Module Balance (Mostly Efficiency)

[Syrchalis]

In my opinion (I did a lot of math on it) the modules are pretty balanced, with some exceptions. Let's go through all nine of them:

Productivity Modules:

Rather well balanced. Lower tier ones return their investment quickly, but give little benefit, while tier 3 ones allow for some incredible resource multiplication if used in multiple steps of a production chain. They are mostly balanced due to the interaction with speed beacons and that this combo needs A TON of power.

My only problem here is that the lower tiers have the same speed penalty than higher tiers and of course that the jump from T2 to T3 is large (4% -> 6% -> 10%) while energy ratio improves as well to T3 (1:10 for T1/2 but 1:8 for T3).

Speed Modules:

Speed to energy ratio improves slightly with tier, offsetting the increasing cost. I think they are probably the most balanced in themselves. Using them with efficiency or productivity modules together is to be honest quite useful and fair.

Efficiency Modules

Now this is where it gets interesting.

I think Efficiency 1 modules are actually very strong. For what they cost to produce (no blue circuits) they are incredibly useful. Energy is not hard to get, but during the long game phase after getting modules and red circuits and before you are rich enough in production to do producitvity/speed beacon setups they allow you to basically ignore building any energy sources. They don't feel like they are way out of balance, but I found myself using them nearly everywhere and throw them around like candy.

Efficiency 2 has kind of a niche use - to double the effectivity in machines with two slots (all oil machines, assemblers 2). They would otherwise need 40% power/produce 40% pollution and with them it's 20% respectively. That's a relative increase of 100%. That seems fine to me since you need much more than four times the resources to make them. In absolute numbers the gain isn't large at all. Their production cost isn't as insane as T3 modules though so I can live with that. They could definitely need a buff, but I honestly can't see how to improve them. Two caps out the energy consumption at -80%, so if they were as good as efficiency 3 they wouldn't actually be any better. Only in combination with other modules.

This does give an interesting possibility though - no matter how powerful you make efficiency 2 modules their use alone will always stay as it is right now - so you could totally balance them for the use with other modules. For example, if you make them reduce energy consumption by 60% you could negate the negative effects of the other tier 2 modules with them and do setups like 2 Efficiency 2 and 2 Speed 2. I'm not saying they should get buffed this much, but it's nice we can essentially balance them for "use alone" and "use in combination" separately because of the -80% cap.

Now the biggest issue: Efficiency 3.
They are just so useless! For what they cost they offer nothing. Using them by themselves is absolutely pointless because in every machine the tier 2 variant is enough to cap out energy at -80%. One could massively buff them, so that their value in combination goes up, but this wouldn't do much in my opinion, because of the productivity/speed interaction.
Anyone who has done the math (if you haven't, look at my link at the beginning of this post, especially at the second table (the column labels are in the first picture)) - knows that productivity modules need speed beacons to bring their energy/unit and unit/time values to reasonable levels (and make the whole "free items" thing worth it).
There is just no place for these.

So what do about them? I say nerf em! Make them reduce energy consumption by 25% only! (Value up for debate) - but make them, and just them, multiplicative with other types of modules (still additive with themselves). So if a machine is taking 400% power due to other modules and you put an efficiency 3 in it, it gets reduced to 300%, put another one in it and it's down to 200%.

Please don't get stuck on the exact values for now, but grasp the two general ideas here - Efficiency 3 -> multiplicative interacting with other modules.

What do you think? What is your balance assessment of the nine modules? Any other ideas what to do about the lower tier productivity and the higher tier efficiency modules?

[greep]

Well, regarding efficiency3, that might involve making weird changes to UI/description and stuff for the special case, it seems weird to just tack on a special case to a tier for one module. One problem I see mechanics wise it it kind of makes them a no brainer for high density beacons setups (each assembler w/ 6 beacons), or worthless for "normal factories" with lower or no beacon density, essentially making them impossible to balance on a veteran/newbie scale. But I guess that's more a problem with beacons being no brainers themselves xD

Personally I'd just buff them to be like -30/-40/-50 -> -30/-80/-200

[Yoyobuae]

How about making them affect the efficiency of the beacons where they are placed.

[BenSeidel]

tldr;
Efficiency modules interact with biter behaviour and your time spent clearing them out so are pointless on anything but a death-world.

