Link to the Spreadsheet I used in the Video. Feel free to comment on any errors you find: docs.google.com/spreadsheets/d/1cCm_DMUyNdQDkVa8wIg-mv1NbCzCMhV9HODJJL3i9MA/edit?usp=sharing
I checked your spreadsheet, and the thing that you talk about in the first few minutes (crafting rare from common is not 0.24*0.24, but 0.24*0.09) is exactly what you did in the very very first step of your calculations. I think the line 12 results from 15K crafst should be 10350.00/4185.00/418.50/41.85/4.65. It is done correctly afterwards, but that first line does change the whole thing :D Somebody correct me if my thinking is wrong. EDIT: Woah, with the correction in place (assuming it was an error), the productivity strategy is vastly, vastly superior to the others for both keeping th epics and for just legendaries.
@@dannyus007 This is what I had gotten before but you can even get better results (for legendary module count) if you mix modules in the same machine. When saving epics and legendarys, 5 quality modules in the machines yields: 714 epic and 197 legendary, 5 productivity modules in the machines yields: 1524 epic and 169 legendary, and 3 productivity + 2 quality modules in the machines yields: 1272 epic and 270 legendary. I think the extra 100L is worth the 250E. **note: all calculations done assuming full legendary modules. The numbers change a lot when dropping to lower quality and also change the fraction of modules within a machine that is ideal.
@@RestartVandelay That is what the wiki entry for quality states, that mixing provides even better results so that tracks. I tried doing the same thing in the table and got similar results, so 3 prod + 2 qual in every machine besides the legendary one is the best way to go.
So I did this exercise myself and my conclusion is that focusing on blue circuits you can create all the legendary circuits you need. It's quite easy to get more legendary blue than the number of normal blue circuits you put in. I've focused on only legendary circuits and my conclusion is what you want is to just build blue circuits. Then you recycle if you want red and green, you get half the amount of blue circuits as red circuits (0,25*2 red/blue), and 5 times the number of green (0,25*20 green/blue) This requires significant blue circuit productivity and you only use legendary productivity modules in Electromagnetic Plants. One you research130 % blue productivity your total productivity becomes: 150 % (base) + 130 % (research) + 125 % (prod mod). This means your total productivity is 405 %. Because this is higher than 4 you actually get more circuits after recycling and producing than you started with. The number of legendary blue circuits you produce at this point is 109 % of the number of blue circuits you put in. Not tried this is practise yet but going to try tomorrow.
I'm just here for the math and spreadsheets DLC. Ho ho holy shit we all appreciate your videos and kudos to Wube for putting a probability mini game that that doesn't require loot boxes.
I was really hoping for micro-transactions this dlc. Quality modules should be replaced by purchasable, one time use gems that have a chance to increase the quality of something. Once the crafting cycle completes, the machine sparkles until you click on it. Then you get a full screen slot machine where you watch the quality roll. Personal armors and equipment should be a gacha game. Maybe you pull a mech armor but it's only uncommon, better luck next week. Unless you just buy the packs that is. I'm really hoping for an epic roboport mk2 but I keep getting rare mk1s. The final fantasy 7 crossover will be awesome. I'm gunna get the spiky hair cosmetic on my mech armor!
" ... but i'm going to drag you through some spreadsheets because sometimes that's how we play factorio" is the most factorio thing I've ever heard. I love it.
That is an epic idea with the combinators. Didn't even notice that there was a new one. I also like the new ui displaying input- and output-signals directly without the need of a power-pole. Legendary!
There's one degree of freedom you didn't consider and there's a pretty good chance it's in fact better: Combining both quality and productivity in the same assembler. That's because you're multiplying the two numbers and that gives you the maximum when they are equal. The same reason why a square has the biggest area amongst all rectangles of the same circumference.
I just go straight with probability calculation and found out mixed modules has mixed result, primarily depends on how good your modules are, and full legendary productivity module always perform significantly better, even for cryochamber with 8 module slot.
22:00 you could use blueprint parameters to specify that some item is the 4th ingredient of a previous item, which then would propagate down to every other parameter. this is especially useful for builds where you have to set the parameter on construction and not dynamically; saving you both combinators and ups
When i read your comment, i tested it and there was a bug that prevented the ingredients for Module2 and Module1 to appear. It is now fixed and it work great :D
If I understand, can wire up all buildings in a group and 'Set Recipe' + 'Read Ingredients', you can set the one Building Recipe and all other buildings will follow what they can make from that. Like, Assemblers 3-2-1 + Gears + Chips all with one press of a button No combinator or parameters...
Honestly, Nilaus using the trick of Rocket capacity for getting the right input in the circuit feels the same as when I do a hack in my game code to get it to work how I want to. It´s like using a metallic spoon to close an electric circuit, not the original purpose of the tool, but it works nonetheless haha. Well done!
When quality Nutrients (or other ingredients) spoil inside a machine they turn into normal Spoilage. When they spoil elsewhere they turn into Spoilage of the same quality.
This makes sense. Imagine the added nightmare of having to account for various different rarities of garbage getting filtered out of your machines hahah
I think a fish farm would do perfect for efficiency modules on nauvis. Just upcycle the fish and wait for the quality fish to spoil. When there’s too much common spoilage you can then craft it back into nutrients with quality modules and wait for that to spoil as well!
