Efficiency or Bust

[Silvador]

Not really looking for advice, here, more opinions.

While googling for tutorials to help me understand circuits, I fell down a bit of a rabbit hole with DoshDoshington videos. While he wasn't exactly aiming for efficiency in the videos I ended up binging, I still noticed something about the production lines he was using in them and it got me wondering... Should I bother being efficient?

I know I don't technically -need- to be efficient, but after a while playing DSP, where outputting production of items at a certain rate is... well, easier, aiming to produce at least 1/s of things has kind of been ingrained into me, which in turn has led me to instinctively trying to design my little production lines in Factorio as efficiently as I can. But... as I get further along, that seems to be getting harder and harder, particularly if I want to keep said production lines not looking like a mangled mess. I also realised that a lot of the items I'm automating, especially early on, don't really need to be efficiently produced because they just sit in a chest until I -eventually- come to collect them, to build more things. Because of this I've had to -force- myself to say "stuff it" and just leave the production line -working- instead of -working efficiently-, while planning to 'make it efficiently when/if I need to, later'.

TL;DR - Do you bother with efficient production lines? Do you think efficiency is important in Factorio? Do you think it makes an overall difference to the speed and/or enjoyment of the game? Obviously this pertains more to casual gameplay rather than speedrunning.

To clarify, when I say "efficient", I'm talking about building a line of production that builds enough components to ensure the final product is crafting without stopping to wait for more of the required intermediate parts, without just "brute forcing" a boat load of materials down the line. And then making sure the final product is producing at or as close to one-per-second as possible.

[Fletch]

I think "efficient" is an overloaded term, and you'll need to quality what you mean by efficiency when discussing it.

Such as: efficient with player's time, or pollution/biter evolution, or efficient with raw resources (avoiding the recycler), etc.

[Silvador]

Originally posted by Fletch:

I think "efficient" is an overloaded term, and you'll need to quality what you mean by efficiency when discussing it.

Such as: efficient with player's time, or pollution/biter evolution, or efficient with raw resources (avoiding the recycler), etc.

Recycler???

I guess I'm talking about ratios? Figuring out an effective layout to produce enough of each item to ensure the final product is being produced without stopping to wait for more resources, and increasing that production to reduce the overall production time of that final item, without just having thirty assemblers making the intermediate components enmasse and shoving them all down the line.

[Nonotorious]

Depends on what you mean by "efficient".

[Xeno42]

it nice when all the production lines up to a factory ins't overproducing or under producing, but i dont worry too much about that except for things like curcuits where they can get really messy if done half assed. but for things like say gun turrets, i just hook 1 gear assemblers to 1 turret assemblers and call it a day

[terry.langeveld]

I often just guesstimate it at first. It is easy to make a mistake and then realize your nice design only produces at half its capacity. You can always learn from the mistakes of the initial design and redo it better if you need to scale up.

[BlueRock]

I don't worry about efficiency too much, but there are a few simple things you can do as far as efficiency to make your life easier. If you only spend a little time on efficiency you can do the following: 1. Limit chest sizes using the red X icon in them to only what you need to reduce pollution (reducing enemy attacks) and allow resources to get to other machines. 2. Put productivity modules starting in your most resource gobbling buildings first. This will create more of your expensive things for free. Labs, rocket silo, the more expensive science packs, processing units, LDS, etc. 3. Consider efficiency module 1's in miners near enemy bases. It will considerably reduce the pollution created reducing the costs of defending the outpost and lower energy use further reducing pollution.

[fx!]

If you are talking about matching outputs/inputs between different factory blocks it depends on what your goal is and bunch of other factors. If your new block is producing too much rest of the factory might suffer until said block fully converges (usually by excess output fully saturating belts/chests/inputs). On the other hand producing less than needed leads to the common problem you've mentioned of not having enough surrounding space for expansion. It was a fun learning experience for me to figure out which parts of factory worked better through careful planning and where I could just wing it. Early automation is probably the latter. There is always room for improvement so I wouldn't worry too much about it.

[Fletch]

Overbuilding intermediate production buildings isn't really a bad issue -- as the factory grows, you'll quickly realize that what you thought you had overbuilt, is now woefully insufficient and you need to double it (and later double it again).

[Chuck]

Originally posted by Silvador:

To clarify, when I say "efficient", I'm talking about building a line of production that builds enough components to ensure the final product is crafting without stopping to wait for more of the required intermediate parts, without just "brute forcing" a boat load of materials down the line. And then making sure the final product is producing at or as close to one-per-second as possible.

This kind of sounds like a main bus type of factory, dumping resources and intermediates on a line. Im not really sure why this isnt desirable, and I dont really know in what way its 'inefficient'.