This topic has come up on multiple occasions, in one form or another and the issue that I have seen in all the arguments so far is that biters are not taken into account. All my points are based upon "late game", as the OP is talking specifically about T3 efficiency modules, where you will be using them the most. (I actually class "late game" beginning the moment you place your first t3 module in a building, but that is me).

So, lets assume "no biters".
Speed & Productivity modules have no trade-offs, only an initial cost (item cost + power infrastructure). They increase the production rate of your factory at the cost of additional power (aka solar, all built by your construction robots). Efficiency modules on the other hand are different, they don't increase the production of your factory, only reduce the amount of power you have to make. By the time you start placing productivity/speed modules in everything with beacons and have +1000% power usage, -50% power usage seems pointless. Also, the amount of resources required to build more power is far less than the resources required to build the T3 efficiency module AND the additional buildings to make up for the loss in production.

Now, lets put biters into the equation.
Speed and Productivity modules now do have an associated running cost to them: biter attacks, and a much higher initial cost: you clearing land. The pollution produced by each building is increased immensely, increasing your pollution cloud and the biter bases that are being hit, producing more biters, causing more attacks, costing your more resources. Efficiency modules, with biters, have a two fold benefit. Firstly they reduce the amount of pollution you produce, thereby reducing the number of biter attacks (and the resources spent on defending against them), and they reduce the amount of space required for your power. In maps with biters, land is expensive in your time. It takes a long time to clear out the biter bases and to build the wall.

I know that once you have a wall with 2+ rows of fully upgraded lasers there is little to no additional cost to defending your base (with solar), so that part is more about personal preferences but the clearing out of land is an extremely long process. Anything you can do to reduce the amount of land you use increases your enjoyment of biter-filled worlds.

[Syrchalis]

Well, regarding efficiency3, that might involve making weird changes to UI/description and stuff for the special case, it seems weird to just tack on a special case to a tier for one module. One problem I see mechanics wise it it kind of makes them a no brainer for high density beacons setups (each assembler w/ 6 beacons), or worthless for "normal factories" with lower or no beacon density, essentially making them impossible to balance on a veteran/newbie scale. But I guess that's more a problem with beacons being no brainers themselves xD

Personally I'd just buff them to be like -30/-40/-50 -> -30/-80/-200

I mostly agree with you. I think that my approach isn't perfect, because from a design perspective it would be better if they would act all the same way. However, what you say is a problem was actually my intended goal.

Your point is that they either become worthless in normal settings or extremely good in late game setups. My intention was to make them worthless without other modules being involved. The multiplicative interaction with other modules should be a big neon sign for all players "hey, these suck alone - use them with other types of modules together!". Why I think this is a good thing? Well, as it is, Efficiency 1 and 2 completely cover any needs in terms of pure efficiency modules. In other words - if you want to use ONLY efficiency modules then you will NEVER use efficiency 3 anyway, even in the current game.

I still agree with you though that the same could probably be achieved by straight up buffs.

For me personally it is clear that efficiency modules are in need of a rather large buff, except for efficiency 1 of course.

@BenSeidel - I get your point. Right now people are either running full efficiency builds when they are still stuck with a low processing unit production or they give no craps about energy/pollution and walled themselves off already. With efficiency buffed you could run hybrid builds that don't have the throughput of pure prod/speed (they really suffer in speed) but reduce pollution considerably.
It allows you to choose to go green or to go black. Maximum throughput and lots of pollution or medium throughput and little pollution. Right now you don't really have a choice if you want to utilize productivity modules (I don't mean the speed/prod combo, I mean you want to use the 20-40% free items from productivity). Because without any speed your throughput is even worse than without any productivity modules and you still have high energy/pollution as drawbacks, so it's just not viable.


Here are some possible numbers:

Variation A
20 / 40 / 80
Now efficiency 1 modules are not as amazing anymore instead they are more in line with other tier 1 modules in terms of usefulness. Assembly machines 3 can still use them very efficiently (hah). Efficiency 2 is the go-to module for anything with just two module slots if you don't want to put anything else in there. Efficiency 3 can be used alone to cap out energy reduction if your factory produces a lot of tier 3 modules anyway or alternatively it is now at least capable of negating the negative effects of T3 speed and productivity modules.

