Quality levels completely break factory

[TV4Fun]

So this has been talked around a little before, but I am starting a new thread because I haven't found any that discuss how completely it breaks your factory to start using quality modules without doing a complete redesign from the ground up and needing to basically build the same factory five times to engage with the quality system. After building an entire interplanetary factory empire without building a single quality I module, I decided to finally try engaging with the quality system in a pilot program on Fulgora, and I was totally unprepared for how broken it is. WHY do filters on splitters and inserters default to only applying to common quality items? WHY WHY WHY OH WHY can assemblers be set to only accept one quality of ingredient? I saw ingredient quality level on assembler recipes, and assumed that had to be a minimum, because, that was the only thing that made sense. I put quality module 3s on each of my scrap recyclers, so each one now outputs a random mixture of different quality levels. This completely broke my downstream filter setup, because contrary to what any logical person would think, an icon with just an item type and no quality level next to it means that filter only applies to quality 1 items and completely ignores higher quality items. And this is the default. Then I realized my belts were getting completely choked because none of my assemblers were accepting any of the higher quality ingredients because they can only accept Q1 items, and I can only pick ONE quality level for EVERY ingredient that assembler accepts. WHY WOULD YOU DESIGN IT THAT WAY????!?!??!?!??! I'm going to just have to remove all of the quality modules and go back to common quality on everything because I don't have the room to make a separate assembly line for every single quality level of every single ingredient. This exponentially complicates any factory design and I can't imagine it being worth it just for the little boost each higher quality item gets. What am I missing here?

[coffee-factorio]

You have to look for technologies that overlap and develop those.

The canonic hint, is to examine beacons, modules, and an assembler at 2 different qualities. The "mistake" is to examine them at legendary.

It is also better to attack the issue with massive economies of whatever you're approaching the problem with. Compared to Seablock which applies calculus, this is easier to get into but harder to operate. It tends to make people either over analyze the problem or over indulge in it.

[XT-248]

Quality breaks Gleba.

Yes, I know mentioning Gleba can be contentious at times, but this is not that kind of topic.


We all know that when we craft something of higher quality, we burn a lot of raw ingredients to nothingness and spoilage.

This is an example with made-up numbers. To craft 1k Gleba science packs at baseline quality, I need enough raw material for 1k science packs. At quality-2, I need at least the raw ingredients for 4k science packs to match the equivalent of 1k Gleba science value*. At quality-5, I need even more raw resources to keep the same science value.


Here are a few quality-related problems that do not apply to other non-Gleba science packs.

The time between crafting and then Gleba packs being consumed at the lab causes science value loss. A baseline science pack loses approximately 1% science value per minute, so 15 minutes of travel time means 15% loss per pack at quality-1. At quality-5, the same 15 minutes means about 60% science loss, or four times as fast per pack.

Waiting for a full rocket load for each quality level slightly decreases potential science value the longer packs sit in storage before being loaded into a rocket payload. I don't know if that has changed in the 2.1 experimental, as I haven't played that version yet. But I assume it will look for any quality and combine them into a rocket payload, which might help mitigate science value loss from waiting for a full payload. Assuming that's how it works in the 2.1 experimental version.

Dealing with mixed-quality logistics during production means those items might spoil before machines have enough to craft something.

A larger harvesting operation is needed to support higher-tier Gleba science pack production, which means more pentapod activity for only slightly lower science value than a non-quality Gleba science pack production line. Those tradeoffs make the higher-tier production line harder to justify.


TL;DR Summary: Don't craft Gleba past baseline quality. The return does not justify the significant investment required to up-tier Gleba science packs.

Don't get me wrong, there are viable scenarios in which higher-tier science packs work great. That said, up-tiering Gleba science packs needs a little tweaking so the value tradeoff is more favorable.


Gleba science value*: a quality-1 science pack gives 100% value, quality-2 gives 200%, and so on, up to quality-5 at 600%. Each quality level also has its own spoilage time, starting at quality-1 with 1 hour and up to quality-5 with 2.5 hours.

[NineNine]

What am I missing here?

It's fun for those of us who want more Factorio challenges. I love that quality makes me have to re-think my whole factory again. It's like getting yet another new version of Factorio.

[coffee-factorio]

Quality breaks Gleba.

Yes, I know mentioning Gleba can be contentious at times, but this is not that kind of topic.
....

This is an example with made-up numbers. To craft 1k Gleba science packs at baseline quality, I need enough raw material for 1k science packs. At quality-2, I need at least the raw ingredients for 4k science packs to match the equivalent of 1k Gleba science value*. At quality-5, I need even more raw resources to keep the same science value.

Gleba science value*: a quality-1 science pack gives 100% value, quality-2 gives 200%, and so on, up to quality-5 at 600%. Each quality level also has its own spoilage time, starting at quality-1 with 1 hour and up to quality-5 with 2.5 hours.

