importantly, correspondingly low pollution. Pollution is hugely important on death worlds for instance. Yes, you pay for it considerably in terms of fixed costs, but the same is true for the speed and productivity gains of the higher tier spd and prod modules as well. Not to mention that prod+speed beacons is incredibly energy intensive and thus extremely pollutive (which is fine if you have the infrastructure to handle the heavy duty defense you will need or the biters are set to passive, but if not, then it's quite prohibitive to even attempt). Also, Efficiency module based setups make more sense for smaller scale setups where you don't need extremely high throughput (especially since beacon based layouts don't become terribly efficient unless they are quite sizable due to perimeter builds getting/providing less than interior buildings in otherwise efficient layouts).

Death worlds have actually bigger problems with resources than ordinary maps (or wtf are your settings for "death worlds"), so the resource efficiency is even more important. As you won't be probably using prod/speed modules there for a long time, eff2/3 just stop maing any sense altogether, as eff1 crushes them. Also you waste space by extra infrastructure. And, when you have blue circuits defense is not such a big issues anymore, so minimizing pollution at all costs for all buildings is not really important, while resources / difficulty of expanding still remains big problem.
Btw last time I played death world I pulled prod3/speed3 combos without much trouble (I abandoned the base at around 2,5GW energy consumption due to UPS issues).

How do eff2/3 modules matter for "small scale setups with low throughput" is also not clear. Once again - why not just use Eff1?

Make some clear points please. Like how many resources you're wiling to spend just to consume 30kW less per assembler. How much pollution do you save. Will it be better than just spending resources on expansion . . . ? When you can't quantify your preferences, and just state "it is better" you won't really convince anyone about anything. And since majority of players view Eff2/3 as worthless the burden on proof is even more on you.

Tev wrote:

Frightning wrote:^ You're missing the point, it's more about being able to make stuff for very little energy, and, more importantly, correspondingly low pollution. Pollution is hugely important on death worlds for instance. Yes, you pay for it considerably in terms of fixed costs, but the same is true for the speed and productivity gains of the higher tier spd and prod modules as well. Not to mention that prod+speed beacons is incredibly energy intensive and thus extremely pollutive (which is fine if you have the infrastructure to handle the heavy duty defense you will need or the biters are set to passive, but if not, then it's quite prohibitive to even attempt). Also, Efficiency module based setups make more sense for smaller scale setups where you don't need extremely high throughput (especially since beacon based layouts don't become terribly efficient unless they are quite sizable due to perimeter builds getting/providing less than interior buildings in otherwise efficient layouts).

Death worlds have actually bigger problems with resources than ordinary maps (or wtf are your settings for "death worlds"), so the resource efficiency is even more important. As you won't be probably using prod/speed modules there for a long time, eff2/3 just stop maing any sense altogether, as eff1 crushes them. Also you waste space by extra infrastructure. And, when you have blue circuits defense is not such a big issues anymore, so minimizing pollution at all costs for all buildings is not really important, while resources / difficulty of expanding still remains big problem.
Btw last time I played death world I pulled prod3/speed3 combos without much trouble (I abandoned the base at around 2,5GW energy consumption due to UPS issues).

How do eff2/3 modules matter for "small scale setups with low throughput" is also not clear. Once again - why not just use Eff1?

Make some clear points please. Like how many resources you're wiling to spend just to consume 30kW less per assembler. How much pollution do you save. Will it be better than just spending resources on expansion . . . ? When you can't quantify your preferences, and just state "it is better" you won't really convince anyone about anything. And since majority of players view Eff2/3 as worthless the burden on proof is even more on you.

Ok let's quantify some things then:
You said yourself: space is at a premium on deathworlds (because expansion is hard with a sea of red in the way).
So with that in mind, upgrading from 2xEff1 to 2xEff2 in things like Electric Furnaces, Oil Refineries and Chemical Plants will cut the amount of space you need for the Solar panels and Accumulators to power them in half, and also reduce pollution (and hence reduce pressure on your defenses due to attacks, which if Laser turrets are a component of your defenses saves more power, and hence more space).
Likewise, running 3xEff3+1xSpd3 in Assem3s yet further decreases how much space your power grid needs. Yes, the fixed costs of doing this are non-trivial, that is the price of higher efficiency. Similar results hold for Prod and Spd module benefits, the increases from using higher tiers are far smaller than the additional fixed costs associated with them. For instance, it's cheaper to just build more assemblers than it is to increase production rate by putting speed modules in them.

Which module types make sense to use are a function of map settings:
Productivity modules are the logical choice for scarce resources (e.g. I have used them for Iron and Oil based items in my Random World map because those resources have been quite scarce and hence I've needed to stretch my supply of them as much as I can)
Efficiency modules are logical when energy is difficult to produce in quantity (this can happen when you have a lack of Coal and/or Oil and lack of space for solar). It's also the only option with no real downsides apart from fixed costs (spd and prod have negative effects as well).
Speed is mainly useful for keeping production areas compact, and ideally in beacon setups to counter prod speed penalty.

