Quality

[jaunty citrus]

Ah, I understand now.

It's an interesting idea but I'd say you're better off scaling somewhere like Nauvis where you already have infrastructure to scale faster.

[fallen urchin]

Does foundry give another recipe for steel?

I think steel prod might be less important at high levels because the recipe isn't used

Do we have parallel research?

Without parallel research, I think it's still better to have a centralized lab, rather than letting the specialized labs idle when not researching their tech

Maybe we can overbuild labs, and let the sci packs stockpile when not researching, and use them in bursts when researching their tech? But this requires huge buffers, which is especially problematic with Gleba

[burnt reef]

I think prod research applies to all recipes for an item, so even the foundry steel recipe gets a boost.

but since the foundry LDS recipe uses molten iron instead of steel, steel prod research becomes slightly less relevant then.

[frank warren]

What are known values for Foundry and EMP Q1 -> Q5 costs? I get 24.39x and 13.8x for proddable items respectively.

[teal wren]

I think it will be wisest to quality up first intermediate which can get productivity boost. This way you can easily make a lot of legendary plates and such and then using exponential productivity boost create a lot of legendary stuff

[fallen urchin]

So this is the idea of "qual in qual out, and qual ingredients make best use of machine craft time"

I think this fits recipes that have low ingredient cost and higher craft time

[exotic saddle]

prod modules are supposed to give negative quality

[teal wren]

Not productivity, speed modules...iirc

[exotic saddle]

so, if bottom recycler just uses a bunch of prod modules + speed to go neutral, then the bottom step to first tier of materials is 100% rank 2 quality as output

actually, you wouldn’t be able to do that, because if the prod ratio went that high, then you could use recycler for free materials forever

[solar osprey]

Recyclers can’t take productivity modules.

[exotic saddle]

what i can say is that this is extremely sensitive to the specific numbers

because of all the high exponents

no matter how you slice it, the difference between a modile getting you to 8% vs 10% is massive

[teal wren]

Exactly. Everything exponential is very sensitive

But combining prod and qual modules into assembler will be better than using all qual modules

[obsidian crescent]

Productivity is not "exponential"; it's multiplicative.

[teal wren]

Yeeeah, but it works

https://docs.google.com/spreadsheets/d/1GpszMksTmaaIV4opnKTx_e7JrrBPRaSp/edit?gid=122855994#gid=122855994

[exotic saddle]

the didference of 8% vs 10% quality up is the difference between one legenary item costing 4.1k vs 1.7k of its base item

it’s not exponential, but it is to the 8th power

which is still stupidly fast scaling

[solar osprey]

Compare 10% and 25%

[exotic saddle]

still at 25% recycling?

25% quality up on assembler/recycler corresponds to 0.345x quality increased output, or one legendary item being worth 70 of the basic version

quality modules in recyclers makes it 8th power scaling. without that, it is 4th power scaling

just generally, it will be a lot easier to balance quality if recyclers reject quality modules AND prod modules

[teal wren]

But that is what makes it interesting :D

[obsidian crescent]

That would be awful; it'd take forever to get quality anything.

[teal wren]

Wait... recyclers cannot be used with productivity modules

[exotic saddle]

they are going to be stuck in a cat and mouse game of optimizing the little digits forever if recyclers can take quality modules. and players will be pretty stuck without a 100% totally maxed out setup

[exotic saddle]

yes. i meant if they rejected both, instead of just prod

[obsidian crescent]

If you couldn't put qual modules in recyclers, then quality cycling end products would be the only effective way of making quality stuff.

[teal wren]

But that exactly why we have whole thread for quality. This discussion would never happen if it was all so simple

This is what makes factorio so interesting and complex

[jaunty citrus]

Recyclers reject productivity for an entirely different reason, not because it would make quality too good

[exotic saddle]

Just to make clear the formulas I’m using:
q = rate of quality up on one machine.
r = ratio of recycled output vs input
Assembler and recycler both have quality-up rate q, then ratio of total quality-up output (with recycling) is:
q [1 + r (1-q)] / [1-r(1-q^2)]

If only assembler has quality modules, then the ratio is: q / (1 - r + rq)

[obsidian crescent]

... so? The ideosyncratic nature of quality production is easily the best thing about quality.

[exotic saddle]

that is for one rank of quality up. For going to legendary, take that to the 4th power

[jaunty citrus]

I'm not sure I understand what issue you're claiming exists

[exotic saddle]

what I mean is that if recyclers take quality modules, the formulas are so extremely sensitive to the quality up rate, that wube will be eternally stuck balancing the little decimals to get the balance right. And players will be forced to have a completely 100% optimized setup to realistically make things of high quality. Which goes agaisnt having the freedom to do things in a few different ways.

[obsidian crescent]

To get what "balance right"? Like, what happens if "balance" is wrong? What does "wrong" look like and why is it wrong at all?

[exotic saddle]

Like the difference between 8% and 10% quality up rate is so astronomical, that you will never see a setup that uses 3 quality modules + 1 of anything else in an assembler

[jaunty citrus]

Have you considered any other methods you could use to get quality items outside of recycling them in loops?

[obsidian crescent]

And that's... bad?

I don't think WUBE cares about trivial issues like that.

[jaunty citrus]

Like yeah, recycle looping is a really bad way of getting quality items, but it's far from the only way

[exotic saddle]

you could do quality up generally across the factory, but if you want a supply of legendary beacons, the end of the road is going to have a mix of qualities, and you’d then need to do the recycling loop to make it all max

[exotic saddle]

afaik, quality modules, and then recycling are the only ways

[teal wren]

Snoop, this is the correct formula! One you've typed above. I had adjusted my program and now everything matches.

[jaunty citrus]

Have you considered that productivity research also exists and has a far greater effect on the cost of quality cycling? Once you reach 300% on a recipe it doesn't matter what the quality chance is it will tend to zero cost

[exotic saddle]

this formula is correct

[obsidian crescent]

Like no, they're not. There are so many ways of getting quality stuff.

Are you making purple science? Guess what? You're always making electric furnaces and prod 1s. You can't prod them, so you may as well stick qual modules in them and collect the good stuff for use in the factory.

You can put quality modules in miners and skim off the quality ore for processing elsewhere.

[exotic saddle]

i do not know about that mechanic

you’d realistically need some other sort of bonus besides what i describe to not burn a whole mine to make some legendary items

[teal wren]

This is just infinite research of productivity for many intermediates. So you can get +300% on assembler and even add some quality modules in.

[exotic saddle]

fyi, as far as formulas are concerned, every productivity multiplier would go as a straight factor multiplying into the 1 rank step

then obviously to the fourth for legendary

[obsidian crescent]

Using vanilla prods, to make 1K SPM, you need over 60,000 iron ore per minute. If you can get just a 5% quality buff on your miners, that's 0.375 Q5 ore per minute. And you can get even more if you put quality modules in the furnaces you use for the quality ore that comes out.

[teal wren]

I will do like that because it sounds very logical

[solar osprey]

Though I’d have to ask why you’d quality your mines and not just foundry it and quality the casting foundry.

[jaunty citrus]

By the way your not wrong about the numbers being sensitive. For some context (assuming all legendary modules) recycle looping normal items into legendary that don't accept prod modules costs:
154x normal for items made in an assembler 3
40x for items made in a foundry
30x for items made in an EMP

Without the machine crafting the recipe back for the extra step, that cost skyrockets to 2727x the cost

[obsidian crescent]

Because you'd probably be shipping molten iron everywhere, so that's a lot of different places to put quality modules and siphon off the remains. By doing it at the miners, you can cull them early and ship them to a dedicated quality facility.

[exotic saddle]

yeah, if there is any one intermediate that can be ranked up in quality at a high effficiency (eg +300% prod), then you’d want to exploit the shit out of that to get high quality

[solar osprey]

It’s a lot of miners to put T3 quality modules in.

[obsidian crescent]

My numbers used +5% quality. Qual module 3s give +2.5% quality each, so that's 7.5%. So obviously, not using those.

[teal wren]

At the point you will have anything more than +50% you realistically already will make most items legendary. And at point of about +100% you would make absolutely everything legendary

[obsidian crescent]

But yes, the downside is a lot more modules. But at least BMDs can lower that a bit.

[jaunty citrus]

Would be 10% for BMDs using qual3s

[solar osprey]

Lot of T2 qualities then.

[obsidian crescent]

Those are pretty cheap.

[solar osprey]

T1 qualities would get you 3%. T2 would get you 4.5%

[teal wren]

+300% is absolutely insanely costly research. 6.5 hours of research to get to the level 30. With 1M SPM

[exotic saddle]

it’s like prod modules: the more a given assembler is worth in terms of total resources/sec, the more the one module is worth. eg labs

[final nimbus]

+250% is the most you'll need out of research before productivity modules are no longer useful on certain items.

[teal wren]

Just 1 hour to level 25 with 1M SPM :D

[jaunty citrus]

Following up on this if an intermediary accepts prod modules and can be made in the EMPlant, (like all circuits) the cost goes down to 13x, even lower for processing units which can also benefit from prod research

[unborn flax]

Close enough - it's 24.25 and 13.23 (computed with no fp rounding until the end). Someone else once agreed with my 13.23 to many decimal places; I don't think the Foundry has been discussed enough for 24.25 to reach the same level of consensus.

[unborn flax]

100% this, and that's assuming there are items that have infinite prod research but can't use prod themselves (e.g. I'd be surprised if there were productivity research for modules).

Once you reach inputs=0.25 you're just saving on UPS as you need fewer cycles to produce your free legendaries.

[teal wren]

I really don't think savings on UPS are a concern when we are talking about producing materials for the factory... unless you already have quite expansive mall and factory at hand

[unborn flax]

I'm not saying it's a concern, I'm saying it's the only benefit of further prod research once you reach the bottom line

[teal wren]

Well, overall productivity research is extremely useful for anything really, because we are finally getting someting else than just mining productivity. Which is on itself is most basic and quite boring

[unborn flax]

Whereas the first level of prod research (+10%) drops the resource requirements by an enormous 21-22%, pretty consistently for all four lines on the graph.

[half raptor]

Fun part about prod research is the multiplicative effect a 6-8 production step process will benefit. I hope we can research module productivity but I won't put all my hopes on that necessarily

[frank warren]

Why do you need module prod?

[half raptor]

To make more modules

¯_(ツ)_/¯

And because we can't use prod modules when making modules (currently)

[obsidian crescent]

We already have module productivity at home.

Module productivity at home: Electromagnetic Plant.

[half raptor]

Don't get me wrong, it's already more productivity than 1.1 Assembler mk3 with full modules

But we could get more pls

[frank warren]

I did the math for the effective output and costs of upgrading to legendary

turns out a legendary setup costs as much as it's benefit for the EMP

(IE - per item/s, it costs the same number of resources to make a legendary setup as it does a basic one - ignoring inputs which are lower as well)

[tall sandal]

Wait! thats wild. So if you want to make (say) 60 q1t3 modules per minute, and you are considering two options:

• All the buildings and modules in the buildings are normal quality
• All the buildings and modules in the buildings are legendary quality
  The legendary quality factory costs the same as the normal quality factory? And what do you mean about inputs being lower? Wouldnt they be the same?

[frank warren]

Because it has more prod, it also needs less inputs from external sources. Here's my rough napkin version of the math.

• Modules are 2.5x as strong
• Beacons are 1.66x as strong
• Machines are 2.5x as strong

Overall buff is ~10x
Quality cycling in EMP has a cost of 13x.

Arguably this undersells the buff - because buffing speed and prod have multiplicitive effects.
This also undersells the cost, because not all intermediates can path through the EMP.

It’s a very rough rough estimate because there are a lot of factors I am missing, but what it suggests is that a legendary setup is not monumentally more expensive. Tell me if you see any clear issues.