Variation B
30 / 80 / 120
I shamelessly stole that from you greep. With some calculations to back up the numbers.
This doesn't change anything about efficiency 1, buffs efficiency 2 so it is now closer to being cost efficient. Usually the problem with efficiency 2 is that you pay four efficiency 1 modules AND 5 more red/blue circuits for basically no gain. Relatively it's a 100% gain, and a 300-500% cost increase, which is already bad. But in absolute numbers it's a mere 20% gain for the same (absolute) cost. So them being this good isn't really a problem. You put a single on into a machine and it's energy reduction is maxed out - nice, but you payed a lot more than for efficiency 2 modules.
For efficiency 3 I wasted an hour of my life and put them into my excel sheet to find good values. 120% seems pretty good. For my calculations I used beacons with 1 speed and 1 efficiency module. I mostly focused on these two situations: A machine with 2 productivity modules and 4 beacons affecting it and a machine with 4 productivity modules and 4 beacons affecting it. I did it for more and less beacons too, but that would explode this post.

You lose considerably speed, but you more than half the energy consumption. So it's a trade off worth taking if you care about energy. If you don't then prod/speed is still better (and should be).

At very high beacon numbers (8+ per machine) you start to max out at -80% energy consumption leading to pretty overpowered values (very low energy/pollution, high speed/productivity) - HOWEVER! my calculations are relative and don't account for beacons requiring a crapton of energy. So 8+ beacons per machine means a lot of beacons, which then again means a lot of energy - so the beacon's high energy cost is a natural diminishing return against these modules, making sure you can't get too good of an energy ratio. At 120% you need enough beacons to reach -80% energy consumption that the benefit isn't worth it at all anymore.

If we would take 200% for example we would end up with -80% energy consumption at 4 beacons for 2 module machines and 6 beacons for 4 module machines. There the beacons themselves aren't drawing enough power, because you can still make rows of beacons to reach these numbers. Having 8 beacons affect a machine and having each beacon affect many machines is pretty hard (maybe impossible? I didn't test).

[greep]

Well, the reason I went with -200 was with the same logic benseidel was using, I was thinking "what would make me use these in a real game". For instance, in a no-biter world, the only values that make sense would be when the amount of energy reduction per item exceeds the cost saved in solar power/accumulators of both itself minus the cost to regain that speed per item in module/assembelr cost of the replaced speed module.

That's some pretty insane math but I'm guessing for that it'd need need like a no loss scaling -30/-350/-2000. I.e. it's not going to be balanced for peaceful/non-peaceful worlds no matter how you slice it.

However, in a real game, pollution can start exploding if you beacon smelters and mines, and this would at least make them useful there in combination with speed/production. Additionally those players that like to play "non-optimal" and just like clean factories will be able to at least be efficient while also being able to use speed and productivity. But in a real game I personally wouldn't use them if not beaconing smelters/mines in any situation.

TBH, I'd be leaning towards straight crazy looking like -30/-160/-800, if only it didn't look so silly. -800% per assembler That would actually effectively be your original suggestion @ ~70%, i.e. one full beacon mostly nullifies energy in a 6 beacon setup.

[Syrchalis]

Well, if we go for crazy numbers then you need 1 efficiency 3 module in ONE of the beacons and you got optimal energy, that wouldn't be great. I'm looking for a value that is balanced so if you use some speed and some efficiency you get a very energy-efficient factory, which also utilizes the productivity bonus without sitting at -60% speed (and rather at +50-100% or so).

I want it to be a viable option. I don't want it to compete with prod/speed in terms of items/sec, I want it to be superior in energy/item while still having okay speed, unlike it is now.

With pure speed you would get the most items per second (as it is now).
With prod/speed you get the most items out of the least resources with similar item/sec values (surpasses pure speed eventually, but you need tons of beacons for that... 11-13)
With eff/speed/prod you should get lots of items, at a decent rate with low energy cost, but the item/sec value is considerably lower. It should still be high enough to be not just usable, but good.

In my base for example I would use the latter, because I don't produce that much raw materials yet that I can handle any reasonably sized prod/speed setup, so I might as well turn down the speed and get much higher energy efficiency in return.