I guess my best advice is dont worry about efficiency. Maybe in the early game, pollution efficiency is good, as you might not have the spare resources to automate gun turrets, nor the bots to replace them, so hiding from biters isnt a bad idea, and maybe if youre really into this game and making endgame factories that basically fight the game's ability to process it you might need to change your factory building thought process, but im gonna venture and say 95% of people, me included, will never get to that point. Once early game is not an issue, stop focusing on efficiency and start focusing on scalablility. Make a factory that you can easily expand upon. I always do main bus, mostly because i havent tried city block, but theres a reason these kinds of factory layouts are popular, its because expansion is simple and youre going to be doing alot of that in this game.

[Hedning]

Efficiency is important, but focus on the right things.

Building 1:1 wire to green circuits (instead of 3:2) is good.
Using a bus to move green circuits between green production and blue circuits is bad.

Mining the entire ore patch and sending all the excess to steel buffer way before you need lots of it is good.
"Upgrading" your steel furnaces to electric furnaces before you need the module slots is bad.

Building a full line of smelters even if you need only half a line is good.
Having lots of assembly machines making assembly machines is bad, especially if you do not limit the chest.

[Swede]

You might considering installing https://mods.factorio.com/mod/factoryplanner as one of the in-game mods.

Beware, however, that it will unfortunately disable achievements while the mod is turned on.

[Chindraba]

Originally posted by Silvador:

Not really looking for advice, here, more opinions.

TL;DR - Do you bother with efficient production lines? Do you think efficiency is important in Factorio? Do you think it makes an overall difference to the speed and/or enjoyment of the game? Obviously this pertains more to casual gameplay rather than speedrunning.

Efficiency is boss. However, I am King. Sometimes I'll go for efficiency and sometimes I'll go for the Nike slogan. Odds are higher towards efficiency if it's part of the science production tree and higher towards just do it for mall or support systems. I've been known to slam together something which works, supposedly, in an hour or two, and spend a week making 'the perfect' something. Sometimes I'll even do what looks like one and is the other - both ways.

Four variations from the same map;

The early mall. Looks rather 'organized' until you notice the belt torture inside. (Especially in the far left end.)
https://steamcommunity.com/sharedfiles/filedetails/?id=3279953709

The expanded mall, tacked on beneath the previous section. Efficiency is not the goal here. Not even for the Mil sci buried deep inside it. Top edge next to the diverted copper plate belt.
https://steamcommunity.com/sharedfiles/filedetails/?id=3303144724

The original, and yet never replaced, yellow sci. Calculated to the point of choosing which ASM to use and odd module choices. All in the name of using everything from the input and wanting nothing in the final assemblers.
https://steamcommunity.com/sharedfiles/filedetails/?id=3303144409

An interesting blend of both. Looks like a mess, perhaps even a plate of pasta, after it's been dumped. Yet it's designed to fulfill the building of rocket ingredients to a set rate. Quite efficient, and a few days (or more since I can't recall that far back) to 'design' within the constraints (cliffs) at hand.
https://steamcommunity.com/sharedfiles/filedetails/?id=3324147742

Bonus content: breaking the rules for how to build a silo. The rocket ingredients all end up here, eventually. Direct insertion for the silo is an interesting experiment.
https://steamcommunity.com/sharedfiles/filedetails/?id=3324147916

The bottom line of all that is: I had fun. And, having fun is the only correct way to play the game.

[knighttemplar1960]

Depends on your goals. If your goal is to produce 1 item per second of each item that is pretty broad as some items take more than 1 second to produce. If an assembly machine takes 20 seconds to produce 1 item you'll need 20 machines to get 1 per second. Calculating that can be a pain and failing to take inserter speeds into account can skew your numbers but you can do it simply without any math for any design you care to test.

Save your game. Lay out your machines and set out your power poles but don't hook it up to your power network. Hook the machines up to supply and when the belts are full turn the power on. Watch while it runs for about a minute. If the output belt is saturated remove any machines that aren't making items. If not add machines until the output is saturated.

If you use up all of one of the inputs remove machines past the point where the supply ends. To easily calculate items per second after you have removed all unsupplied machines. Turn the power off and clean out the output belt and any chests. Let the belts fill again. Put a chest and high capacity inserter at the end. Turn the power on for 60 seconds and shut the power off. Count the number of items left on the output belt and in the chest and divide by 60. That's the output speed in items/sec for that bank of machines. If that's sufficient you are good to go. If not you'll need to build a second set of machines that receive full supply.

Once satisfied blue print your set up, place the blue print in your book and exit the game without saving (or save that game as a test file) reload the game you are playing and use the newly created blue print to build your efficient design. You can type in the consumption and output rates into the blue print description so you don't forget what the blue print does.

[Shurenai]

Personally, I dont care so much about efficiency. My personal goal is perpetually to reduce the footprint of a part of my factory as much as possible. If I can get a tiny area to output a satisfying amount of resulting product, that's good enough for me even if I could've gotten 100% efficient input->output by spreading it out more and adding more machines.

I just enjoy cramming as much production into as little space as possible, enjoying the process of puzzling out how to compact it just a tile or two further whenever I can.