This also doesn’t account for a quality setup needing less downstream inputs as well because it is at higher prod. It also doesn’t account for any recipe prod.

[tall sandal]

Fulgoria first chads just keep winning

[winter saffron]

But Bwuhuo probably has the spidertron

At least I hope so, since IMO it needs something quite powerful to compete

[tall sandal]

Its funny that we know nothing about anything it unlocks that is useful on other planets

[frank warren]

@tall sandal What do you think overall of this analysis?

Because speed and prod multiply, you could argue the difference is even 2.5x what this one is

[tall sandal]

I dont fully follow it honestly.

Is the goal of this to calculate the cost to build the factory or the cost to run the factory?

[frank warren]

to build

cost to build a factory that produces X output/s

I found that legendary only costs 30% more roughly, ignoring prod benefits downstream, and multiplicative effect of speed and prod together. (also ignoring that you can't make everything with EMP)

but most of the cost is modules, and all of their intermediates can be cycled with EMP, so I imagine it's close.

[exotic saddle]

bro, your axis goes all the way to 100 of an infinite research

[tall sandal]

So then it would only apply to items that the EMP itself can produce right? Ie, we can use to to make modules (hugely important) but not the EMP itself or assemblers or inserters.

[frank warren]

yes, but if this is most of the cost I think it's still fair, and I imagine it is. p3s are expensive

[obsidian crescent]

The EMP is made at the EMP.

[frank warren]

Still doesn't exactly count because I am talking about EMP intermediates, IE circuits.

That's the EMP making it cheaper - but that applies to both basic and legendary quality

What we have are two values: a cost multiplier for legendary, and a benefit multiplier. Divide the two and you get the 'true cost' of upgrading your factory quality. It seems this cost is far closer to one than people thought initially. Could even be less than one given just a few prod researches 🤯

This would have really weird effects because it would suggest that a legendary speed 1 is a better idea than a speed 3 for ROI - that may even already be true.

[winter saffron]

only have a speed penalty, right?

[frank warren]

yes - maybe power too

[obsidian crescent]

Quality modules of all tiers have a small speed penalty.

[jaunty citrus]

You're interpreting it wrong, it only goes up to 100% bonus, which is 10 levels of an infinite prod research

It's not the most clear graph to be fair

[jaunty citrus]

While it does undersell the cost, it's interesting that the 13x applies almost all the ingredients in T3 modules and beacons as they are entirely made from circuits/wire. We'll have to see about the T3 module components but I don't think steel is that bad to get quality of either, just set up an iron mine recycling barrels or something.

[winter saffron]

Do module penalties scale with quality?

[frank warren]

no

[jaunty citrus]

Module penalties not scaling tempts me towards making a high-efficiency base.

Would be an interesting challenge imo

[obsidian crescent]

I think overly focusing on efficiency would be too much. But I think there's a pretty big transitional period where it would be really reasonable to use higher quality beacons with speed&eff modules that cause no extra power consumption coupled with prods in the big buildings.

Once you get more power, you can speed up your builds by just swapping out the efficiency modules in the beacons.

[jaunty citrus]

Yes overly focusing on efficiency would be the point to make it a sort of challenge. I would not consider it optimal given how "free" power is after a certain point.

That transitional period you mention might exist seems quite interesting though.

[frank warren]

that period at least exists in some places, maybe indefinitely (space platform)

[jaunty citrus]

Yes I expect it to be extremely dependent on the surface you're building on.

[wary beacon]

we need a youtube tier list of best items to have quality early on for, surprised i haven't seen that clickbait title yet on reddit

[jaunty citrus]

You will once the expansion releases I reckon

[obsidian crescent]

It's kinda hard to have that a priori, since it depends on many factors: how easy it is to get quality versions of that item, how useful those quality versions will be, when you'll be able to get them, etc.

For example, the purple science precursors of electric furnaces and prod 1s are good candidates due to their ease of manufacture if you're getting purple science (which you undoubtedly will). But even then, you're not saying that "prod 1s are the best item to get quality versions of". You're saying "quality prod 1s are pretty cheap early game."

Though putting some Q3 prod 1s into your first rocket silo is kind of a must...

[wary beacon]

early on it will be things you are naturally making, and items that just slot in naturally in your bp's and are just better and having more doesn't hurt, and will naturally benefit you.

[jaunty citrus]

Q3 prod 2s are about the same as Q1 prod 3s no?
9.6% vs 10%?

[wary beacon]

beacons, solar panels and labs all come from you just 'playing the game', and every uncommon/rare one that pops out is real nice, and then when you launch your first platform you have a decent base of upgraded equipment for babys first platform

[jaunty citrus]

Yeah you've got the right idea there

Even if you don't want to use solar to power your base, you'll still make a bunch for the satellites you need initially. Make some quality ones as well to use on your space platform

Something I realise I haven't considered is upgrading steel furnaces to electric furnaces is my started smelters, mostly because I don't want them to be 1.5x as long.
But with the 30% speed bonus from Q2 that negates the need to expand it as much. I'll have to consider that when I get to it.

[winter saffron]

On that note: Any news what's going to happen to eff. modules?

They already max out at 80%

[obsidian crescent]

That's probably not going to change. With all of the power-sucking machines and lower-power environments in SA, being able to hit that -80% cap while also using prods and/or speeds is going to be very useful indeed.

The more interesting issue is that most other modules need 2 quality levels to match the bonus from the next highest tier. But efficiency modules don't work that way; a Q2 eff1 is almost as good as a Q1 eff2. And a Q3 eff1 is almost as good as a Q1 eff3.

[jaunty citrus]

As far as I know efficiency is staying the same, but quality actually makes efficiency modules far more appealing. With higher quality modules only the benefits scale, the negatives don't. This makes the relative effectiveness of efficiency modules drastically increase.

E.g. one cancels out ~71% of a power increase, but one cancels out ~178% of one , legendary or not.

[obsidian crescent]

It also gives you a reason not to immediately switch to higher tier modules. I could see some setups just using higher quality prod 2s instead of trying to upgrade to prod 3s ASAP.

[jaunty citrus]

I'm definitely interested in utilizing quality like this early on. Regular prod mods are fine for most things like furnace stacks, but higher quality ones would be much better in places like rocket silos, processing units and science. That and the new beacons being more viable early on, it'll be awesome to play around with.

[iron root]

this is something interesting to consider, because if you go the quality ore route, you definitely get material in a way that it's viable to upgrade everything to quality versions
but it's going to be much more rate-limited than building materials, to the point that the time it'd take to kit a whole factory in might take long enough that youll want to be slotting lower tier modules to get their benefits here and now

cause if youre going the quality ore route, you wont ever even be making modules unless its for science (and in that case, youre not making t2/t3 modules)

[unborn flax]

Meow. How would you make it better? It's split into two halves (levels 0-10 and levels 10-25) otherwise it's not clear what's happening at high levels of bonus prod.

[jaunty citrus]

It's just the bottom label, I can see how it was misinterpreted. I would make it clear it's talking about the productivity % instead of just mentioning research.

[unborn flax]

Things do turn out cleanly (with almost straight lines) if you take logs of the y axis; this will look better for some people and worse for others

[jaunty citrus]

Oh that actually shows the interesting relationship between quality and productivity there.
You can see that the lower productivity in the EMP outweighs the productivity in the asm3 for a while due to having a significantly higher quality upgrade chance.
But as you increase the productivity, the asm3 with prod mods approaches 300% faster, asymptoticly increasing the number of quality rolls you get with the same amount of materials.
The rolls you get approaches infinite at 300%, at which point chance to upgrade quality doesn't matter as it will upgrade eventually without loosing material.

[unborn flax]

Also the kinks in the lines (e.g. the asm3 one between 90% and 100%) represent strategy changes; in this case going from P3Q1 to P4Q0

[frank warren]

I’m very surprised nobody has analyzed how much stronger a quality setup is. For many items I have found that their cost of being upgraded is virtually free.

We’ve been comparing the costs - but this is pretending like a legendary item in the same. In truth you need less of them.

[jaunty citrus]

It is an interesting comparison. Going by raw ore costs, an costs around 6x as much as an while only being 2/3rds better
If an uncommon could be considered "free" depending on how you make it while being 3/10ths better.
A legendary one costing ~20-160x (depending on how you make it) while being 2.5x better (more than ~10x better when combined with other legendary items) isn't that ridiculous considering possible production scaling in space age.

[frank warren]

A legendary setup overall will cost about 15x cost. It’s >12x better by my rough math. Has to be done on a case by case basis.

The napkin math however is that quality is not expensive at all. We were totally wrong about it.

[fallen urchin]

Faster speed can be achieved by placing more machines, but higher qual have better prod

[jaunty citrus]

You should still consider that it's only "not expensive" once you can take advantage of max quality items, scaling to get those items will either be expensive, take a long time or be somewhere in between.

Probably not a super long time, and I wouldn't think it would be very expensive, but not as cheap as it winds up being once your quality engine is fully up and running.

[frank warren]

It’s a lot less than we expect though. One you have your first quality quality modules you’re off to the races.

[frank warren]

I have made more breakthroughs - but first I must ask the chat a trick question.
How much more expensive is a compared to a ? Assuming you start with .

[jaunty citrus]

Ballpark guess 60x instead of 80x while looping because you get the upgrade step from crafting the plates into gears so you're starting with ~25% of your gears being higher quality.

[frank warren]

(starting with base iron)

Have another guess?

[jaunty citrus]

idk, it depends on what outcome you want I suppose.

[frank warren]

wdym exactly?

[jaunty citrus]

Are you okay with sending the normal gears to make something else or do you want to recycle those into legendary as well? would cut the cost down a lot. Not sure how much I haven't done the math.

Problem with the question is that the answer depends on the methods you use, and I don't know what is optimal.

[frank warren]

All of everything must become a .

It's a black box, iron goes in, gears come out. The question is how much more iron do we have to put in this box to get a legendary gear out compared to getting a normal gear out.

[burnt reef]

also depends on how much prod research you have. the obvious assumption is "none", but the research required to unlock is probably more expensive than at least the first level of gear prod research, so why wouldn't you have it by that point?

[frank warren]

Let's assume no gear prod.

[frozen fiber]

Gear prod research? Is that a thing?

[frank warren]

Not to our knowledge.

[burnt reef]

maybe! we don't know yet.

[jaunty citrus]

From common knowledge, it's just shy of 80x the input of normal items to legendary items.

Although that assumes you're starting from the item and recycling first I think, with the extra quality step starting from the ingredients it might be less?

[frank warren]

That common knowledge is correct, but is not correct for the question I asked 😉

That assumes the alternative is using no modules at all to make basic gears.

It costs 80x more than making basic gears with 0% prod...

Or, prased another way, a legendary gear requires 80 inputs.

[jaunty citrus]

80 normal gears, yes

[frank warren]

not quite.

You could have made those gears with prod instead.

80 normal gear-inputs.

That makes 160 gears actually, best case. (100% prod)

So it is actually 160x more expensive than a basic gear.

[jaunty citrus]

But that's not the modules used in the recipe for optimal recycling, you use two prods and two quals

[frank warren]

correct.

I am comparing it to making basic gears.

A basic gear does not cost 1 basic gear of ingredients. it costs 0.5 basic gears of ingredients.

[jaunty citrus]

Yes, which would have 50% more productivity than the first step in legendary gear production.

or 33% more item output

[frank warren]

That leads to a 160x cost difference between the two.

Gear costs 2 iron.
A legendary gear costs 80 * 2 = 160 iron
A gear made with max prod costs 1 iron.

[jaunty citrus]

80 * 4/3 = 106.7x

[frank warren]

this is much simpler - I have no idea where your math goes wrong - probably with one of them being looped and one not being looped, but this is the correct analysis.

[jaunty citrus]

I mean I assumed that we were always dealing with normal inputs of the product itself when talking about the 80x number. In which case this just doesn't follow.
If that isn't the case then it reframes the costs of certain items and makes looping non-prodable items more appealing in comparison to prodable ones.