Legendary quality modules in machines making legendary things hurt my soul (At 32:30 you put legendary qulaity modules into the em plant making legendary prods...)
My actual favorite part of these videos is the circuit network design process. Learning circuits as a professional software dev makes me giddy, like the first time learning Python!
if u want an exact amount of reciclers to process everything, I derived this formula: y = 4/(4-ax), x is the total productivity (base product + productivity)(if not all the machine have the same productivity, like this video, do a weighted average counting all the machines), a is the amount of items that stays in the belt (less then 100% if u take out higher quality items, the amount is 100% - the sum of the % of the higher quality u are taking ), y is the multiplier of the amount is entering ( if u have a red belt entering, 30 ITEMS/S and y = 4/3 u will need to consume 40 items/s). if a=1 and x=4 u will need an infinite amount fo reciclers
First, love the math! Really tickles the engineer brain in great ways. Also, I'm endangering the mission, I shouldn't have come. For what it's worth, you put legendary quality modules in the legendary crafting lines. Just loving the series, thank you!!
Nilaus you are kind of missing the forest for the trees, the easiest way to make quality is to go to gleba. I am producing every part as straight legendary, because everything there is based on bioflux. So i built a large base that specializes in legendary bioflux, which is quite simple really, and done. full productivity from there Also no headaches in recycling, bioflux just recycles into itself
This calculation is good, but it looks like you are starting with the strategy of tapping off your mass produced normals and recycling them. Another option would be to have raw ingredients coming in and start crafting with quality right away. A metric that is worth tracking is also items/s, because at the end of the day that is what you want to maximize and just feed it with as many raw materials as it needs.
Right. Nilaus demonstrates a static amount of resources-in to a static number of quality items-out. What it doesn't answer is the rate. If you're resource constrained, rate isn't as important to you. But since we are essentially unconstrained by resources in Factorio, rate again becomes a very important metric.
The problem with making legendary/epic circuits is you still need quality of the planet-specific item for each Tier 3 Module. So you end up with no legendary biter eggs for example, and tons of legendary T2/Circuits. That's why it's best to not do any of this, and just craft regular T3 modules (with quality modules in the EMP), recycle everything bellow legendary, and re-craft the lesser tier modules (with quality in the EMP). Don't mess with quality circuits at all, they'll come out of the recycling of T3s in the correct ratios and then you get the quality Egg/Tungsten/Superconductors as well. You can see the problem at the end of your video - every machine has legendary circuits but is missing the planetary items. I've got 1k of every T3 Legendary module this way, I do it on Nauvis so I can ship in Tungsten/Superconductors and keep eggs local so I don't have to worry about spoilage. You can create quality farms for the planet-specific items, but that's just inefficient. I have a quality biter egg farm and it pales in comparison to recycling T3 modules even with 20 legendary nests because the eggs are always regular quality whereas an epic T3 recycled gets you Epic/Legendary eggs every time an egg gets produced from the recycler.
i agree but it is nice if you want to quality everything to do circuits etc so you can build a variety of products my plan is to do both, generic materials quality for general items and specific upcyclers for annoying soecific items that have hard to upcycle base materials
I'll also be doing both. Prioritize upcycling normal quality final products as you describe. Then after that start working on a bottom up approach for some of the most common components.
Pretty sure (haven't done it yet) that recycled spoilage recycles into itself, so one option that immediately comes to mind is installing single-stage-recyclers into your spoilage-handling circuits on Gleba with quality modules in them. They will output some quality spoilage, which can be shipped back to nauvis for module crafting. (further processing may be necessary to get higher tier spoilage from the mostly-uncommon spoilage you'll be producing. You can also set up a multi-stage recycler loop to get the higher quality spoilage ON gleba but if you aren't careful about how its' set up it could actually starve your factory of the spoilage it produces, ironically enough (*especially if the majority of the nutrients you use is coming from spoilage and you don't have a dedicated bioflux->nutrient build to supplement the 75% losses the recyclers will cause) (*also, haven't tried it yet, but the excess of uncommon/rare quality spoilage you get from such a method may be able to be used to make quality nutrients without much re-factoring of your builds, which may help offset the losses the upcycling stage generates... set up a set of buffers for each of the quality tiers of spoilage the recyclers generate, and then any excess above...say, 500, gets converted into quality nutrients and the rest get "kept" on gleba until requested by the gleba-export ship to deliver to nauvis for module crafting. Also....you put quality modules in your tier 3 legendary module makers, you dufus xD
Hello, I am looking at the spreadsheet and I'm pretty sure, it does not apply in the case of your playthrough. You always supplement recycling lane with full belt of normal red circuits, which makes productivity on regular quality equal to throwing them out in the trash, because it can only produce regular quality product, which you have plenty of. I think the productivity strat should use quality in common assemblers, not productivity. That's why the combined method was so much better
Really good point. The spreadsheet doesn't assume you keep feeding it with new Normals, but that fact makes it obvious that Normal crafting should use Quality
My spreadsheet resulted in a different numbers from the one in the video, I think there has to be a mistake somewhere. especially seeing the fractional leftovers of lower qualities. But the idea of using different modules for each quality tier is brilliant, I only used different for legendary tier.