Of the 3 module types, Speed is actually the hardest to justify in isolation (without other modules/interactions), as the speed bonus costs more than it gives in energy, and the modules are more expensive than just building another assembler. Productivity comes with heavy downsides: increased energy costs and even more pollution due to +% pollution modifier, and lowered speed while doing it (which is another multiplier to energy costs and pollution generated per item made), but the upside is your get more output for the same input materials, which stretches available supply of resources to give more items than it otherwise would. Which, of course, is only useful if that resource is actually scarce, otherwise, why bother? Efficiency by comparison is almost a no-brainer: no downside apart from fixed costs for much reduced energy demand? Why not?

So why do so many people use Prod+Spd setups in megabases? That is mostly due to two things:

Productivity multiplier stacking: High tier items have many steps from raw to final, and most of those are intermediate products, and hence the productivity bonuses can be stacked many times for a massive gain in finished products from initial materials

Speed-productivity module interaction: Speed penalty of prod modules is off-set additively by speed modules, hence energy per item made is actually lowered by using speed beacons in an efficient layout compared to no speed modules.

These two facts together, with abundant resources and space that is typical of megabase favoring map settings means that, if you plan to build a megabase. There's no real reason not to go with prod+spd since the large energy demands are a non-issue and, in fact, provide another thing for your massive factory to make for you (power infrastructure buildings).

In the OP, the idea of adding total energy consumption modifier is one I suggested before.

• Eff1: -30% energy consumption
• Eff2: -30% energy consumption, -10% total (multiplicative) energy consumption
• Eff3: -30% energy consumption, -20% total energy consumption

The idea is the TEC modifier is applied first, so on a machine with a consumption of 1.0, the eff2/eff3 work exactly as they do now: (i.e. 1.0 * 0.9 - 0.3 = 0.6: a -40% modifier). On a machine with +500% consumption, the change would be: 6.0 * 0.8 - 0.3 = 4.5 : equal to a -150% consumption modifier.

Another possible scheme could look like this:

• Eff1: -30% energy consumption
• Eff2: -40% energy consumption, -40% module energy consumption
• Eff3: -40% energy consumption, -80% module energy consumption

The module energy consumption reduces the energy consumed by other modules, but cannot reduce base consumption, so an eff2 module by itself only provides -40% energy consumption. The advantage of this expression, is that it makes it clearer what the primary purpose of the higher tech eff modules is: for combining with other modules and it's also easier to reason about because the "-80%" is in direct proportion to the energy consumption of machines, i.e. in an Electric Furnace, -80% is equivalent to 180kW*0.8 = 144kW.

Frightning wrote:Pollution is hugely important on death worlds for instance.

This was true of 0.14 "death worlds", but isn't for 0.15 Deathworld. The major evolution factor on Deathworld (and Dangerous) is time-based, once past the first hour (i.e. once you have some solid defenses) it hardly matters how much pollution you make, or how much you go around stomping nests because time is driving evolution so quickly anyway. Sure, making more pollution will attract more attacks, but you can knock the nests back to make the attacks stop. I actually find the balance pretty good. However the fact that pollution doesn't really matter, but the techs cost 4x as much makes the use of productivity modules insanely favorable under Deathworld, there is heaps of time for them to pay for themselves. Also, while expansion isn't that hard, it is a bit tedious and resources are always running out (assuming unmodified resource settings) so stretching resources is very helpful.

Productivity modules are already more of a no-brainer in 0.15 because the higher tier science packs have such high "resources per second" that the payback time on prod3 is very quick, but on Deathworld/Marathon they are 4x the no-brainer.

Anyway, I don't disagree wrt to 0.14 deathworlds, but the assumptions of what a "Deathworld" entails are going to change now, and a new word will be needed for "normal settings, max biters" (if that's something people continue to play despite the Dangerous/Deathworld presets) and the devs seem to be inclining towards making pollution less important for these game styles.

IMO the whole "additive multiplier" thing is wrong. I've done quite a bit of modelling (mostly epidemiology & bio-stats), and modifiers are almost always multiplicative.

E.g. speed 3 modules have a 70% increase in energy consumption, productivity 3 modules have a 80% increase, so 2 prod + 1 speed should result in 180% * 180% * 170% = 550% of base consumption (450% increase), not 330% of base (80+80+70=230% increase).

End result: if speed modules reduce efficiency (energy cost per item output), then they always reduce efficiency the same way.

Of course, this would make your 2 productivity + 8 speed beacon setup increase power usage 226 fold (22500% increase) with approximately 26 times the output-per-second, thus increasing power-per-item 8.5-fold (750% increase)