[BenSeidel]

then you need 1 efficiency 3 module in ONE of the beacons and you got optimal energy

It's a bit more complicated than that because you don't actually want -80% energy, but to minimise the energy per unit of work done. More beacons with the appropriate number of speed and efficiency modules reduces the energy per item, but the ratios of speed to efficiency is not a constant X:Y ratio. (it also varies based on the number of productivity modules and base energy consumption of the factory).
For example, you always place 3 speeds in first, then start putting in efficiency modules. Anyway, there are a bunch of topics dedicated to this and I'm only bringing them up because you must have an understanding of this aspect of the game before you are able to talk about how "unbalanced" something is.

There are other additional factors that affect Efficiency modules, such as the game update time. The increase in entity count because of the reduction in speed and increase in efficiency modules everywhere is a major drag (the factorio developers also don't seem to care about optimising long running "processes" such as a the 20 seconds it takes to make an engine).

The other issue with balancing them is that you are trying to balance people's goals. At a guess I would say that 95% of people's map-goals would be to rape & pillage the earth and build soooo much stuff. eg: "Can I get 200k/min iron plate?" or "how do I play". Not many people would do a no-wall/no turret challenge (omg do you need efficiency modules in EVERYTHING), or do "no solar" (steam engines are a pain) or other such goals that dramatically change the dynamic of the game.

If however the efficiency module were to be split into a power reduction module and a pollution reduction module then you would have a far more explicit use-case for each type, but trying to balance two competing concerns is difficult.

[greep]

Here's the thing syrchalis, balancing eff 2/3 only makes sense when considering a mix of speed and productivity. Mostly because eff1 handle all other cases well enough perfectly or in the case of two slots, good enough.

And as ben says, speed + prod mix has a synergy in reducing energy costs. Because a 4 prod setup significantly increases energy and lowers speed, and speed modules work additively, the best way to lower energy costs with prod modules is actually speed modules interestingly enough.
So eff3s have to compete with that. Eff2/3 only are useful in vanilla if you're using them in combination with speed modules and no production, but in vanilla of course the cost is too high. I think at a 30/160/800 balacing, I might even only use them if I were mine prod beaconing xD But it would at least give others an alternative to energy building spam and clean factories.

Regarding energy/item see the wiki here:
https://wiki.factorio.com/Module

As you can see, with only 4 full beacons of speed, a filled prod3 assembler uses less than 1/3 energy/item what it would have without speed modules. Plus the insane gains from needing far less modules for the same productivity effect.

Tl;dr, if efficiency modules are to be balanced in a speed/prod module setup, their vaules need to look pretty silly. 800 is less than it sounds, since it'd be 400 due to beacon halving, you'd at least need 2 modules in one beacon, for 6 beacons. And as you've noted that doesn't even take into account beacon energy. Anyways, not married to that number, just saying that's the ballpark I'd use if we really cared to balance this thing.

[Syrchalis]

I guess you ignored the link I posted? I did the math quite extensively.

The values I suggested are based upon that and take in account prod/speed setups, because those are the relevant setups. I rarely see people use a different setup if I'm honest, though I wanted to do a topic about that for a while.

Let's assume -120% for Efficiency 3:
4 Productivity Modules in the factory in both cases

4 Speed Beacons - +600% Energy Consumption - +140% Crafting Speed
6 Speed Beacons - +740% Energy Consumption - +240% Crafting Speed
8 Speed Beacons - +880% Energy Consumption - +340% Crafting Speed

4 Mixed Beacons - +220% Energy Consumption - +40% Crafting Speed
6 Mixed Beacons - +170% Energy Consumption - +90% Crafting Speed
8 Mixed Beacons - +120% Energy Consumption - +140% Crafting Speed

Mixed = 1 Speed 1 Efficiency

The items per second for speed beacons is around twice as good. The energy used per item stays between 50% and 70% for speed beacons, but for mixed beacons it goes to 100% at 6, and 150% at 8, so at 8 mixed beacons you actually get more items per energy than without any modules, that's something prod/speed doesn't achieve. At 1 million speed beacons energy efficiency for prod/speed is still 86%, it's an asymptote. I focused on 8 beacons because that's generally the amount you get affecting rows of assembly machines or furnaces.