As a followup, if you consider common materials to be free (yay lava), then for red circuit cycling, you want: Common crafting: 5 quality modules Uncommon: 3 quality, 2 productivity Rare: 2 quality, 3 productivity Epic: 1 quality, 4 productivity Legendary: 5 productivity I just tested all combinations, and apparently on higher tiers productivity is better, because it preserves ingredients better, leaving you with a bit more chances of upscaling to legendary. It is optimized for legendary production speed, aka minimalization of machines needed to produce the same output.
What about if you qualited the miners on nauvis and recycled the ore until you get legendary and then productivitied the now legendary ore into its intermediaries?
Is there a reason to want anything other than the highest available quality? Other than the effort needed to produce, but that's the machines making the effort.
You're trying to scare me away with spreadsheets? I clicked specifically for the spreadsheets! Designs? Pft! Who needs 'em? As for qual vs prod, I hadn't considered using qual in one stage and prod in others, but I had considered using prod and qual in the same machine. The ratio between them seems to depend mainly on the strength of the prod modules, and full productivity is really only worth it with legendary prod 2s or 3s. (I use prod 2s because fuck biter eggs!) Using full prod does have a bonus benefits of enabling the use of speed beacons, which could be considered worth the extra material cost. Lastly, infinite prod research is so powerful, it pretty much singlehandedly makes cycling red and green circuits pointless. LDS cycling requires more levels to become lossless, but is even more powerful than blue circuit cycling thanks to the alternate casting recipe, letting you basically print legendary copper and steel for free.
ngl I am bad at math so didnt understand most of it. nice spreadsheet I guess. On the topic: have you considered having a combination like 2 quality and 2 productivity? if its in the spreadsheet I wont know so I ask
Given that blue circuits have the additional productivity research bonus, does it make sense to make vast quantities of legendary blue circuits and then recycle them into legendary green and red circuits?
Yes. Absolutely. You give it n common circuits and it will spit out n legendary circuits if you have enough levels of research prod. Any less than +300% prod and you lose material exponentially as the items go through the cycles.
@@Nilaus But with enough research you can reach the cap of +300% productivity so you just need to supply sulfuric acid and you can recycle -> craft in an infinite loop to upgrade N common blue circuit to N legendary blue circuit. This allows you to have a conversion rate of 100% for red and green circuit since you would craft the initial batch of blue with 300% prod to offset the final recycling step. You can do the same with LDS to get a 100% conversion rate for plastic, steel and copper.
Rant add, if you wire up all buildings in group and 'Set Recipe' + 'Read Ingredients', you can set one building recipe and all other buildings will set recipe of that selected one that they can make. Eg. Module Tier 3-2-1, Assember 1-2-3 with gears and Chips No combinator or parameters... 23:08 isnt 'Use Fillters' 'Whitelist' inserter above >. Normal more simpler and without wires??? Or if your brain works backwards, blacklist Normal OR
if the secondary quality rolls are capped at 10%, would it not be best to use like 2 legendary quality modules and 2 legendary productivity modules? You would get 12.4% quality and 50% productivity.
Have you considered only recycling blue circuits to take advantage of their productivity boost and then recycling legendary blues to make the red and green?
you need to supplement red, recycling blue gives you a glut of green but never enough red for anything you really want, my fulgora base can attest cause it's always waiting on plastic to make red despite recycling most blue
How are you moving the selection of quality without clicking on it with your mouse? It use to be shift+option and scroll but since the update that does not work anymore.
I would like to see from anyone, in simple enlgish, how the math works, side by side comparison, for each strategy, assuming infinite resources and power (so the end game)!
got a question for you or the community, is it possible to connect the planetary material request with the landing pad - space. here is my scenario, i put down a kinda big blueprint, a lot of the needed parts are produced on the other planets, i got a cargo ship going from planet to planet with all the basic stuff like belts, power poles etc. so my thought was that it would be amazing to give the landing pad the order to connect with the network, get the info what is currently missing for building projects (the info from the yellow cogwheel icon) and send the request to space so any bypassing ship of mine that has the material will drop it automaticly. hope there is no obvious solution to this and i get chased by you guys, i'm hours deep in and my head starts to smoke :D
I love the design, but you had a little error. You put legendary quality modules on the Legendary module build, which cannot get upscaled, that should have speed/efficiency as it cannot have prductivity nor more quality.
You can't because in the select input from the selector combinator it orders them not by their position in the recipe but by the quantity in the recipe. Since the efficiency modules need 5 spoilage and 4 modules it is in a different position than the quality modules that need 1 superconductor and 4 modules.
@@RestartVandelay isn't there a part in the blueprint parametrization that literally takes the 4th ingredient? cause on the wire the order gets thrown away, but saying it's the 4th ingredient of a recipe directly should work
@@jonasgajdosikas1125 Nilaus tried that on stream a few days ago but it wasn't working with the tier 1 modules. I just checked the vod to see what his problem was and it may have been that he overdefined red and blue circuit parameters. So it might work with just a constant combinator with 3 parameters for the first ingredients of the recipes but he was just having issues until he switched to this method.
idk what he means by "productivity vs quality" as you cant put productivity modules in recyclers and if you put productivity modules in the crafting machines your sacrificing your quality % for less resources used. Shouldn't you just put productivity modules if your crafting the highest quality available whenever possible?