[teal wren]

I wonder how you get legendary plastic. Just, you need a ton of additional petrolium?

Well, that is not a big problem I think

Just still needs to be crafted normally

[frank warren]

It is the case, but it’s 80x more expensive than handcrafting basically. We have cheaper ways to make basic quality stuff so the real difference in costs is much larger.

A gear can be made for 1 iron, a legendary gear, 160 iron.

[teal wren]

That's another reason why fluids are infinite

[jaunty citrus]

Where are you getting the 160 number from

If you start from gears, recycling them into legendary would get to 80 normal per legendary.
If you start at plates, that's an extra step you can use quality modules instead of prod modules, making it cheaper, not more expensive

[teal wren]

How assembler with both qual and prod modules will function when fed legendary plates?

[jaunty citrus]

Unless I'm misunderstanding where the 80x number comes from in the first place

[teal wren]

As I get quality modules turn into prod ones when fed legendary stuff??

[fallen urchin]

Do you take into account that qual can increase by more than 1 tier at a time?

[teal wren]

No, the case when quality modules give you more legendary items if you already craft from legendaries

How that works, I didn't researched that mechanic

[quick rampart]

When a speed module is inserted into a machine, it gives a negative quality bonus, so will the product decrease its quality then? (i guess no)

[jaunty citrus]

Quality cannot decrease, thankfully

Well it kinda can if you mix qualities together in an assembler, which isn't reccomended

[iron root]

it only counters positive quality from quality mods yeah

[frank warren]

I believe the 80x number is the cost of ingredients -> end product, but the difference of one more loop is minimal I believe anyhow. The way I am getting that number is by comparing the cost to the altnerative cost of making a gear, and not the material cost. See this list of costs:

Cost of gear (no prod): 2 iron plates
Cost of gear (max prod): 1 iron plate
Cost of gear (optimal legendary): 160 iron plates

[frank warren]

A quality loop will cost 80x more than a machine with no modules, but 160x more than a machine full of prod modules.

[hoary current]

I think the point is, we forgot that when we make normal quality products, we can have full prod which makes the cost be 0.5x instead of 1x

[frank warren]

yes exactly

[hoary current]

so 0.5x relative to 80x is 160x

[frank warren]

Not that 80 is a relevant number though - almost all will flow through the EMP route

[jaunty citrus]

beacons and modules being such a significant cost of every late game build yeah

[frank warren]

But the principle of comparison here works nonetheless. - the cost difference is not the difference from the base recipe, but the difference from the best other way of making it.

[jaunty citrus]

Yes, which is why I assumed the 80x number was from product inputs, not ingredient inputs

In which case this doesn't apply.

[frank warren]

It still applies because you had to craft them, it’s like replacing the first step in the loop with full prod. The difference in cost remains the same I believe. I can see how that makes it way more in the weeds

[jaunty citrus]

It wouldn't because then you'd be counting the bonus prod in both methods

I'm going to need to run the calcs again before I can be happy with the numbers now

[unborn flax]

My tool produces output like

--------------------------------------------------------------------------------
free prod 0%, 4 modules: input (1.0 x normal) -> 0.0125190042245584 legendary outputs (reciprocal 79.8785575963231)
legendary outputs are 159.757115192646 times the expense of a normal output```

as both 79.9 and 159.8 are useful depending on what you're thinking about. [I think of 1 input as being '1 set of input ingredients' to avoid complexity from what the recipes are - so '1 input' means 2 plates in the iron gears recipe.]

[jaunty citrus]

Okay I've redone some numbers (extremely poorly) and I'm happy to agree with you @frank warren. The 80x number is from inputs.
This does recontextualise these optimised intermediate loops. Although I've been moved in recent weeks to caring less about the cost anyway due to the immense production scales that seem possible with space age and quality.

[frank warren]

Well, I'm doing some math right now for how much more output you get out of a prod setup, because I realized everyone is calculating the cost very wrongly.

First, we should measure the cost difference from our best alternative route, not handcrafting (this is where 160x comes in)

Secondly, we should normalized this cost based on the benefit it gives us - if quality items are 100x better, but 10x more expensive, they are actually 10x cheaper.

[jaunty citrus]

It's easy to think about one legendary inserter/assembler/etc. costing 200x a regular inserter or something and dismissing quality as too expensive, but I don't think I've ever put effort into mass manufacturing anything other than modules and science.

A typical hub making ~2 assemblers per minute is fine for almost all my use cases. Bumping that up to 200 assemblers per minute sounds like a cool challenge tbh.

[frank warren]

I'm currently working on a calculator for the benefits of quality, and I am still getting very favorable numbers. A legendary setup should still cost < 3x what a normal one does given this metric.

[jaunty citrus]

It's hard to put a number to it due to fairly forgettable benefits like lower power consumption or benefits that don't translate nicely between quality levels like productivity

[frank warren]

My 2x number is not accounting for decreased input values.

At the very top the value easily goes sub one. (legendary being cheaper than normal)

For a one beacon EMP setup, I currently have a legendary setup outputting 12.7x more than the same setup but normal quality.

[jaunty citrus]

Yes, the decreased input requirements for the same output provide benefits outside of the setup, so now you have 2 different metrics, the relative cost of the setup compared to normal setups for the same output, and how quickly that cost sees ROI with higher quality prod mods.

[frank warren]

Yeah, the downstream effects aren't possible to directly measure because it depends how you make those downstream things.

It's a stronger effect when your downstream production is lower quality.

This means when you're first getting started, quality at the very top is probably far beyond free.

It's likely in the order of 2-3x cheaper I would guess. (than base quality)

So this idea that quality is something you need to wait for at all is not neccesarily true.

[jaunty citrus]

Getting high quality into expensive science production will have massive benefits while requiring very few high quality items for sure

[frank warren]

It suggests that you could recycle some of your downstream production and turn it into a higher quality end machines, and end up with more output.

[jaunty citrus]

replacing 10 assemblers with 3 fitted with modules, beacons, etc. would probably make more while consuming less input

[frank warren]

yes, but the fact that it will go sub 1 means that if you removed the now 'extra input' - that is more than enough material to pay for it 🤯

Basically negative cost quality upgrades; you end up with more than you started.

[jaunty citrus]

reframing it this way has big implications for speedruns, right?

[frank warren]

It means on resource counts alone, you will have to do quality

We thought before it was a way to scale, but this suggests that it's not a more expensive high density option, but cheaper.

That said, it's still a lot of hassle to actually make the items.

So we can't really be sure if they will actually use it to any appreciable degree, and say, recycle loop - but it's at least in the search space now I think.

If making quality was as simple as paying for it, they would 100% use it in speedruns and recycle loop a lot.

[jaunty citrus]

With all the possible strategies discussed I expect something to come out on top as "best"

Even if it isn't necessarily

[frank warren]

Interesting fact is that quality becomes more impactful the more speed modules you apply, but speed modules themselves become less impactful.

[jaunty citrus]

Intuitively I feel it would be pretty easy to stick some quality modules into module 1 production and some circuit assemblers to make some quality T2 modules, which are very good even in small quantities

[frank warren]

I feel like they'll do that for sure.

This math suggest there's room for full fledged recycling though.- that surprised me.

[jaunty citrus]

Speed modules are easy to get more of their effect, just use more beacons. Within a limit of course.

I'm more interested in quality prod and qual modules, maybe some eff modules for a space platform.

[frank warren]

that is not what my math is suggesting. They multiply with everything else.

[jaunty citrus]

Making me want to go to fulgora first just to get my hands on a recycler

[frank warren]

Fulgora is really looking good. Reminder that these cost calculations rely on the EMP route for resources.

So it's not just the recycler.

[jaunty citrus]

I'm agreeing with you, speed is powerful with quality as it multiplies other qualities, but you can also substitute higher quality speed modules with more speed beacons.

[frank warren]

And that is somehow more expensive...

I believe that is because of the quality beacons as well.

Legendary beacons with legendary speed are generally a bit cheaper than base beacons with base speed, when you account for it all.

[jaunty citrus]

Yeah getting the EMP for cheaper circuits/modules/beacons and the recycler for dealing with stockpile buildup seems really nice for playing around with quality more...

[frank warren]

If you want to lean into quality, you want to instantly get fulgora setup to make Q mods.

But also, I think this is why they restricted the other qualities behind some sort of unlock.

You would instantly go for legendary following the math. It is optimal.

[jaunty citrus]

Oh yeah fulgora is also where you get Qual 3s

That sort of makes it a no brainer if you want to play around with quality

[frank warren]

Yeah, and dramatically cheaper modules directly. Reminder that EMP doesn't just make the circuits. It makes the modules too.

[jaunty citrus]

That's a decent point, getting a lucky legendary beacon right off the bat would be pretty nuts

Same with equipment, especially power armour

[frank warren]

It's not even luck. This suggests that you'd be better off to explicitly recycle for it.

IE if you are building a new line, it is cheaper to build a small legnedary line than a large normal one (in resource cost)

[jaunty citrus]

That's the math for when you already have a legendary recycling setup though

[frank warren]

that is frankly absurd.

[frank warren]

Expensive, but not so expensive.

It's a problem like kovarex

[jaunty citrus]

So if you somehow got lucky with some early legendaries it would skip so much progression

[frank warren]

in the bootstrap phase? yes.

[jaunty citrus]

Yeah it would be like if mining uranium ore had a 0.001% chance to give you 40 U235, except far wider reaching.

[frank warren]

'and u235 could be crafted into prod 3s'

somehow still wider than that

this thing looks realistic to me

play with this if you want to see how much better legendary setups are

edit: added automatic beacon calculation https://www.desmos.com/calculator/twvyxezlwu

Most of our setups are 12-13x better when all legendary. This is roughly the base cost via the EMP as well, so in general, legendary setups have a cost premium of the EMP's prod - so 1.75x cost premium.

They generally require 0.75x input compared to a normal setup.

So, in conclusion, any time that your input machines cost 2.33x more than your current machines, upgrading to legendary is break-even.

I would say for the majority of the top of the crafting chain, this is the case.

[fallen urchin]

I wonder whether the speedruns will actually take the hassle of starting the quality engine

[quick rampart]

only the 100% achievement runners maybe

[solar osprey]

Quality looping: no. Using quality: yes.

Or maybe actually they will.

Because a quality looping machine will only require going to fulgora, then building some assemblers, recyclers and some belts.

And once you’ve built one, you can blueprint it and then copy it for everything else you want.

[fallen urchin]

"Starting the quality engine" means making enough and to cheapen items, so it certainly needs looping

[solar osprey]

And a device that makes quality quality modules can be blueprinted and used for quality other modules.

Similarly a machine that makes quality assembler 3s can be copied to make quality electric furnaces.

While I don’t think speed runner would be making a bespoke quality loop for every item, or running a quality cascade through an entire crafting process, a generic quality looper that’s built once, copied and used everywhere else would probably be useful.

[heavy ember]

Assuming the recipes for every module stays the same

[solar osprey]

Feed in common modules, recycle them, feed the recycled ingredients into assemblers.

[half raptor]

A lot of our speculation is based on 1.1 recipes. I wouldn't be surprised if mk3 modules require planet specific resources.

Speed makes sense for Gleba since you have resources that have an expiration

[heavy ember]

Using speed at gleba makes sense.

Having it require, say, vulcanus to make would be :fun:

[half raptor]

Might just be 1.1 recipes, but it would be an interesting opportunity to include a bit of Holmium/Tungsten/Green Goop

[heavy ember]

Honestly gleba making materials for eff modules makes more sense to me

[obsidian crescent]

... why though? Speed modules on most planets are just more buildings with less stuff. But on Gleba, they're can do something that "more buildings" couldn't do: have less stuff. A larger setup takes up more space and thus can't make ingredients as freshly as a smaller one. "Less stuff" means more on Gleba than on any other planet.