For -150% Energy Consumption it's already above 300% energy efficiency at 8 beacons. And if we look at machines using only 2 productivity modules (like furnaces, chemical plants etc.) you max out at 6 beacons already, giving you 1320% energy efficiency.

With those numbers, why would you even go for pure prod/speed anymore? Why not just make twice the amount of assembly machines, furnaces etc. with this setup and have A FRACTION of the energy cost? It's not like the buildings cost a lot to produce. Building two or four screen-sized solar fields for each production step however does cost quite a bit.

Important: I didn't account for beacons using lots of energy. It was to simplify the whole thing. I do however believe that something should be done there as well. Beacons costing so much energy kind of defeats the point of efficiency modules in my opinion. Either their base values should be lower or they should get affected by eff modules themselves - it would make sense, because speed/prod don't give beacons any benefit (how would they even?) so they also don't make them cost more energy, but eff modules could totally affect them.
Even without this however, your total energy consumption at 8 beacons is still less than 40% of what it is with pure speed/prod while your crafting efficiency (items per second) is around 60% of what it would be for speed/prod.

[greep]

I saw you mentioned an excel sheet but I didn't see a link, I guess I just missed it o.O
Yeah those numbers look right, so at 6, it's be ((100 + 740)/(100+240)*1.4) = 176% energy/item; for mixed its nearly dead even at 101%

I was just saying the numbers I'd use them at (cost savings in energy building, module costs), you're right that in pure energy/item efficiency, mixed/pure efficiency wins. As ben points out x3 modules are post-endgame, and it's kind of hard to imagine a scenario where I would care about just the energy/item alone (pollution concerns essentially) since you're forced into a double laser wall anyways to stop biter expansion. RIP victory poles

In any case, regarding this is why I'd prefer a much higher number:

With those numbers, why would you even go for pure prod/speed anymore? Why not just make twice the amount of assembly machines, furnaces etc. with this setup and have A FRACTION of the energy cost? It's not like the buildings cost a lot to produce. Building two or four screen-sized solar fields for each production step however does cost quite a bit.

You're forgetting the module3 costs, if it were just the assemblers/furnaces, that would be peanuts. Some math here, calling plastic 2, iron 1, copper 1 for simplicity.
Going with 6 beacons, 6 assemblers, +amount of energy buildings needed to support them + modules for the speed setup, compared to the mixed.

per Beacon cost: ~300
per Assembly3 cost: ~400
per eff3/prod3 cost: ~3800
per speed3 cost: ~3000
per solar+ fractional acc: ~100

Speed:

raw cost: 6*300 + 6*400 + 12*3000 + 24*3800 = ~131400
#solar to support assuming = 40MW avg. over the day (60*~.7) Beacons->6*(400/40) = 60, assemblers = 6*840%*210/40 = ~270. cost in energy buildings: 33000.

Total: 165,000 stuff

Now, mixed would need (100% + 240%)/(100% + 90%) = 1.78 * the raw cost above since it need that many more beacons/modules for the same output speed. eff3 modules cost a bit more, but I'm lazy so... that's

131,000 * 1.78 = ~234,000 stuff. Let's call the energy free.

Way more stuff xD. 8 beacons is worse @1.83 * raw cost. Although I'm not sure how you fit 8 beacons average into assemblers anyways

[Syrchalis]

per Beacon cost: ~300
per Assembly3 cost: ~400
per eff3/prod3 cost: ~3800
per speed3 cost: ~3000
per solar+ fractional acc: ~100

Speed:

raw cost: 6*300 + 6*400 + 12*3000 + 24*3800 = ~131400
#solar to support assuming = 40MW avg. over the day (60*~.7) Beacons->6*(400/40) = 60, assemblers = 6*840%*210/40 = ~270. cost in energy buildings: 33000.

Total: 165,000 stuff

Now, mixed would need (100% + 240%)/(100% + 90%) = 1.78 * the raw cost above since it need that many more beacons/modules for the same output speed. eff3 modules cost a bit more, but I'm lazy so... that's

131,000 * 1.78 = ~234,000 stuff. Let's call the energy free.