You have not changed the formula for second and so on procs of quality in the very first line of calculations (10k input of normal ones). It is the quality chance from modules in power of procs now, but should be the quality chance * 0.1 in power of procs, as you did in all other lines.
@@dannyus007 Could you share your copy of spreadsheet? Also I found productivity strategy is only worth it with legendary modules. Epic modules seem to almost equalize the strategies, but still before legendary modules, quality strategy all the way.
@@vary01 I dont have a good way to share now, but you can do this yourself, just change the total crafting productivity and Crafting quality field to have the combination of modules. So for 3 prod + 2 qual its 2.25 productivity and 0.124 quality (1.5+3*0.25 and 0.062*2)
You don't need quality modules in your setup when the building is building the legendary version because it has a 100% chance to be legendary already. They should be productivity modules instead.
Btw u have to take in account the factory size into the calculations, the strat for no quality as input is way smaller and legendary put put is way more. For example get craft regular electric plants with quality inside rececyle the output then input into other rarities. Keep only legendary. Get insane the more common plants u craft. 1 day full chest with 5-10 regular plants crafts, way smaller setup and faster speaks for itself no mathing needed
Can't you just go algebraic math? So amount of qualities is just a vector and then multiply with matrices. 10% pro module would be obvious a matrix with 1.1 on the diagonal and quality matrix would be more or less the table from factorios own tutorial screen. 🤔
Thinking of that maybe you'll need even more dimensions in vector space to account all intermediates. But that just makes the matrices just bigger, math should stay simple multiplication in vectorspace.
You can get away with only 1 extra dimension to hold finished legendary products that skip the recycling step, and it collapses back to a 5x5 matrix after composing the recycler and the assembler together. If you use quality and productivity in the same machine, then you can basically just multiply the quality matrix by the productivity, though make sure to reset the passthrough component to 1. The easiest way to find the amount of output is to just take this 5x5 matrix to a large power and then apply it to a vector with the amount of input material of each quality. My personal code just uses an input of one and then reports the reciprocal, giving me how many commons it would take to get one legendary, though most scripts, including Nilaus's spreadsheet, seem to use a high input quantity and then report the output quantity for that directly.
I will not stand for this selector combinator slander! I use one to set hte stack size on my automatic wall resupply train stations! Totally not a niche use case that could have been done with three arithmetic combinators instead!
I'm not sure if you're sharing that as a negative. Nilaus explicitly is asking for comments/feedback/critique throughout the video and description so I'm not sure how anyone is surprised seeing critique in a comment section where the creator invited critiques. (in the case that you mean "the large amount of requested engagement is a testament to how good this creator is", I would agree)
You really should have a "conclusion" part. I don't have time or patience to watch for 33 minutes. Where do you make your conclusions? What's the result? Tell us. In an easy to find way.
Link to the Spreadsheet I used in the Video. Feel free to comment on any errors you find:
docs.google.com/spreadsheets/d/1cCm_DMUyNdQDkVa8wIg-mv1NbCzCMhV9HODJJL3i9MA/edit?usp=sharing
@nilaus what if you quality module base + uncommon, then prod, instead of just base?
I checked your spreadsheet, and the thing that you talk about in the first few minutes (crafting rare from common is not 0.24*0.24, but 0.24*0.09) is exactly what you did in the very very first step of your calculations. I think the line 12 results from 15K crafst should be 10350.00/4185.00/418.50/41.85/4.65. It is done correctly afterwards, but that first line does change the whole thing :D Somebody correct me if my thinking is wrong.
EDIT: Woah, with the correction in place (assuming it was an error), the productivity strategy is vastly, vastly superior to the others for both keeping th epics and for just legendaries.
@@dannyus007 This is what I had gotten before but you can even get better results (for legendary module count) if you mix modules in the same machine. When saving epics and legendarys, 5 quality modules in the machines yields: 714 epic and 197 legendary, 5 productivity modules in the machines yields: 1524 epic and 169 legendary, and 3 productivity + 2 quality modules in the machines yields: 1272 epic and 270 legendary. I think the extra 100L is worth the 250E. **note: all calculations done assuming full legendary modules. The numbers change a lot when dropping to lower quality and also change the fraction of modules within a machine that is ideal.
@@RestartVandelay That is what the wiki entry for quality states, that mixing provides even better results so that tracks. I tried doing the same thing in the table and got similar results, so 3 prod + 2 qual in every machine besides the legendary one is the best way to go.
So I did this exercise myself and my conclusion is that focusing on blue circuits you can create all the legendary circuits you need. It's quite easy to get more legendary blue than the number of normal blue circuits you put in.
I've focused on only legendary circuits and my conclusion is what you want is to just build blue circuits. Then you recycle if you want red and green, you get half the amount of blue circuits as red circuits (0,25*2 red/blue), and 5 times the number of green (0,25*20 green/blue)
This requires significant blue circuit productivity and you only use legendary productivity modules in Electromagnetic Plants. One you research130 % blue productivity your total productivity becomes: 150 % (base) + 130 % (research) + 125 % (prod mod). This means your total productivity is 405 %. Because this is higher than 4 you actually get more circuits after recycling and producing than you started with.