[heavy ember]

You're talking about usage of modules.
I was talking about making the module and the ingredients for it

I agree: speed module usage on gleba makes a lot of sense

[obsidian crescent]

What ingredient for modules would suggest efficiency rather than speed? I mean, Fulgora's qual module 3 uses superconducting wire, and that doesn't really suggest "quality" to me.

[hoary current]

Nauvis making eff modules from makes a lot of sense for me

are basically mini fusion reactor

[obsidian crescent]

It almost certainly won't be from UFCs. It would likely be from some new uranium derivative (uranium needs more uses if it's going to be Nauvis's special thing). The "bluranium" could be an RTG intermediate which gets used to make eff3s as well as some PFR-like device.

[heavy ember]

That's the energy part.
But what about the pollution?
Bio reactor that eats pollution and generates electricity.
Also aren't superconductive wires part of superconductive magnets which are used in devices like MRI? So quality module could be a sort of high tech imaging to find defects

[opal haven]

that makes quite a lot of sense, and decent reason to visit

[obsidian crescent]

I'm not sure I understand the pollution question. Nothing has been confirmed on whether pollution is a problem on Gleba, and there's certainly been no statement about "Bio reactor that eats pollution and generates electricity". So I don't know where any of that is coming from.

From all of the (admittedly quite limited) evidence we have, speed is a much more natural fit for Gleba's mechanics than efficiency. Just as productivity is a natural fit for Vulcanus, whose primary building is all about getting stuff cheaper. And quality is a natural fit for Fulgora which is all about recycling stuff (and has a building that makes module recycling way more cost efficient).

Also, if there is a mechanic of pollution being a problem for growing things (thus encouraging you to put your farms farther from your main base and thereby stretching your freshness logistics), it should be noted that the Biochamber is at least partially a burner device. It may not produce pollution at all. But the first step in any of the fruit-processing we've seen are not Biochamber processes: red mash is done at an assembler, and baked pinecones come from a furnace.

So it could be that the Biochamber itself is quite eco-friendly; efficiency modules would be about reducing fuel consumption.

[heavy ember]

I meant the pollution reducing aspect of the modules.

[obsidian crescent]

Yes, but why does that matter specifically on Gleba? To the extent that non-Nauvis planets have enemies that key off of pollution, the pollution-reducing effects of efficiency modules will always matter. So why would they be of particular importance on Gleba?

Molten metal processing introduces an additional step in the refining process which allows prod modules to be even more effective. The recycler makes producing quality goods a matter of resource throughput rather than some carefully calculated process, also giving you an additional step that can use quality modules. In both of those cases, the modules have particular interactions that wouldn't exist without the special buildings/mechanics on those planets.

How do Gleba buildings or mechanics interact uniquely with efficiency?

[half raptor]

I still think it should be speed, not only because we need to handle products in a speedy manner but also just seems like the plants might keep things fast and lubricated

[heavy ember]

What I'm trying to say is tgat adding gleba product to a speed module doesn't make sense to me thematically.

Also why do you need to have the module that uses hleba specific stuff to be hyper relevant on gleba?

[half raptor]

It's just a shower thought. They haven't actually said mk3 modules will change their recipe. Odds are it'll all be

[obsidian crescent]

I don't think we know enough about "gleba products" to say one way or another. We have no idea what any of these things are or do, nor do we know all of Gleba's intermediates.

[kindred crater]

I like to think is just a special exception because of what it can do

[obsidian crescent]

As to the design question, you can design a game however you like, but it's usually a good idea to make the design make sense.

Prod modules on Vulcanus make sense, as its mechanics make iron/copper processing more productive (X ore in, 2.25*X metal out, minimum). Quality modules on Fulgora make sense, as the recycler really opens the door to hardcore quality usage.

You don't have to design this way. They could stick the efficiency module 3 on Gleba. But it wouldn't be as good of a game design fit as speed modules.

The main reason I feel that efficiency module 3s are on Nauvis is that efficiency modules get increasingly less important the farther you get into the game. Power constraints matter early, but slowly bleed away as you trivialize enemies and can expand at will. Space platforms will always care about power, but the other planets increasingly won't care.

So if eff3s are on Gleba, then a player who goes to Gleba last is likely to pick up a module that's basically worthless. Maybe Aquilo will have extreme power constraints that necessitates its usage. But nobody's going to be thinking "Great, I just got eff3s. Now I can do something cool!"

[kindred crater]

aquilo as the speed module planet... playing devils advocate for a moment, I definitely think thats the most fitting T3 module for the final planet. you still get T1-T2 before that, beacons are buffed, and of course there is , both in the modules themselves and in the buildings for faster crafting speeds. so still plenty of speed boosts mid-mid late game

even if it doesn't require some special material like holmium, just having it as an unlock there (with the same recipe as 1.1) might make sense

[heavy ember]

No matter where you put eff3:s the uselessness of them doesn't change.

Personally I think it would be good game design on going to one planet to get something that makes a second planet easier, which has an item that makes a third planet easier.

It would connect them together atleast somewhat and push people away from routing everything though a single planet

[obsidian crescent]

The utility of eff3s change over time. There is a point early on where their utility is high, and it goes down with time. That's why it should be on Nauvis, to maximize their usage at high utility.

[heavy ember]

That's eff:1s and maybe eff:2s
Eff3 is too expensive to be be worth it when pollution is a consideration when making them, which it might not be on gleba

[obsidian crescent]

You shouldn't think of efficiency modules as a pollution control mechanism. First and foremost, they are a power control mechanism.

The main issue with eff3s as the currently known numbers stand is that quality beats them. A Q3 eff1 is almost as good as an eff3. So if eff3s are going to be useful, they need a buff.

[heavy ember]

And unless we're massively limited on power generation, that power just isn't a factor.

Unless you have a machine that eats hundreds of MW before modules, I don't see using eff modules to reduce power usage.

To me they're a pollution reduction primarily.

Exception might be the space platforms due to how cramped they are

But unless a single eff module can remove the same amount of resources needed to generate that much power, it's just not worth it.

[obsidian crescent]

Both Vulcanus and Fulgora start off "limited on power generation". Being able to run 5 Foundries for the price of 1 is a game changer for your initial Vulcanus start. You can use that calcite for more lava processing, which means you can switch over to solar sooner. Your infrastructure can be better and tighter. You can climb the tech tree faster because you're not throwing calcite into a neutralization plant.

[kindred crater]

I don't see the issue personally with them, they have their niche use

as for lategame/postgame, maybe their primary use will just remain on space platforms

[obsidian crescent]

Fulgora is a similar story. Until you get Mk2 collectors, lightning isn't likely to be a great power source, so you're going to have to be careful about how you use it. And recyclers and EMPs are not cheap buildings to run.

But as each planet matures, you'll rely on efficiency modules less and less.

[heavy ember]

Do I need to remind you that eff1:s can hit the cap and eff2:s can hit it with one less module

You don't need eff3:s to make meaningful progress at the start

Nor can you if the module is unlocked at the planet

[frank warren]

‘can hit the cap’ misses the point because other modules effectively raise the cap.

[kindred crater]

thats when you can combine speed/prod with them and cancel out or largely reduce their power demands

[heavy ember]

If fulgora and vulcanus are as power starved as you say, I would rather have gleba be the place for the eff3:s so I can get them sorted before I go to fulgora/vulcanus.

Not hammer my head through a wall without enough power to get the good stuff, only to not need it anymore

The latter sounds like worse design to me

[kindred crater]

a big part of the game is increasingly more efficient/effective technologies and builds as you progress. nothing is stopping you from just building more power, but you just pay for it with footprint size, pollution, related infrastructure, fuel demands, etc

[obsidian crescent]

I'm not sure I understand. You're saying that you can see the justification for earlier eff3s, but you use that as justification for gating them on Gleba? That doesn't make sense. If it's good for the player to get them early, if those modules are at their most useful early on... then just give it to them early on.

[heavy ember]

If eff3:s are made on fulgora, that doesn't solve vulcanus issue and vice versa.
Gating them away from one of the planets is just as big of an issue

All that said my original argument was thematic and yours was gameplay.
I still don't see thematically a better option than gleba materials for eff3:s

[obsidian crescent]

And I don't understand how efficiency modules fit with anything thematically on Gleba as has been revealed to us. Is power a particular problem on Gleba? Is pollution a special problem on Gleba different from other planets? Besides the fact that plants are involved on Gleba, what is the theme that links Gleba to efficiency?

[kindred crater]

I'm open to the possibility of pollution on gleba affecting fruit grow speed/harvest rate

[heavy ember]

Like I said earlier: advanced organic material that could plausibly make a bio reactor which feeds on pollution to make power

Thus the reason why efficiency modules reduce power usage/pollution amounts

[obsidian crescent]

So the reason you think Gleba has a theme of efficiency is because of a machine and process that is purely speculative? A thing for which there is not one hint from the developers that it is in the game?

[kindred crater]

in all fairness, we know far less about gleba than any of the other starter planets

[obsidian crescent]

I get that we don't know as much about Gleba as we would like. But you can't take that unknown and say "that's where the efficiency theme is". Especially since the current knowns pretty consistently point to a theme of speed.

For another example of that theme, Gleba has small-but-rich mineral patches. Patches that would definitely benefit in throughput from speed modules. Wouldn't it be nice to have an even faster speed module for those?

We don't know much about Gleba. But what we do know is just not pointing in that direction.

[kindred crater]

in case I haven't made it clear, I've never said my guesses are the way the game will be. I just enjoy speculating based on what we already know. mostly it turns out wrong, but sometimes I've been proven right. and based on what we know so far, assuming planets each have their own T3 module unlock, seems like the place for it. if for nothing else than the fact its a green lush planet, plants on earth don't grow well in polluted areas, and efficiency modules reduce that.

[frank warren]

I see zero reason why Eff. modules wouldn't be on Nauvis if we follow the 'each planet has it's own' idea - It's clearly the weakest module, and also most useful early on. Think gameplay instead of thematics, and it seems to be the obvious conclusion.

[heavy ember]

I give up. Everything goes back to gameplay reasons within few sentences

[kindred crater]

I understand that, but we still have T1-T2 efficiency there before even leaving the planet, not to mention uncommon/rare quality. Going back to my guess earlier #1215078107334057984 message , maybe there is some more pressing reason to use them on gleba, such as combining them with speed/prod while still keeping close to 20% power/pollution

[frank warren]

I do agree with you that on thematics, Gleba seems like a good fit.

[frank warren]

I can see this being fair.

We know that in some cases they use rails on Gleba, and that the team generally loves rails.

There's no good reason for stuff to be far apart unless.. you have to make it far away

[kindred crater]

and that plays in to the spoilage challenge as well

either you have lower fruit yields and have your processing machines closeby, or you have orchards on the outskirts, and you must design a rail network to quickly transport goods to your processing far enough away from it

[heavy ember]

Further more: if orchard require a specific conditions to function and you have more than one type, you might need to transport longer distances

[burnt reef]

"lower fruit yields" can be solved by overbuilding unless the yield approaches zero. it'd be interesting if polluted trees gave you partially spoiled fruit.

[obsidian crescent]

As previously mentioned, efficiency modules are useful on every planet at the start. The only thing this might change is why. On Vulcanus and Fulgora, it's about power consumption. On Gleba, it could be about pollution control. But even then, the best control measure is distance.

There's also the issue that faster consumption of highly fresh nutrients may result in less pollution generated than slower consumption of less fresh nutrients.

[obsidian crescent]

If pollution is going to be a factor, I doubt it would reduce freshness. The reason being seeds.

Mashing even nearly spoiled fruit still returns seeds (and presumably seeds don't spoil). And it would be a more interesting challenge if pollution could make it difficult to get enough seed stock to plant the next crop.