Way more stuff xD. 8 beacons is worse @1.83 * raw cost. Although I'm not sure how you fit 8 beacons average into assemblers anyways

Uhm, I think you misunderstand. You trade speed for energy efficiency and reduced pollution (opposed to insanely high pollution). Trading means that you don't compensate for the lack of speed by building more beacons. You build the exact same number of modules and beacons. I doubt you can even fit enough beacons to reach prod/speed modules crafting speed with the mixed setup.

Basically you have these three options, if efficiency modules got buffed:
1. Horrible pollution/energy demand - maximum output in lowest time for the least resources (prod/speed)
2. Very good pollution/energy demand - normal output in normal time for normal resources (low tier efficiency modules)
3. Good pollution/energy demand - good output in good time for the least resources (prod/speed/eff all mixed)

The investment for #1 and #3 are the same - #2 is a lot cheaper (that's why it is even viable).

The only difference is that for SOME REASON I DONT UNDERSTAND AT ALL speed 3 modules cost 1 T2 module less than the other two (wtf Factorio dev team?). And that really is close to negligible.

Again - the goal isn't to get the same speed as pure prod/speed setups. The goal is to give players an actual choice that both have benefits and drawbacks. Making efficiency modules give -800% energy consumption would just make the mixed setup the new optimal strategy. And that is not what I want. I want more meaningful choices, because right now it is quite thin when it comes to late-game setups for modules. For end-game it probably will always be speed/prod just because of throughput, but that is okay.

[greep]

Well, it kind of depends on what you mean by "speed".

If you have something you need more than one producer of, like green/red circuit assemblers, copper cables, or furnaces producing metal, than there's functionally no difference in having 2 assemblers/furnaces producing at speed 1, and 1 assembler/furnace producing at speed 2. All you're doing when you build a second assembler of the same type as another in your factory is increasing speed.

That means you really do need to multiply the values for any case except where you aren't building multiple assemblers/furnaces. And you'd have to have a really bizarre factory that both doesn't have multiple furnaces/circuit assemblers and also builds tier 3 modules en masse.

Those extra solo assemblers won't make up the majority of your factory (probably 5%?) and are mostly end products anyways that won't need beacons (furnace assemblers, railroad assemblers, etc) because they wouldn't even need the speed of the mixed case, in which case it's eff1 or nothing.

If you already knew to think about it this way, then I believe you're arguing for a case that doesn't exist: A mid-game that produces bulk tier3 module. You kind of need the mass production to support these things in the first place unless you want to spend several hours watching one module get made. Anyways, I wouldn't mind seeing an actualy use case of where you personally would consider a setup where you'd use these things at 120%.

As for making the number too high resulting in it being the new norm, well maybe if it were 800 or something, I just threw that number out there. But I know I'd never use them anywhere at less than 350 or so, and wouldn't even use them even in mines at less then 150 probably.

[Syrchalis]

I guess you just don't care about energy or pollution. That doesn't mean everyone plays that way. When it comes down to personal preference the balance is right, so at 120% they wouldn't suit you, because they are not so overwhelmingly overpowered that you feel forced to use them. But many others might find them appealing enough to use them at these values. And that's the goal, to make them a viable alternative - not "THE one optimal setup".

Also the amount of space you have to designate to solar fields for prod/speed is pretty insane. Usually several times the actual size of the assembly machines/belts.

No matter if 120%, 200% or more, right now they are just worthless. Eff 1/2 have their place, but Eff 3 just seems so edge-case useful it might as well not be in the game, I wouldn't miss it at least. I will try if I can make it work in a 3/1 eff/speed setup in assembly machines, but I doubt it. For me the constant pasting of solar panels is really annoying, that's why I don't find prod/speed so attractive.

My current game is in a thick forest, so pollution is actually not much of an issue. I never triggered a biter attack, except once where they expanded basically less than one screen away from 30 speed beaconed oil pumps with 50+ pollution each.

[greep]

Ah sure, my point was just a mixed 120 eff3 module setup doesn't save resources, it costs way more to build. And while one needn't necessarily balance them so they're "optimal", I wouldn't use them in mines, let alone assemblers or even furnaces at that rate; that's a wide range of potential power and use and I wouldn't use them at even their best whereas I might use them in some places with a more favorable value, and only everywhere at sickeningly high values. It's just a matter of timing: t3 modules just don't become viable in general until way after pollution is a concern so to me its more about resource saving vs convenience in energy building numbers and only pollution in far off mines.