The number of legendary blue circuits you produce at this point is 109 % of the number of blue circuits you put in.
Not tried this is practise yet but going to try tomorrow.
And then at the end you put legendary quality mods in the legendary module buildings. I guess you were just exited to have it all finished. =)
Nilaus: Using Quality module to make legendary Quality is silly
Also Nilaus: *Uses Quality Module to make Legendary Quality*
Silly Nilaus :D
I'm just here for the math and spreadsheets DLC.
Ho ho holy shit we all appreciate your videos and kudos to Wube for putting a probability mini game that that doesn't require loot boxes.
I was really hoping for micro-transactions this dlc. Quality modules should be replaced by purchasable, one time use gems that have a chance to increase the quality of something. Once the crafting cycle completes, the machine sparkles until you click on it. Then you get a full screen slot machine where you watch the quality roll.
Personal armors and equipment should be a gacha game. Maybe you pull a mech armor but it's only uncommon, better luck next week. Unless you just buy the packs that is. I'm really hoping for an epic roboport mk2 but I keep getting rare mk1s. The final fantasy 7 crossover will be awesome. I'm gunna get the spiky hair cosmetic on my mech armor!
@@OniNaito When are we getting NFTs? I want my thousand dollar behemoth biter .png that can only appreciate in value.
" ... but i'm going to drag you through some spreadsheets because sometimes that's how we play factorio" is the most factorio thing I've ever heard. I love it.
if you are using excel, you can use the solver add-on to solve the system of linear equations or min max it.
11:56 "Let's go back to nauvis"
12:00 "Look at this beautiful thing we've here at fulgora"
12:01 Bye, beautiful thing
As a current engineering student, thank you for spreadsheets
Seven? C'mon man give us some credit. There's dozens of us. DOZENS!
That is an epic idea with the combinators. Didn't even notice that there was a new one. I also like the new ui displaying input- and output-signals directly without the need of a power-pole. Legendary!
There's one degree of freedom you didn't consider and there's a pretty good chance it's in fact better: Combining both quality and productivity in the same assembler. That's because you're multiplying the two numbers and that gives you the maximum when they are equal. The same reason why a square has the biggest area amongst all rectangles of the same circumference.
Yeah, I'm curious about mixing.
@@Seawolf159 so the mixed mixed result
@@janstoots1772 no. Nilaus' mixed result was about mixing *steps*. I'm talking about mixing *within a step*
I just go straight with probability calculation and found out mixed modules has mixed result, primarily depends on how good your modules are, and full legendary productivity module always perform significantly better, even for cryochamber with 8 module slot.
22:00 you could use blueprint parameters to specify that some item is the 4th ingredient of a previous item, which then would propagate down to every other parameter. this is especially useful for builds where you have to set the parameter on construction and not dynamically; saving you both combinators and ups
did you test this?
When i read your comment, i tested it and there was a bug that prevented the ingredients for Module2 and Module1 to appear.
It is now fixed and it work great :D
If I understand, can wire up all buildings in a group and 'Set Recipe' + 'Read Ingredients', you can set the one Building Recipe and all other buildings will follow what they can make from that.
Like, Assemblers 3-2-1 + Gears + Chips all with one press of a button
No combinator or parameters...
Honestly, Nilaus using the trick of Rocket capacity for getting the right input in the circuit feels the same as when I do a hack in my game code to get it to work how I want to. It´s like using a metallic spoon to close an electric circuit, not the original purpose of the tool, but it works nonetheless haha. Well done!
When quality Nutrients (or other ingredients) spoil inside a machine they turn into normal Spoilage.
When they spoil elsewhere they turn into Spoilage of the same quality.
I wonder if this is a bug. It definitely sounds like something that would get patched down the line.
This makes sense. Imagine the added nightmare of having to account for various different rarities of garbage getting filtered out of your machines hahah
If you recycle them, they turn into 2.5 spoilage is the same quality.
I think a fish farm would do perfect for efficiency modules on nauvis. Just upcycle the fish and wait for the quality fish to spoil. When there’s too much common spoilage you can then craft it back into nutrients with quality modules and wait for that to spoil as well!
Legendary quality modules in machines making legendary things hurt my soul
(At 32:30 you put legendary qulaity modules into the em plant making legendary prods...)
Nilaus. MATH IS MATHING! WE LOVE THE MATHS. ❤️
My actual favorite part of these videos is the circuit network design process. Learning circuits as a professional software dev makes me giddy, like the first time learning Python!
3:39 is actually really funny after hyping up the spreadsheet
Really enjoying watching this series. Keep up the good work!
you still put those legendary quality modules on the legendary recipes
Minor oversight. Easy fix.
if u want an exact amount of reciclers to process everything, I derived this formula: y = 4/(4-ax), x is the total productivity (base product + productivity)(if not all the machine have the same productivity, like this video, do a weighted average counting all the machines), a is the amount of items that stays in the belt (less then 100% if u take out higher quality items, the amount is 100% - the sum of the % of the higher quality u are taking ), y is the multiplier of the amount is entering ( if u have a red belt entering, 30 ITEMS/S and y = 4/3 u will need to consume 40 items/s).
if a=1 and x=4 u will need an infinite amount fo reciclers
First, love the math! Really tickles the engineer brain in great ways. Also, I'm endangering the mission, I shouldn't have come. For what it's worth, you put legendary quality modules in the legendary crafting lines. Just loving the series, thank you!!