That being said, there seems to be a simple way to trivialize that pollution problem: don't harvest fruit from some farms. Basically use some farms as a pollution barrier by just not letting the Ag tower harvest stuff. Those trees will (presumably?) absorb pollution, shielding those behind them from the pollution spread.

[kindred crater]

if thats the case, I could see their pollution absorption nerfed vs normal trees, just so that isn't as viable of a strategy

[obsidian crescent]

In any case, we should probably move this to the main SA channel instead of the quality thread.

[quick rampart]

i think that explains why epic substations cover 24x24 while legendary ones cover 28x28

interesting that legendary is 2 tiers higher than epic

[obsidian crescent]

That's how it's always been. Q5 is a double-bonus. That's why it goes 100%, 130%, 160%, 190%, 250%. Internally, base quality is Q0, then there's Q1, Q2, Q3 and Q5, with Q4 being skipped.

[heavy ember]

The wanted legendary to be more special

[obsidian crescent]

That's how Q5 provides a double-bonus.

[quick rampart]

that's great, i initially thought it was like tier 0, 1, 2, 3, 4

[obsidian crescent]

We will still call it Q1-5, but internally, it's [0-3] and 5.

[half raptor]

Just say and make sure to enunciate the dots when you speak

[quick rampart]

five dots substation

[jaunty citrus]

There are 3 parts to quality that give the final number.
The quality bonus defined in the item. e.g. 30% speed bonus for assemblers, +1 tile radius for power poles, +0.2 transmission efficiency for beacons, etc.
The quality levels, which each item is able to be upgraded to in order. e.g. normal, uncommon, rare, epic, legendary.
The quality power, which is defined independently for each quality level. e.g. 0, 1, 2, 3, 5 for normal, uncommon, rare, epic and legendary respectively.

I should clarify that this is what I know from dev comments on this discord.

[half raptor]

And for mundane items like chests and belts, quality increases hp

[burnt reef]

so quality is 1-dimensional? there's no way for different tiers of quality to give different kinds of bonuses?

[quick rampart]

that would be a coding nightmare

[half raptor]

Also a balancing nightmare. Instead of different kinds of bonuses they give the same bonuses just in larger values with each tier

[fallen urchin]

The problem with eff3 is much more severe than that

Even if it's unlocked for free, it still sucks at its intended usage: speed/eff mix

[frank warren]

not on space platform

[fallen urchin]

In a speed/eff mix, 1 speed3 = 1.66 speed2 + 0.75 eff2, but 1 eff3 = 1.25 eff2

Upgrading speed2 to speed3 is much more important than upgrading to eff3, even if you're aiming for max eff

[daring siren]

And then we get fusion reactors that reduce the power issues

[real crown]

@little orchid the problem i have with quality is one of game design philosophy

i am of the belief there should be no amount of randomness in factorio

[gloomy torrent]

Ok the expansion of biters is random should we now remove them?

[daring siren]

There is randomness in the map generation

[real crown]

you guys are nitpicking what im saying

there should be no randomness in the actual factory part of factorio

[jaunty citrus]

Why though?

[little orchid]

Law of large numbers successfully eliminates randomness and turns it into ratios

[daring siren]

We had these discussions in the past, and it just takes a while to see it's not a big issue

[real crown]

because i do not believe it leads to a fun experience

[daring siren]

Have you played with it?

[little orchid]

But it’s pretty much functionally the same as a non-random implementation

[real crown]

none of us have played with it

[jaunty citrus]

Not really true

[eternal verge]

Janky quality (BETA)

Implements a janky version of Factorio Space Age quality

Author

Soul-Burn

Downloads

1893

[real crown]

and also, we are working off of numbers we dont even know

[little orchid]

We do, actually

[jaunty citrus]

Yeah we know the numbers

[daring siren]

We have played with a close analogue to the real thing

We know almost all the numbers

[real crown]

does it use the actual numbers space age will use

[daring siren]

As far as we know, yes

[real crown]

confirmed by a developer?

[little orchid]

By information provided in the fff

[daring siren]

Streamers like ColonelWill have some nice huge megabases using quality

[jaunty citrus]

A quality module 3 gives a 2.5% quality chance, this goes up to 6.2% at legendary

[daring siren]

Also confirmed by dev after fff yes

[little orchid]

Oh cool

[daring siren]

But the numbers aren't the issue for you, but rather the whole idea

[gloomy torrent]

Just admit you hate quality because you can't build big

And are scared that random numbers will not play out

[jaunty citrus]

Chill man

[daring siren]

And the mod is janky, giving even more "random" results, and you can still build around it

[little orchid]

I just think summers is severely underestimating the law of large numbers, and there’s not much else to it

[daring siren]

Lets keep discussion civil, no need for ad homs

[real crown]

yes, i dont believe that randomness should be a part of a factory in a game design sense

as it adds a level of uncertainty previously foreign to the experience, which i believe is good that it was foreign

[gloomy torrent]

So if there was an implementation that had several progress bars, and at the end of each one you would be guaranteed a certain rarity item, you would be satisfied?

[daring siren]

Quality has 2 phases: early game skimming of resources (big numbers) and later game recycling (big numbers)

[little orchid]

In practice it’s simply too small a level to make any impact

[swift sedge]

Do you dislike uranium processing now? That's random, and it adds an interesting complication

[daring siren]

There are methods to reign in that randomness. I find them quite interesting!

[real crown]

no, because uranium is insanely simple

its a pratically infinite resource that you just leave it there and it will make enough uranium for 1 reactor

[jaunty citrus]

Building for quality early on is a trap, you're better off letting some intermediary assemblers with quality modules run in the background to stockpile some ingredients you can use to make quality machines.
You can also put the modules into mall products and get a few higher quality machines that way as well to use in space and power constrained environments like space platforms.

[gloomy torrent]

And quality isn't???
It outputs the same exact resource by its uses, just that the final output will be inherently better at its job

[swift sedge]

It's the same.. percentage chance of different product ..

[real crown]

can you let me fucking finish saying my thoughts please

[jaunty citrus]

Mind you you don't need to do this, it would just be slightly "better" if you wanted to do it.

[real crown]

and yet kovarex completely removes the randomness

also lets not act like uranium and quality are on the same level

uranium is a very low level mechanic, 1 centrifuge will always make enough enriched for 1 reactor

meanwhile, i might be waiting for hours for 1 legendary tier 3 assembler, even with dozens of recyclers running and tier 3 quality modules

depending on the numbers

[little orchid]

Ok they’re really not that low

[jaunty citrus]

I mean you won't be waiting hours for a legendary assembler, I expect you to have significantly higher production by the time you unlock legendary quality

[swift sedge]

I think it's more similar than you think. You won't be waiting for hours

[daring siren]

Legendary is a big feat

[jaunty citrus]

Legendary is also not the only level that has major benefits

[daring siren]

You're not expected to have any of those for a long time, and definitely not a lot of them

Even just is great

[little orchid]

There’s just as much “uncertainty” here as there is with quality. It’s statistically possible for the ore processing to get bad rolls long enough for your reactor to turn off, but you don’t seem to consider a possibility

[jaunty citrus]

if 12.5% of the modules you make are or better (you've been to fulgora and got the T3 modules and EMP), thats great! you can use them in the most resource intensive recipes like science and rocket parts to great benefit!

[real crown]

so is speed 2s or prod 1s, but almost everybody uses speed 3s and prod 3s because they are the best

[daring siren]

Well time to change that line of thought

[real crown]

just because you dont need the best, doesnt mean the best should be a pain in the ass to get

[gloomy torrent]

But those are expensive, so are high quality machines

[daring siren]

It's exactly to give you a longer progression

You're not expected to switch to once you get it

[jaunty citrus]

I would encourage you to try using the lower tier modules as well, they're getting a significant boon in 2.0 with the beacon changes imo

[daring siren]

High quality is for post win runs

[real crown]

by artificially padding out the playtime, due to the fact quality takes longer than making speed 3s or prod 3s?

[little orchid]

1 centrifuge will “always” be enough for a reactor in the exact same way a quantity of materials will “always” be enough to make a certain quality item

[daring siren]

It's not artificial. It's a progression. Something to work with.

[jaunty citrus]

quality takes no more time if you just realise that using and stuff is also really useful.

Might as well use the quality modules when you're making mall items to get some that are better!

[real crown]

except uranium doesnt cost me processing units or flying robot frames

yknow, valuable resources that cost alot

[jaunty citrus]

Personally I'm excited for the soft-requirement to make hundreds of times more mall items to feed into loops, I've never had to make a full belt of assemblers before

[real crown]

yeah, for good reason

[little orchid]

I don’t get what you mean. Do you want buildings to be free?

[real crown]

i dont want a single assembler 3 to cost 8 thousand speed 1 modules

[jaunty citrus]

Well good thing it doesnt lol

[real crown]

bad thing that it will with quality

[gloomy torrent]

I will ask once again, do you want the results to be made with progress bars as in productivity?

[jaunty citrus]

It just won't

[little orchid]

Well, if it’s a really good assembler, it’s only fair for it to be expensive. I think this is pretty standard. Thing is good therefore thing is hard to get

[real crown]

no because that defeats the entire purpose of quality

its just a mechanic that if its fixed, isnt random, and if it is random, its terrible

[gloomy torrent]

🤯

[real crown]

thats why it should not be in the game

[jaunty citrus]

The math has been done, quality items just aren't as expensive as you think they are, and it's not like everything you place down has to be legendary quality

[real crown]

post the damn stats then

[little orchid]

You could have just asked sooner if you were interested…

[real crown]

what are the actual chances of a legendary item being produced

[jaunty citrus]

10% if you feed an assembler with epic ingredients

[real crown]

if a legendary assember 3 is 100x as expensive as a regular assembler 3, then i expect it to be 100x better

[jaunty citrus]

that just isn't even how the game works right now

[real crown]

and how likely are you to get epic ingredients

[daring siren]

Lets say there was only and , would that be fine?

[little orchid]

Yeah this

[jaunty citrus]

same thing but one tier lower

[daring siren]

And then someone adds and as a mod

[jaunty citrus]

thats with an assembler using quality 3s though

[real crown]

so its 10% of 10% of 10% of 10%?

is that correct

[jaunty citrus]

If you use an emp with higher quality it's a way higher chance

goes up to 31%

[little orchid]

Like you said, it’s not quite as simple as uranium. This is a relatively complex mechanic, there’s some nuance to it

[gloomy torrent]

The point of quality is that you have 3 paths instead of 1:
Do things the regular way
Touch the quality, and make some parts of your factory more powerful if you need, like faster assemblers or more productivity on iron
Or make the shiny legendary factory that is quite literally the best at it's job, but also hella expensive. By the laws of large numbers, if you try enough, you will get everything, not within days, but hours, for the entire thing.

[jaunty citrus]

And it's not as simple as percentage of percentage of percentage either

[real crown]

so with legendary tier 3 quality modules, its 31% of 31% of 31% of 31% to get ONE legendary item

[real crown]

then what is it?

[gloomy torrent]

If your inputs are common

[jaunty citrus]

It's not like the other items don't exist, you still get them just at lower quality

[real crown]

why cant you just post the math equation

[gloomy torrent]

If they are uncommon at least, you're 3x more likely to get legendary

If we go by simple math

[jaunty citrus]

If you have an "ideal" setup, a legendary item can cost only 13x what a common item does

Technically that's not true, the perfect setup actually has 1x cost, but the 13x number doesn't includ prod research

[daring siren]

You're expected to recycle at that stage

[jaunty citrus]

13x requires you to be using all legendary modules in an emp and recycler

and for the item you're making to accept prod modules

[daring siren]

With the huge prod bonuses, it becomes reasonable

[real crown]

which you are gonna take 40 hours alone to get but go off

[jaunty citrus]

It's expected for the average player to take 80 hours to beat space age so sure

[daring siren]

Fwiw you're not expected to have a full in 80 hours, just beat the main game

[jaunty citrus]

But hey, you could always just build more production and it'll take less time

Just like everything else in the game

[real crown]

yeah, so whats the purpose of quality existing

[little orchid]

Vertical growth as a supplement to the standard horizontal growth, without horrendous recipe bloat

[jaunty citrus]

So that you can have verticle growth while playing the game, and get even crazier scaling in the post game

[sour iron]

plus wouldn't you have built a number of intermediate quality machines and factories before even unlocking legendary?

thereby knocking down that 40 hours significantly?