Anyways, argued to death on this, I think I've made every point I can xD

[Syrchalis]

I think though we can agree that right now eff 3 has just pretty much no place in the game. Something should change, so that it might find some actual use. Actual numbers aren't that important, that's for the devs to figure out.

The other modules seem pretty fine at least.

[meems]

i think the productivity modules are weak and oddly balanced.
Clearly the devs were worried about the geometric effect of prodmods over mutiple machines. Which is a fair point, but to my mind the current state is overkill. Prodmods are nerfed in their strength and not allowed in beacons.
So even with full investment of tier 3 prods over the whole assembly process its going to around +150%% productivity. Compare that with speed mods that can gain +300%..400% quite easy with beacons.
But productivity is only a bottleneck for advanced items and oil. So players aren't going to run prodmods over their whole assembly line, so prod bonus will rarely go over +100%.

I buff my prod mods to +5%, +9%, +12% prod, -10%, -12%, -15% speed.
I don't get why there's a need start at 4% then give a meagre +2% for 2nd tier to 6%, only to return to +4% -> 10% for tier 3.

[meems]

The jump from +30% to +50% for the speed tier 2 to 3 mod doesn't make sense. Tier 3s are too strong.
There is little question that when speed is a bottleneck that tier 3s are worth the investment nearly every time. The player doesn't have to weigh up the creation costs, they can always just go for tier 3.

Suggest nerf tier 3 speed mod to +45% or even +40%. Then tier 3 speed would be a luxury for very late game or necessary only for major bottle necks. As it stands tier 2 speed mods are a very brief steeping stone, used only to speed up creation of tier 3 speed mods.

As for efficiency mods... who cares? Just dump another 100 solar panels ( boring ).
Only a consideration on mega alien worlds.

[Syrchalis]

Who cares? I do - either remove them or make them worthwhile. Anything in between is just avoiding the issue. Efficiency 1/2 are good and worthwhile in many situations, but Efficiency 3 is just meh.

As for productivity, I agree that it should probably be 5, 7,5 10% or similar - but you should probably check your math again. It's not 150%. A chain with 4 steps yields 207% (+107%) if every part only uses 2 productivity modules. A chain with 4 steps where every machine has 4 productivity modules yields 384% productivity (+284%). An example is miners -> furnace -> green circuits -> red circuits -> blue circuits.

Energy Efficiency goes to 132% in a scenario with 4 steps with each 4 prod modules. But even for machines using only 2 prod modules the energy efficiency after 4 steps is 107%. And this is really the whole point. You get TONS of free items and you don't even have to pay more energy for it in the end. (this is assuming 4 speed beacons)

If productivity was buffed than that would become just more on the strong side. I do personally really like how productivity is good at giving you free items, no matter what, but also that it eventually gets energy-efficient when having equipped many factory parts with it.

Here math (3 steps, speed modules = speed beacons obviously, energy efficiency means energy used per item, crafting efficiency means time used per item):

[meems]

crude -> petrol -> plastics -> reds - > blues -> rocket part - > rocket

this is the crucial productivity bottleneck.

mod sockets = 2 at ref, 2 at c plant, 4 at red, 4 at blue, 4 at rp asem, 2 at r silo.
I calc 474% productivity with 10% prodmods in all sockets, and 618% with 12% pmods. That last figure is what desuades the devs off 12% pmods, but its a extreme figure and also comparable to speed mod extremes.

In practice, its always worth taking 10% tier 3s over 6% tier 2s. But if tier 2 was buffed to 7%, then it would not be as clear cut. Same with 9% to 12%.


Efficiency just isn't a factor in a game where u can place 100 solar panels without any logistic work or penalty. For efficiency to be a crucial consideration, pollution needs to be a crucial consideration : i.e. NOT using them causes the base to be overrun by aliens. Also this is another reason for waste management to be a part of the game : the most requested feature in the game by the community. But I'm beginning to think the devs have some extra bias against waste management. Saying it would make the game too complicated is like saying having a queen in chess makes chess too complicated. It does make it more complicated, but chess is more fun with queens on the board.