Fly casual.
Nilaus you are kind of missing the forest for the trees, the easiest way to make quality is to go to gleba. I am producing every part as straight legendary, because everything there is based on bioflux. So i built a large base that specializes in legendary bioflux, which is quite simple really, and done. full productivity from there
Also no headaches in recycling, bioflux just recycles into itself
Do you mind sharing a blueprint?
Just don’t let that bacteria die 😅
This calculation is good, but it looks like you are starting with the strategy of tapping off your mass produced normals and recycling them. Another option would be to have raw ingredients coming in and start crafting with quality right away. A metric that is worth tracking is also items/s, because at the end of the day that is what you want to maximize and just feed it with as many raw materials as it needs.
Right. Nilaus demonstrates a static amount of resources-in to a static number of quality items-out. What it doesn't answer is the rate. If you're resource constrained, rate isn't as important to you. But since we are essentially unconstrained by resources in Factorio, rate again becomes a very important metric.
The problem with making legendary/epic circuits is you still need quality of the planet-specific item for each Tier 3 Module. So you end up with no legendary biter eggs for example, and tons of legendary T2/Circuits. That's why it's best to not do any of this, and just craft regular T3 modules (with quality modules in the EMP), recycle everything bellow legendary, and re-craft the lesser tier modules (with quality in the EMP). Don't mess with quality circuits at all, they'll come out of the recycling of T3s in the correct ratios and then you get the quality Egg/Tungsten/Superconductors as well. You can see the problem at the end of your video - every machine has legendary circuits but is missing the planetary items.
I've got 1k of every T3 Legendary module this way, I do it on Nauvis so I can ship in Tungsten/Superconductors and keep eggs local so I don't have to worry about spoilage.
You can create quality farms for the planet-specific items, but that's just inefficient. I have a quality biter egg farm and it pales in comparison to recycling T3 modules even with 20 legendary nests because the eggs are always regular quality whereas an epic T3 recycled gets you Epic/Legendary eggs every time an egg gets produced from the recycler.
i agree but it is nice if you want to quality everything to do circuits etc so you can build a variety of products
my plan is to do both, generic materials quality for general items and specific upcyclers for annoying soecific items that have hard to upcycle base materials
I'll also be doing both. Prioritize upcycling normal quality final products as you describe. Then after that start working on a bottom up approach for some of the most common components.
Pretty sure (haven't done it yet) that recycled spoilage recycles into itself, so one option that immediately comes to mind is installing single-stage-recyclers into your spoilage-handling circuits on Gleba with quality modules in them. They will output some quality spoilage, which can be shipped back to nauvis for module crafting. (further processing may be necessary to get higher tier spoilage from the mostly-uncommon spoilage you'll be producing. You can also set up a multi-stage recycler loop to get the higher quality spoilage ON gleba but if you aren't careful about how its' set up it could actually starve your factory of the spoilage it produces, ironically enough (*especially if the majority of the nutrients you use is coming from spoilage and you don't have a dedicated bioflux->nutrient build to supplement the 75% losses the recyclers will cause)
(*also, haven't tried it yet, but the excess of uncommon/rare quality spoilage you get from such a method may be able to be used to make quality nutrients without much re-factoring of your builds, which may help offset the losses the upcycling stage generates... set up a set of buffers for each of the quality tiers of spoilage the recyclers generate, and then any excess above...say, 500, gets converted into quality nutrients and the rest get "kept" on gleba until requested by the gleba-export ship to deliver to nauvis for module crafting.
Also....you put quality modules in your tier 3 legendary module makers, you dufus xD
Okay that design is reaaaaally cool
Hello, I am looking at the spreadsheet and I'm pretty sure, it does not apply in the case of your playthrough. You always supplement recycling lane with full belt of normal red circuits, which makes productivity on regular quality equal to throwing them out in the trash, because it can only produce regular quality product, which you have plenty of.
I think the productivity strat should use quality in common assemblers, not productivity. That's why the combined method was so much better
Really good point. The spreadsheet doesn't assume you keep feeding it with new Normals, but that fact makes it obvious that Normal crafting should use Quality
My spreadsheet resulted in a different numbers from the one in the video, I think there has to be a mistake somewhere. especially seeing the fractional leftovers of lower qualities. But the idea of using different modules for each quality tier is brilliant, I only used different for legendary tier.
As a followup, if you consider common materials to be free (yay lava), then for red circuit cycling, you want:
Common crafting: 5 quality modules
Uncommon: 3 quality, 2 productivity
Rare: 2 quality, 3 productivity
Epic: 1 quality, 4 productivity
Legendary: 5 productivity
I just tested all combinations, and apparently on higher tiers productivity is better, because it preserves ingredients better, leaving you with a bit more chances of upscaling to legendary.
It is optimized for legendary production speed, aka minimalization of machines needed to produce the same output.
Stamping down 4 blueprints and box deleting 4 times is fewer operations than stamping down up to 20 individual blueprints.
don't mind me replacing the legendary quality modules out of my machines making legendary...
i blame tired me
15:47 « its an older code but it checks out » nilaus dropping good stars wars ref here
Its time to gift yourself legendary mech armor soon (and maybe some higher quality construction bots)
Hello, at the end of the episode, why put quality modules in the machines with the legendary recipes?