[real crown]

if you are willing to afk for hours so that you get enough ledgendaries to make a setup

[gloomy torrent]

Who said we need to

[jaunty citrus]

Or you could spend your time building more produciton instead of afking

[gloomy torrent]

Quality is not for blueprints

[little orchid]

You can also go afk for hours in 1.1. Except people don’t do that, they just make their factories bigger and faster

[jaunty citrus]

I could make a 10spm base in 1.1 and beat the game by afking, but that wouldn't be very fun

[gloomy torrent]

Noone said you can't just slowly upgrade the factory

[real crown]

yeah, which returns me to questioning what is the purpose of quality if its better to just make a bigger factory

[jaunty citrus]

You know what's better than a bigger factory?

A higher quality bigger factory

[little orchid]

A tall factory

[real crown]

wasting my fucking time praying to god that i will get legendary modules?

[daring siren]

To make a bigger factory.

[jaunty citrus]

Or just building the necessary setups to know you'll get one

[real crown]

speed 3s and prod 3s already gave you enough horizontal scaling to begin with

[daring siren]

The factory must grow and all that

For you

Many of us want more

[sour iron]

again, ramdom numbers become statistics at a point. so it's not wasted time when you know the ratios

[little orchid]

You fool, in factorio the word “enough” has ceased to exist long ago

[real crown]

at the cost of precious time?

[little orchid]

Just make it work faster

[jaunty citrus]

You could probably say the same about fusion reactors. Why add them to the game when nuclear reactors do fine enough, just build more of them?

[daring siren]

Yes. Playing games wastes time and gives enjoyment.

[little orchid]

I think this is a pretty good analogy

[real crown]

id rather wait 5 hours knowing that ill have 500 speed 3s than wait 5 hours not knowing if ill have 50 or 5 legendary speed 3s

[gloomy torrent]

Higher quality machines work faster, making you loose less time

Even at uncommon quality, mind you

[daring siren]

You're extremely likely to have around that number. Law of big numbers.

[jaunty citrus]

Why are you waiting 5 hours instead of playing the game

[little orchid]

It all keeps coming back to the law of large numbers… You will never have a range like “5 or 50”, you will have ranges like “498 or 506”

[real crown]

that is simply not true as was stated explicitly in the FFF, where nuclear reactors suffer from the fact shipping in water to the space platform

fusion reactors fill that niche of not needing water

[jaunty citrus]

And regular Factorio suffers from not having a 3rd dimension of scaling

[gloomy torrent]

Who said that

[real crown]

thats literally a argument AGAINST quality

[jaunty citrus]

????

[gloomy torrent]

Do you know the ingredients of coolant and energy cells?

[daring siren]

????

[sour iron]

no it isn't?

[jaunty citrus]

How, just play the game lol

Build your recycling loop and continue playing the game

[daring siren]

Expand your factory to produce more stuff

[jaunty citrus]

no one will force you to wait for the legendaries to appear before playing more

[real crown]

the FFF posted TODAY

[sour iron]

then, as you obtain intermediate quality machines, decide where best to use them

[gloomy torrent]

Okay but where did it specifically say that fusion does not need water?

[sour iron]

y'know, the act of playing factorio

[real crown]

i want to say such vile things to you right now

[little orchid]

How did we switch topics to the water requirements of fusion reactors

I thought we were talking about quality

[real crown]

you guys are seeing things in the theoretical

[jaunty citrus]

When it had a picture of it that doesn't have water inputs

https://cdn.factorio.com/assets/blog-sync/fff-420-fusion-generator-alt.png

[real crown]

of the "just go play the game while waiting for the legendaries" meanwhile im seeing the practical of how you already have to wait for fuckin speed 3s to be made so that you can make your best setups

[jaunty citrus]

And a diagram with no water

https://cdn.factorio.com/assets/blog-sync/fff-420-fusion-prototype-1.png

[gloomy torrent]

There's no productivity for outputs other than intermediates or science, therefore i would rather have faster machines than make them faster(as in, making more machines)

[gloomy torrent]

Yeah but there are inputs of coolant and energy cells, what are those made of?

[little orchid]

I think you are looking at the theoretical more than we are. You’re looking at the unlikely possibility of getting incredibly bad rolls many times in a row just because it’s statistically possible. But in practice, you will pretty much never have such bad luck that it really impacts the state of your game

[daring siren]

Guys, this is the quality channel. I suggest you move to #friday-facts or #space-age

[real crown]

even if its statistically improbable, you are already banking on the statistically improbable nature of getting good rolls

the stupid fucking law of big numbers you love so much

[rotund egret]

it's statistically probable*

[daring siren]

Law Of Big Numbers

[sour iron]

why is the law of big numbers stupid exactly?

really what is the issue with it and how it works?

[real crown]

it doesnt matter if its 1/1000 chance to not get a single legendary machine in 5 hours, that should just never be possible in a game about building factories that will always output the same amount of items no matter what as long as the starting conditions are the same

[jaunty citrus]

Try one in (not going to happen within the existence of the universe)

[daring siren]

Then build more factory

[rotund egret]

"should"

[sour iron]

why? i don't see why random chance shouldnt play a part in the game

[real crown]

thats NOT THE FUCKING PROBLEM IM SAYING

[little orchid]

Instead of always outputting the same items, they “always” output the same items. It’s really not such a big difference in practice

[real crown]

THE PROBLEM IS THAT RANDOM CHANCE EXISTS

[daring siren]

The larger your number, the lower the probability to get nothing.

[real crown]

NOT THAT ITS NEVER GONNA HAPPEN

[sour iron]

random chance is a new genre of problem to solve

[rotund egret]

yeah, random chance does exist, that is a fact

[sour iron]

and it is solvable

is the thing

[rotund egret]

well actually

[daring siren]

I know statistics are hard to grasp, but they work.

[rotund egret]

#true_randomness

[gloomy torrent]

Just admitted to always getting hard pity in gacha games

[daring siren]

There's a chance a cosmic ray flips a bit that causes your factory to die.

[gloomy torrent]

Just say that you want the hard pity in factorio if you play with quality

[jaunty citrus]

In that sense it's possible for someone to build a uranium processing facility and not get 40 U235 in 2 years. Does it happen? No.

[jaunty citrus]

Can you not

[rotund egret]

@real crown you can make use of quality by cycling low intermediates like ore and then crafting stuff with 100% chance of quality output

[real crown]

i hope, i pray to god jesus buddha every fucking deity under the sun that quality is not as bad as i am saying it will be, but i know for a fact it will

[sour iron]

how do you know that?

[real crown]

i would genuinely commit seppuku before playing a gacha game

[rotund egret]

it was revealed in a dream

[daring siren]

tbf that's why hard pity exists in gacha games. But Factorio is not gacha. It trusts the players understand statistics well enough to play it.

[sour iron]

you knkw leople have been running these statistics for months now right?

[gloomy torrent]

But isn't that what we get to hear here? Summer hates the fact that there are chances involved and they may need to spend far too long because odds don't play out

[real crown]

game design shows how a game will play out

[sour iron]

quality isn't an unknown by now

[daring siren]

ColWill's base is almost fully

[jaunty citrus]

Has there ever been a case of someone being unable to get kovarex enrichment setup because they didn't get 40 U235 in an engame save?

[daring siren]

And only took like 200 hours, without EMP and Foundry

[rotund egret]

yes (I don't use nuclear)

[real crown]

you people are all either just not reading what im saying or are serially incapable of connecting my messages together to understand my greater point

[rotund egret]

yes

we're all wrong

[sour iron]

no thats just it

[rotund egret]

and maliciously misinterpreting you

[sour iron]

we do get what your saying

[jaunty citrus]

In that case it will be very good, as the factorio dev team are pretty good at designing their game. They seem to enjoy playing with quality

[sour iron]

it's just that you're off base

[daring siren]

People who played with the Janky mod enjoyed it

[sour iron]

it's just a base disagreeance

[gloomy torrent]

So your greater point is that there exists a chance that even with enough production, you will not get the quality, and that's not factorio-like because if you make something, it is guaranteed to produce predictable results, am i right?

Yes or no

[jaunty citrus]

We do get what you're saying, I'm connecting your messages together, but I just can't agree that you're making a bigger point than disliking probability

[real crown]

and i do believe the devs are some of the best game designers in the industry, but clearly sometimes even geniuses make mistakes, and they have very clearly made a huge mistake

[gloomy torrent]

If no explain why

[sour iron]

in that randomness in factorio is straight up, not bad

[daring siren]

So it's a disagreement between one person who doesn't like it (which is fine) and dozens who like it and even played with it.

[real crown]

yes, factorio should at a BASE level always be predictable

2 copper wire and 1 iron plate will ALWAYS make a green circuit no matter what

[sour iron]

and that's just it

[jaunty citrus]

Quality is very predictable, it tells you exactly what you should expect...

[sour iron]

it is

[rotund egret]

since we all like it and you don't, the "very clearly" is debatable since it's not unanimous

[gloomy torrent]

Here's the point: quality is a deviation from the base

[charred harbor]

since an FFF said quality was one of the earliest and most play-tested feature, im hopeful!

[sour iron]

even with quality

[daring siren]

Quality is an expansion on that base

[sour iron]

roll 2d6 10000 times and i will be able to very accurately tell you what you shpuld expect to get

[daring siren]

P.S. do you think the same about recycling scrap on ? It's even more random, and required.

[real crown]

there shouldnt be a single significant mechanic that deviates from such a base affirmation that 2 copper wire and 1 iron plate will always make a green circuit and all follows from there

[sour iron]

why?

[rotund egret]

2 uncommon green wires and 1 uncommon iron plate will 100% make an uncommon green circuit

[gloomy torrent]

Here's another kick: 2 copper wires and 1 plate will still make a green circuit

[sour iron]

what if deviation is interesting and fun?

[jaunty citrus]

https://cdn.factorio.com/assets/blog-sync/fff-399-scrap-recycling-recipe.png

[gloomy torrent]

You just have a possibility for it to be more useful

[daring siren]

Nothing changes base Factorio. This is the Space Age expansion, there to bring new things.

[real crown]

because that is not a factor in any other significant recipe in the game

[jaunty citrus]

Um....

[real crown]

i do not believe that deviation and relying on chances makes for a good experience

[gloomy torrent]

Another thing is that quality does not lock you out from anything, you can still make 50k spm gigabase, no matter how bad your luck is

[jaunty citrus]

Have you read the Fulgora FFF?

[rotund egret]

you are entitled to your opinion.

saying it's objectively bad when most folks disagree is going to result in most folks disagreeing with you

[jaunty citrus]

[real crown]

yes, but also, holmium ore is entirely possible of acquiring outside of scrap

so it is not luck based

[final nimbus]

2 copper wires and an iron plate won't make anything.
You need 3 wires.

[real crown]

shut the fuck up

[rotund egret]

where'd you get that one from?

[real crown]

thats not what i meant

[jaunty citrus]

No, it isn't

[real crown]

you are all so fucking dense its crazy

what part of "randomness goes against the design philosophy of the game" do you not seem to understand

[jaunty citrus]

Scrap is one of two "natural" resources on Fulgora, the other is heavy oil.