What about if you qualited the miners on nauvis and recycled the ore until you get legendary and then productivitied the now legendary ore into its intermediaries?
The best way to get legendary red circuits I have found is to recycle the epic blue circuits that you can make way too many of.
You think your spreadsheets can scare me? I am a PoE player, we LIVE for spreadsheets, data and optimisation!
Is there a reason to want anything other than the highest available quality? Other than the effort needed to produce, but that's the machines making the effort.
You're trying to scare me away with spreadsheets? I clicked specifically for the spreadsheets! Designs? Pft! Who needs 'em?
As for qual vs prod, I hadn't considered using qual in one stage and prod in others, but I had considered using prod and qual in the same machine. The ratio between them seems to depend mainly on the strength of the prod modules, and full productivity is really only worth it with legendary prod 2s or 3s. (I use prod 2s because fuck biter eggs!) Using full prod does have a bonus benefits of enabling the use of speed beacons, which could be considered worth the extra material cost.
Lastly, infinite prod research is so powerful, it pretty much singlehandedly makes cycling red and green circuits pointless. LDS cycling requires more levels to become lossless, but is even more powerful than blue circuit cycling thanks to the alternate casting recipe, letting you basically print legendary copper and steel for free.
ngl I am bad at math so didnt understand most of it. nice spreadsheet I guess. On the topic: have you considered having a combination like 2 quality and 2 productivity? if its in the spreadsheet I wont know so I ask
Given that blue circuits have the additional productivity research bonus, does it make sense to make vast quantities of legendary blue circuits and then recycle them into legendary green and red circuits?
Yes. Absolutely. You give it n common circuits and it will spit out n legendary circuits if you have enough levels of research prod. Any less than +300% prod and you lose material exponentially as the items go through the cycles.
Yes and no. You are only getting 2 red from each blue, so you need to creator a ton of blues to keep up with your red consumption
@@Nilaus But with enough research you can reach the cap of +300% productivity so you just need to supply sulfuric acid and you can recycle -> craft in an infinite loop to upgrade N common blue circuit to N legendary blue circuit. This allows you to have a conversion rate of 100% for red and green circuit since you would craft the initial batch of blue with 300% prod to offset the final recycling step.
You can do the same with LDS to get a 100% conversion rate for plastic, steel and copper.
Love waking up to this
love falling asleep to this. (it’s otherside of the earth)
@premiumyoutube-hy4jq divided by borders, united in the factory
I've never noticed the logistic group request multiplier before 🤦♂
Rant add, if you wire up all buildings in group and 'Set Recipe' + 'Read Ingredients', you can set one building recipe and all other buildings will set recipe of that selected one that they can make. Eg. Module Tier 3-2-1, Assember 1-2-3 with gears and Chips
No combinator or parameters...
23:08 isnt 'Use Fillters' 'Whitelist' inserter above >. Normal more simpler and without wires??? Or if your brain works backwards, blacklist Normal OR
have you stopped producing circuits at nauvis or are you just using the vulcanus factory to supply the other planets?
if the secondary quality rolls are capped at 10%, would it not be best to use like 2 legendary quality modules and 2 legendary productivity modules? You would get 12.4% quality and 50% productivity.
Hey Nilaus! Do you plan to update your designs for Space Age with time?
Wow i am too dumb to understand what is happening but I love you videos keep up the good work😊
what about being able to use speed modules in beacons with the productivity setup?
Have you considered only recycling blue circuits to take advantage of their productivity boost and then recycling legendary blues to make the red and green?
you need to supplement red, recycling blue gives you a glut of green but never enough red for anything you really want, my fulgora base can attest cause it's always waiting on plastic to make red despite recycling most blue
How are you moving the selection of quality without clicking on it with your mouse? It use to be shift+option and scroll but since the update that does not work anymore.
I would like to see from anyone, in simple enlgish, how the math works, side by side comparison, for each strategy, assuming infinite resources and power (so the end game)!
19:25 why the heck didn't devs include "don't sort" option?
anyone knows why does he use quality modules to create legendary modules?
it's by accident i believe
got a question for you or the community, is it possible to connect the planetary material request with the landing pad - space. here is my scenario, i put down a kinda big blueprint, a lot of the needed parts are produced on the other planets, i got a cargo ship going from planet to planet with all the basic stuff like belts, power poles etc. so my thought was that it would be amazing to give the landing pad the order to connect with the network, get the info what is currently missing for building projects (the info from the yellow cogwheel icon) and send the request to space so any bypassing ship of mine that has the material will drop it automaticly.
hope there is no obvious solution to this and i get chased by you guys, i'm hours deep in and my head starts to smoke :D
I love the design, but you had a little error. You put legendary quality modules on the Legendary module build, which cannot get upscaled, that should have speed/efficiency as it cannot have prductivity nor more quality.
For the parameterized blueprint, can you not set the recipe for the second EM plant to be whatever is ingredient number 4 of the final plant?
You can't because in the select input from the selector combinator it orders them not by their position in the recipe but by the quantity in the recipe. Since the efficiency modules need 5 spoilage and 4 modules it is in a different position than the quality modules that need 1 superconductor and 4 modules.
@@RestartVandelay isn't there a part in the blueprint parametrization that literally takes the 4th ingredient? cause on the wire the order gets thrown away, but saying it's the 4th ingredient of a recipe directly should work
@@jonasgajdosikas1125 Nilaus tried that on stream a few days ago but it wasn't working with the tier 1 modules. I just checked the vod to see what his problem was and it may have been that he overdefined red and blue circuit parameters. So it might work with just a constant combinator with 3 parameters for the first ingredients of the recipes but he was just having issues until he switched to this method.
I love that he has the same brain virus as me
You're using quality modules in the plants already making legendary modules. Should be productivity.
It can't be productivity, and there's probably no need for speed at legendary quality. It should be efficiency or empty.
efficiency finally being the best after 200 years
idk what he means by "productivity vs quality" as you cant put productivity modules in recyclers and if you put productivity modules in the crafting machines your sacrificing your quality % for less resources used. Shouldn't you just put productivity modules if your crafting the highest quality available whenever possible?
You have not changed the formula for second and so on procs of quality in the very first line of calculations (10k input of normal ones). It is the quality chance from modules in power of procs now, but should be the quality chance * 0.1 in power of procs, as you did in all other lines.
I already commented about that under his pinned post, it changes the whole thing to heavily benefit the prod strategy when corrected.
And that seems to make pure productivity strategy the winner.
@@vary01 Gets even better for 3 prod + 2 qual modules in everything but the legendary crafter. Extra 100 legendary at cost of 200 epic
@@dannyus007 Could you share your copy of spreadsheet? Also I found productivity strategy is only worth it with legendary modules. Epic modules seem to almost equalize the strategies, but still before legendary modules, quality strategy all the way.
@@vary01 I dont have a good way to share now, but you can do this yourself, just change the total crafting productivity and Crafting quality field to have the combination of modules. So for 3 prod + 2 qual its 2.25 productivity and 0.124 quality (1.5+3*0.25 and 0.062*2)
They changed the keybind to change quality of the current item held in hand to ALT+Scroll..
thanks for the tip, that's SO much better (in retrospect, I wonder why it didn't occur to me to rebind it like that myself)
You don't need quality modules in your setup when the building is building the legendary version because it has a 100% chance to be legendary already. They should be productivity modules instead.
In general yes but you can't put productivity when making modules. It is a waste of good quality modules though.
Wait how did you get legendary biter eggs?
I'm confused. How can productivity help with getting higher quality? Without quality modules, upcycling is 0 surely?
Btw u have to take in account the factory size into the calculations, the strat for no quality as input is way smaller and legendary put put is way more. For example get craft regular electric plants with quality inside rececyle the output then input into other rarities. Keep only legendary. Get insane the more common plants u craft. 1 day full chest with 5-10 regular plants crafts, way smaller setup and faster speaks for itself no mathing needed
Don't need quality for legendary recipes
ok I just lost half a day trying to write a python script to simulate this (with chatGPT)
Can't you just go algebraic math?
So amount of qualities is just a vector and then multiply with matrices. 10% pro module would be obvious a matrix with 1.1 on the diagonal and quality matrix would be more or less the table from factorios own tutorial screen. 🤔
Thinking of that maybe you'll need even more dimensions in vector space to account all intermediates. But that just makes the matrices just bigger, math should stay simple multiplication in vectorspace.
You can get away with only 1 extra dimension to hold finished legendary products that skip the recycling step, and it collapses back to a 5x5 matrix after composing the recycler and the assembler together. If you use quality and productivity in the same machine, then you can basically just multiply the quality matrix by the productivity, though make sure to reset the passthrough component to 1.
The easiest way to find the amount of output is to just take this 5x5 matrix to a large power and then apply it to a vector with the amount of input material of each quality. My personal code just uses an input of one and then reports the reciprocal, giving me how many commons it would take to get one legendary, though most scripts, including Nilaus's spreadsheet, seem to use a high input quantity and then report the output quantity for that directly.
@@angeldude101 very cool that you already did a skript. Maybe I'll try myself next week when I'm back home just for the funs.
legendary modules on legendary recipes aren't necessary
Proudly one of the seven
I will not stand for this selector combinator slander! I use one to set hte stack size on my automatic wall resupply train stations! Totally not a niche use case that could have been done with three arithmetic combinators instead!
But I want more spreadsheet content
Bruh, I've never seen a comment section with so much critique. Always happens when you good at something.
I'm not sure if you're sharing that as a negative. Nilaus explicitly is asking for comments/feedback/critique throughout the video and description so I'm not sure how anyone is surprised seeing critique in a comment section where the creator invited critiques. (in the case that you mean "the large amount of requested engagement is a testament to how good this creator is", I would agree)
You made a big mistake on this blueprint eventhough you were talking about it 10 Minutes before. Quality Modules on Legendary Production how silly!
Yes, hello, I am 7th viewer.
I'm viewer #8
First
You really should have a "conclusion" part. I don't have time or patience to watch for 33 minutes. Where do you make your conclusions? What's the result? Tell us. In an easy to find way.
Use Productivity Module for Uncommon+ Crafting. Quality on Recyclers + Normal Crafting