[gloomy torrent]

Point out the fff, and show us where you see the holmium ore be generated in the map, or confirmed to be generating outside scrap

[jaunty citrus]

The part where I just don't think that's true at all

[real crown]

i will literally not buy space age if so many parts of it are luck based

[daring siren]

Buy it for 🙂

But yes, so many parts are luck based

[final nimbus]

🧌 spotted

[real crown]

i might as well be playing a gacha game if its so depedant on luck

[rotund egret]

tungsten, baby

[jaunty citrus]

Yeah """luck""" based

[sour iron]

the part where you're insiting that you know the design philosophy of the game more than the people who are making the game, playtesting the game, and playing mods to replicate the quality mechanics

[daring siren]

Luck can be reigned in

[real crown]

this is not going anywhere

[sour iron]

no it's not, you're not doing a good job of convincing us

[jaunty citrus]

If I roll a dice 12 thousand times I'm not lucky if I get a six two thousand times

[sour iron]

why might that be?

[daring siren]

Lets finish this conversation amicably

[real crown]

you all just refuse to hear what i am saying and just keep saying that its fun because you can just wait 5 hours for 1 legendary assembler 3

[jaunty citrus]

You can, or you can just build a setup that will make it faster!

[real crown]

are you stupid

[jaunty citrus]

Nothing stopping you from waiting 10 hours for 10 T3 modules in 1.1

[daring siren]

We're saying this is not how we're going to play, as it doesn't sound fun.

[sour iron]

you know you're not the first person to make this argument right?

[final nimbus]

Do you build a single assembler and complain that it takes days to launch a rocket?

[sour iron]

it's already been refuted

[rotund egret]

Am I wrong to say you're being a bit rude by calling folks names? That's a bad quality™️ interaction

[sour iron]

we know we arent going to be doing that

[jaunty citrus]

No, you're just projecting at this point. At no point have I said that it's a good idea or fun to wait for slow production instead of playing the game and increasing production.

[real crown]

yeah and im guessing all of you acted the same to all those people who had valid points

[sour iron]

the people playing the game with quality are not doing that

[real crown]

im going back to what i was doing since you all are incapable of doing any sort of interpretation

[little orchid]

People have already responded to that point…

[sour iron]

but it's not a valid point!

we have SEEN proof to the opposite

[obsidian crescent]

The problem is this: you don't have to "wait 5 hours for 1 legendary assembler 3". You can just do quality cycling, and it will output "legendary assembler 3s" at a regular schedule. The problem with your argument is that you're ignoring how the mechanic works.

If you describe the mechanic in a way that is not how it actually works, then it's hard to take your position as valid.

Like, I understand in the abstract that you personally do not like the idea of random outputs being a significant part of a factory game. That's fine. But if you're using that to say that nobody can design a factory that consistently produces quality goods, you're just incorrect.

[gloomy torrent]

Their entire argument is "i believe that given enough people there will be an unlucky one who will have an extremely bad luck getting any good quality stuff, and that may be me"

[final nimbus]

I have never understood this argument people keep making that randomness contradicts the DeSiGn PhIlOsOpHy

[jaunty citrus]

Claiming you don't like probabilistic recipes is fine. Going on to claim that it is "bad game design", "the worst mistake the devs have made", "will force you to wait 5 hours" is just absurd

[little orchid]

I think they actually left, feel free to stop echochambering now 😭

[uncut helm]

if they are so out of luck they may play for 50 hours and not find any other iron ore patch except the one they started with

[jaunty citrus]

JG spitting facts

[obsidian crescent]

Even the "bad game design" or "worst mistake" is at least a personal opinion. I don't like the whole "force you to wait 5 hours" bit because it is just objectively wrong.

[gusty trench]

Has anyone brought up that space platform resource collection is also random, and we don't even get numbers for it? The amount and type of asteroids we get for any given location is not assured

[final nimbus]

Relying on exaggerations is a clue that someone is dishonest.

I see this all the time.

[daring siren]

Random things:

• Uranium
• Quality
• Fulgora scrap
• Asteroid collection

[final nimbus]

Well apparently they just hate randomness and refuse to think of ways to deal with it.

[uncut helm]

biter nests, ore patches, cliffs, almost everything connected to map generation

[gloomy torrent]

"oh but that's not about the factory part of the game" was their argument

I can scroll up and look for that, they said it for real

[uncut helm]

uranium processing is the same randomness the quality is

you put stuff in machine and get different product depending on luck

but i kind of understand the difference they try to point out

you can treat uranium processing and scrap recycling like a kinder suprise you need to open to see whats there. while the quality stuff and recycling is a bit different. you are relaying on stuff you made yourself

[kindred crater]

why are we spending this much breath on the kneejerk quality reactionaries still

[gusty trench]

Because someone is wrong on the internet and that can't be allowed

[kindred crater]

let's have the truly important debate

[gloomy torrent]

The better question than "why quality is bad" would be "how it would be played if focused from the early midgame"

[jaunty citrus]

Because the fewer people with a misunderstanding of the mechanic the better.

[kindred crater]

how stupid the quality names continue to be

[jaunty citrus]

They're whatever

They work and no majority can agree on a better set

[kindred crater]

wow i produced an epic inserter

the biters will write epics about this one

[jaunty citrus]

Hell yeah! I love me an epic inserter

[kindred crater]

it can't be epic it is a mere cog in the machine of my vast empire

the factory is the only thing allowed to be legendary

[jaunty citrus]

Every piece of my empire is epic

[kindred crater]

no your empire as a whole is epic

[uncut helm]

#1090704808509444176 message

[jaunty citrus]

it's what make the factory legendary

[kindred crater]

oh my god it all makes sense now

this is why you haven't been talking about the enemies

the biter libraries and biter movie theaters aren't finished yet

[gloomy torrent]

I honestly get the unfactoriness of your average rpg rarity for the quality of components, and feel that letting players edit the names would be the best

Like, some easy-to-access file where you can change the display name of certain quality

[kindred crater]

oh that would actually be a great compromise

[jaunty citrus]

It'll be easily modable. Day 0 mod: "Custom quality names"

Just change the names in the mod startup settings

[kindred crater]

me playing multiplayer with my friends and all of us refusing to call the quality tiers by the base game names or the names the others chose client side

so i ask someone if they have any superior assembly machine 3s and they have to struggle to remember which quality that is on my end

(just to fuck with them)

if any of you name your anything other than flawless i will be very disappointed in you

[jaunty citrus]

I think I'll name mine

[kindred crater]

if the custom names mod lets you use icons like how you can in train stops and chat and such then you could actually name it

[uncut helm]

flawless is a bad quality name

[kindred crater]

why

[uncut helm]

5hp remaning wall is far from being flawless

[kindred crater]

it has been analyzed for flaws in its construction and none were found

[kindred crater]

5hp remaining wall is far from being legendary

[uncut helm]

not really

[jaunty citrus]

I think anything the factorio engineer makes being described as flawless would just be lying

[gloomy torrent]

There was no flaw to let it break from taking that much damage, pretty good name imo

[uncut helm]

5hp remaning means there is some actuall story behind this entity

[jaunty citrus]

I mean there's no way this cobbled together reactor doesn't have flaws

[sour iron]

a cheap/low quality item irl could still be flawless

[uncut helm]

95% spoiled agricultural science pack is also far from being flawless

[sour iron]

a toothpick could be flawless

[kindred crater]

by factorio engineer's blueprint it has no flaws

well most things mass produced irl are all

[sour iron]

why is why i dont really think flawless is a great descriptor

a cheap hatchback and a supercar could both be made flawlessly

[uncut helm]

common quality is flawfull

[gloomy torrent]

But can we all agree on calling the "best"?

[jaunty citrus]

For the same reason any other descriptor wouldn't feel great, it requires interpretation

[kindred crater]

yes everything in factorio can be made ly

[kindred crater]

too informal, dorky factorio engineer would never

maybe the quality tiers should just be named based off how effectively it warcrimes

like

[gloomy torrent]

"Murderous Solar Panel"

"Wanted Dead or Alive Automation Science Pack"

[kindred crater]

deadly
killing
murderous
slaughtering
extincting

[daring siren]

I feelawful

[uncut helm]

slaughtering rapair pack

[kindred crater]

exactly

[gloomy torrent]

"Slaughtering Wall"

[kindred crater]

i was Joking

capital J

[daring siren]

grey
green
blue
purple
orange

[kindred crater]

can we get this pinned

[gloomy torrent]

Porpol borglar alarm

[uncut helm]

orange red logistic chest

[kindred crater]

it's perfect

[daring siren]

There

[uncut helm]

green red circuit

[gloomy torrent]

Blue green circuit

[kindred crater]

jg your problem can be solved by a mod that colors quality items by the quality icon's color

[uncut helm]

perfect now you wont be able to tell inserters from each other

[kindred crater]

yes exactly

i was about to make a joke about an

[gloomy torrent]

Time to place burner inserters at train stops instead of stack ones

[kindred crater]

what about a mod that just picks randomly from #1263831220756942938

so you never know what to expect the qualities to be called

[trail cradle]

In all seriousness: QI, QII, QIII, QIV, QV is prob what I'll actually call them XD

[warped obsidian]

Why not just 1-5?

[gloomy torrent]

That's roman 1-5

[warped obsidian]

Yeah ik why not just regular 1-5

[gloomy torrent]

And objectively cooler way to call them that

[kindred crater]

come on you know better than that it should be QIIV not QIII

[gloomy torrent]

And tis QIIII and not QIV

[kindred crater]

no roman numeral should be repeated 3 times in a row i don't care what society tells you

[warped obsidian]

QCCXXV

[warped obsidian]

They have to? They don't work otherwise?

[gloomy torrent]

Is that some youtubers name

[kindred crater]

where

show me a number where they have to

[warped obsidian]

I may be stupid

[kindred crater]

society normalizing III instead of IIV has been a disaster for the human race (/joke)

[warped obsidian]

Real

[gloomy torrent]

That's some QIIIII() opinion you got there

[warped obsidian]

You've reinvented tally marks

Base 1

[kindred crater]

even tally marks isn't that stupid IIII IIII II

[warped obsidian]

What's your profile picture from?

[kindred crater]

https://homestuck.com

[trail cradle]

I just like how it looks for modules: Someone fetch me a T3QV prod. module!

[kindred crater]

well you'll at least make it QIIV right? right?

[half raptor]

is it just me.... or is calling something legendary not that hard?

even when I'm on vc with friends, we use the terms when planning out our base, like "when we get the research tier to unlock epic, make sure you adjust your factories to include the new tier prior to researching or you will get jammed up on the Rare production"

[jaunty citrus]

It isn't hard. It's just easy to point at being silly as well.

Honestly I see it as a good thing that this is a common topic. Clearly the game has few glaring issues.

[swift sedge]

Yeah terms have def. grown on me. It's simple and recognizable to anyone who's played an RPG. No confusion over the progression.

[half raptor]

once you set up your automation to deal with quality you will probably not care about as those will be taken by the system and sorted, shipped, and crafted until only and remain

for the naming, you can call it Q5MK3 Prod module, but it is more awkward to say verbally.

embrace the legen-dairy terminology and be grateful this game is on Steam and not

[fallen urchin]

Green blue logistic chest and blue green logistic chest

Or green blue inserter and blue green inserter

Qual logistic chests are useless, but qual inserters will actually show up in a real base

[kindred crater]

only if you're megabasing

lower qualities have plenty of use before that

is just so expensive

obviously you will still want some things before that

[half raptor]

for sure. In the early game right when unlocks, I plan to put them on machines on my bus and have only the normal intermediates go back onto the bus while I use a filter splitter to store any I get into a storage chest. When it comes time to start making my personal equipment and first space platform I'll probably use those stored up rares to make them

even normal mk1 quality modules have a 1% chance and on assembler mk 2, that's 2 module slots for a 2% chance, which is actually pretty good considerinig how many materials you'll probably make for your bus anyway

[kindred crater]

i will not settle for anything less than guns and armor

i must be ready to rip and tear when i get to bwuhuo

[half raptor]

rip bwuhuo

[kindred crater]

rawket lawnchaaa in one hand and concrete in the other

[obsidian crescent]

How many of which materials though? Assemblers, sure; you made those by the dozen. But oil refineries? You might make 30 before the mid-game. Pumpjacks? The same. Chemical plants? Do you make that many that early?

You have to pick things you make in bulk. Belts (for making quality labs), inserters (because green science), maybe stone furnaces (for eventual use in quality boilers). If you want the rest in any quantity, you'll need to do something else.

[half raptor]

while you have a chance to produce , you can also use your stored up uncommon to make them by default and even go for rares. Early game quality management might not be hyper automated and optimized, but storing quality products for later does give you a boost when making personal gear

[kindred crater]

they meant intermediaries like and

so that you can then make quality assemblers or such

[half raptor]

until I get access to higher than I'll probably let my base run on normal machines and stockpile quality ingredients for solar panels and more modules

[obsidian crescent]

But that's where prod modules go. I say put them in your miners, so you can siphon off quality ore and prod your way up from there.

[kindred crater]

maybe actually

hmm

in and then siphon off the quality plates into a chest

[half raptor]

here's my mentality, most machines require 3-5 steps from ore to make, modules can go as high as 8 steps if you look at it from the copper side. If you go for quality on ever step you can, your 10% chance for quality is actually the inverse of 1/(0.90^x) where x is the number of craft steps

[obsidian crescent]

Well, you get quality modules before electric furnaces, so I'd probably stick with those for a while. But even with electric furnaces, you only need to quality module the ones that get Q2 stuff, since you're capped at Q3 early-on.

[warped obsidian]

Let's go gambling!

[half raptor]

You can use your stored up to attempt for and even if a craft fails for quality, you at least end up with an uncommon red circuit

[obsidian crescent]

In a vacuum, that sounds great. But you now have to design a way of doing that that won't mix qualities of inputs (ie: make a green circuit from Q2 copper cable but Q1 iron plate). That adds a lot of complexity to your mall. To every step of your mall.

There's a big difference between having a large mall with a small mall and having a single highly complex and integrated mall.

[half raptor]

train interrupts and stations dedicated to certain qualities

more of a mid-late game idea thouugh

early game I'm just sortinig anything greater than to a chest to store and I'll manually load assemblers myself going for quality armors and guns

so the bus is still all normal ingredients and the mall is still makinig normal machines, but I can then take my stockpile of quality steel, copper, and circuits to try for some quality solar panels

[kindred crater]

no i just use runoff accrued over a long time

[obsidian crescent]

Also, how do you get quality iron plate if you don't start from quality ores? At least, before electric furnaces (which are blue science while qual modules are green)?

[half raptor]

oh see that's a super late game idea I have. All of my product train loading stations will use instead of to load trains. If a lands in a logistics storage chest, a robot will grab it up and take it to a bot mall that needs it. It'll mean that I'll rarely see my legendary intermediates accumulate, but it also means that my legendary mall will have access to any luckky rolls

electric furnaces AND foundries can have quality modules. Totally easy to get quality plates

[heavy ember]

The tech for modules might only need green science, but modules themselves need oil products so I would say it's reasonable to say that by the time you start thinking quality modules, you already have chem science automated.

[half raptor]

so my trains can still pick up all the varied plates I've made and are stores in the chests, but robots can see and access them if they need them for the mall. Means my product stations have both trains and bots competing. This means I'll have to make sure I put an OR condition on my train to roll if inactivity > 5 seconds

[gusty trench]

A splitter seems easier tbh

[half raptor]

absolutely. I will say though that if you're an experienced player you can design your early base to include quality filters long beforre you even make your first quality modules

if you build for the future then you can just simply slot in your modules when you get them

[obsidian crescent]

For myself, I tend to go for modules early, and for SA, I plan to beeline for qual modules. Even if only a handful.

[frank warren]

The problem is that once you hit late game, you are better off making dedicated quality setups I believe.

[half raptor]

yes, my mall will not be a long standing feature in my game

i meant bus

I will transition to a rail grid base because I find I enjoy the modularity and with stackable green belts and faster legendary machines you can squeeze a ton of production into each rail grid

[frank warren]

I believe this is optimal given what we've seen so far. My preliminary quality cost math was showing that higher quality items are actually cheaper than normal items in a lot of key places.

[half raptor]

However each planet will probably have a different strategy considering each one has different production goals and resources

by the time you go to space you'll have access to as well as so each planent will have different goals and strategies

[heavy ember]

I would say limited access to them

[half raptor]

sure, if you only have a handful, use them on the machines that run the MOST often

also people forget too that we can take modules out and move them around. One redditor suggested if he only had 20 he would set up chest buffers to feed each step and store the resultant, then move the modules to a new factory to process to the next step. Time consuming sure and not a great example of automation, but he was technically correct that he could reuse the same modules for the whole process

I could imagine someone making a system with recipes being set by circuit condition that could use this philosophy

[swift sedge]

Is qual multiplicative w productivity?

[frank warren]

to some degree yes

[swift sedge]

Hmm I wonder if it's better to get multiple rolls on the same input resources (e.g. roll on iron, on the plate, on the GC.. same iron rolled more than once) or to focus where you can get max productivity and let the productivity do the extra rolls

[half raptor]

the way prod CAN be multipliciative with qualit (in the late game) as you unlock special buildings that give +50% prod and you get legendary prod modules and you research levels of productivity bonus for intermediates, you approach the 300% prod bonus, or getting 4 items per craft. You then cycle those 4 items through a recycler with legendary quality modules and get back a quarter of the ingredients, which since you're recycling 4 items you get back enough (on average) to craft 1 item

this becomes a mostly neutral transaction that gives a chance for quality on every recycle step

(uncertainty for RNG)

[real crown]

@jaunty citrus @daring siren @gloomy torrent sorry for my actions earlier, i suffer from some pretty bad anger management issues and it got out of hand, and i regret being such an ass to you three in specific

[daring siren]

Hope you will enjoy the expansion even if it's not your cup of tea

[half raptor]

At least enjoy 2.0 which requires no purchase

[winter saffron]

Steel always recycles into iron, does it? Even if created from molten iron

[daring siren]

It always recycles to itself, never to iron

[real crown]

i will still purchase the expansion, but try to just not interact with quality

[open halo]

Quality was designed to be optional

[real crown]

ive spend way too much time playing this game to not play space age frankly

[open halo]

(though to play Space Age, you still need to activate it)

I think it's a really clever design decision

you can engage with it if you want to optimize your factory, but you don't have to

[daring siren]

If anything, the recycling is the more controversial part, as it is required to beat the game

[winter saffron]

. What about the foundry gears?

[daring siren]

and asteroid catching as well

[daring siren]

Recycles to iron plates 🙂

[solar osprey]

Iron plates.

[winter saffron]

weird

but alright I guess

Probably won't get the prod research tho.

[daring siren]

Because it's an assembly recipe

[solar osprey]

Well, no, as you can craft gears from iron plates in an assembler

[half raptor]

pretty sure all recycle back onto themselves at a 25% rate with your module affected chance for quality

[daring siren]

Remember recycling only returns 25% of the value. Think of it like cutting up the gears to find the parts of the gears that can be reused

[winter saffron]

turns into

[solar osprey]

No

[daring siren]

They do not

can turn into

[winter saffron]

Who was responsible for the wiki again?

Part of the upcoming expansion pack. Gives a 25% chance of returning items created from some types of recipe. Smelting and chemical processes cannot be reverted.[1]. For items that can't be reverted, the same item is returned instead with a 25% chance.[2] All fluids get lost on recycling.[3]

[half raptor]

connfirmed by devs. Plates turn into plates, period

[daring siren]

Smelting and chemical processes cannot be reverted.

Except that can be recycled

[winter saffron]

Ahh, I need to sleep

forget I wrote anything

[half raptor]

so steel stays steel, plastic stays plastic, but anything crafted like gears and cables can be recycled back into plates

[daring siren]

honk

[winter saffron]

Can't even read no more

[fallen urchin]

I think the randomness is not the main problem, as you will be dealing with them in millions, but they usually won't be giving items in the desired ratio

[daring siren]

LoBN doing work again

[fallen urchin]

LoBN does work with such big numbers, but it won't work with a wrong ratio

Suppose you need Fe/Cu in a 2:1 ratio, but the recycling gives 50%/50%, so you need either voiding or HUGE storages

I think it could be a good challenge to try optimizing

Void them if you don't want the hassle, but if you want to efficiently use those materials, then you need to think about what to craft with the overflowing resources

[jaunty citrus]

I think you'd be able to brute force a solution with train stop priorities.

If every item produced by scrap has a recycling stop with lowest priority, with it's outputs also having lowest priority, then everything will be consumed unless the ratios are off

[half raptor]

so here's 2 things to consder when dealing with inexact ratios due to quality. first you can set up overflow of items to go to a recycler. You can either build it locally or you can set up a central recycle yard for any and all excesses. Train interrupts will make it easy to deal with 40 different items as a result of mass recycling. Second if your excess of items made its way all the way up to Legendary quality, you can in fact use them in lieu of lower quality ingredients if you set the assembler to accept any quality item. This means if you wind up with too many you can set up an overflow train station at the plant and using priorty splitters use the excess legendary cable instead of the epic. You can do this all the way down to the normal circuit plant, though if you're still overflowing to these stations by then, you REALLY need to increase your production

trains are a heck of a balancing tool in this regard. If all of your demand stations are full and you have inactivity for 5-10 seconds, your train can decide to go to a recycler because you're over producing normal copper and you might as well try for

alternatively you can just replace the for anytime you get too many quality items jamming your system and just let it starve itself down some before changing it back. It is a loss of potential quality though

[merry spindle]

Anybody tried to calculate something about early quality?
I can define some stages that can be interesting:

• Nauvis (maximum modules are T2, "quality" max level I don't remember but not more then 3. No new assemblers)
• Fulgora (maximum Qmodule is T3, Pmodule is T2 "quality" is likely up to 4. EMP unlocked)
• Fulgora + Vulcanus (maximum Qmodule and Pmodule are T3, "quality" is likely up to 4. EMP and Foundry unlocked)
• Endgame (maximum modules are T3, "quality" is up to 5. all unlocked)

Last one was discussed a lot but early ones are likely more interesting for non-gigabase game

[obsidian crescent]

Really, I would say that the meaningful phases are "pre-recycler", "post-recycler" and "end-game".

Before you get the recycler, you generally can only get quality from some process for which the base quality product is being constantly or frequently consumed. Quality mining (putting quality modules in miners and skimming off the higher quality stuff) is a good technique for handling quality because you always have a use for the base quality ore.

Upsides:

1. You can make quality anything, since you're making them from raw materials. This is especially important for one-off items like personal weapons or armor.
2. It's simple to do.

Downsides:

1. You don't get all that much quality stuff. You get some, more if you're consuming a lot, but not a huge amount.
2. You need a separate infrastructure setup to make the intermediates you need.

[daring siren]

Also quality in mall items

[obsidian crescent]

Of course, even before the recycler, there are clever tricks. The inputs to green and purple science cannot be prodded, and they're actually quite useful. Quality belts can be used to make quality labs (and can be recycled for plate/gears when the time comes), and quality inserters can be used for quality fast/bulk/stack inserters later. Quality furnaces and prods are obviously extremely useful, and even when you have "enough", once you get the recycler, they can give you quality green and red circuits relatively cheaply.

[daring siren]

Even just a couple of quality assemblers and power poles are worth it

[obsidian crescent]

Also, there are a lot of infinite techs that use purple and green science, so you may as well do something with it.

[daring siren]

Quality pipes will eventually be recycled for a guaranteed 1/4 ratio for higher quality

[quick rampart]

are quality science packs more durable?

[frank warren]

We believe so.