That was never the turn off for me. Because my assumption is that you do a quality burst and not do something like try and shred all undesirable quality. At that point I have a (3/4)*2*p bonus, where p is added productivity. And (2/4) * 2 * p bonus. So (1.25)*2*p in total. But that's in theory, in practice I can't scale it with speed without losing something.

Or I can just throw productivity in, and have a 1.666*2*p and solve more meaningful issues. Note that it is 1.666 because the ratio of max productivity is 2.5 to 1.5 in planetary machines where you will always make science. On Gleba something extra sneaky happens though, as you noticed: if you are willing to do multiple bursts you increase your resource demands on a level by level basis, including to seeds. So when we do bioflux and pentapod eggs, that's now a 1.5 bonus but our farms grow by 1.666 size. The last stage causes you to lose seeds too, which constrains your operation to use the standard ratio instead of the more productive one. I won't have an immediate effect on farms but it will make soils harder.

My experience with this was that it always looked beautiful until I desk checked it. Then it's like...
...why are you showing me these things that people seem to enjoy...

[Tertius]

What am I missing here?

I assume you're unable to make sense from the answers given so far, so I try to explain again.

What you tried was probably putting in some quality modules very early into your existing production chain, probably into mining drills and smelters, so you got all kind of qualities as ore and all kind of qualities as plates, and of course with all intermediates from that. And your items exploded by a factor of 5.

This is how quality works. Every item and every recipe has 5 versions, one normal, one uncommon, one rare, one epic, one legendary. Every item has actually 5 different varieties. In theory, you need to build every factory line 5 times. One for normal, one for uncommon and so on.

Actually, people don't do this. Instead, people do 2 different approaches with quality.
Approach 1: In earlier game stages, they put quality modules into selected machines with useful recipes. For example solar panels and accumulators: quality panels and accumulators have higher yield and capacity, so they're especially useful for space platforms. Most output is still normal, but the occasional quality item is useful for platform use.
Or they put quality modules somewhat early into the production chain of intermediate and siphon quality intermediates from the main belts full with normal items, with the goal to craft single higher quality end products.

Approach 2: for the end game and post end game, people build dedicated quality upcycling production lines.
These upcycling production facilities input normal quality ingredients, upcycle them to some target quality (for example legendary) and output items with the target quality only. Inside they deal with intermediate quality, but that's a blackbox if seen from the outside: input normal, output legendary.
Upcycling is crafting some item with quality modules, and if the resulting item isn't the target quality recycle (with quality modules as well) and craft again with the hopefully quality-upgraded ingerdients. This is a loop.
There are some key items and recipes to craft with this approach. For example circuits: The higher your processing unit productivity research, the cheaper the upcycling of processing units. That's one of the key recipes: it gives you red chips and green chips with high quality with a minimum amount of input. And casting low density structures: results in vast amounts of copper plates, as well as steel and plastic.
Some items can just be upcycled efficiently, for example electromagnetic plants and stack inserters, other items are crafted best by upcycling their ingredients with other means then craft the item itself.

Since you seem to be at the end game with everything unlocked and having stable production lines everywhere, I propose you try and explore approach 2.

[quineotio]

Actually, people don't do this. Instead, people do 2 different approaches with quality.

I put quality modules in miners and furnaces/foundries. You have to accommodate quality throughout the design of your entire factory, or at least at the miners and foundries. It's a way to essentially get free quality ingredients. It is a bit tedious to set up, and you can't do it until after Fulgora because you need recyclers to process any excess.

But then you can't really do much at all with quality until after recyclers, which is a big complaint I have with the quality system.

[XT-248]

That was never the turn off for me. Because my assumption is that you do a quality burst and not do something like try and shred all undesirable quality. At that point I have a (3/4)*2*p bonus, where p is added productivity. And (2/4) * 2 * p bonus. So (1.25)*2*p in total. But that's in theory, in practice I can't scale it with speed without losing something.

Or I can just throw productivity in, and have a 1.666*2*p and solve more meaningful issues. Note that it is 1.666 because the ratio of max productivity is 2.5 to 1.5 in planetary machines where you will always make science. On Gleba something extra sneaky happens though, as you noticed: if you are willing to do multiple bursts you increase your resource demands on a level by level basis, including to seeds. So when we do bioflux and pentapod eggs, that's now a 1.5 bonus but our farms grow by 1.666 size. The last stage causes you to lose seeds too, which constrains your operation to use the standard ratio instead of the more productive one. I won't have an immediate effect on farms but it will make soils harder.

My experience with this was that it always looked beautiful until I desk checked it. Then it's like...
...why are you showing me these things that people seem to enjoy...

It is not about whether something is a "turn off". It is about looking at different design choices and picking the better one that gives me, the player, the most return on my time and investment.

Quality fell a bit short compared to module-less designs on Gleba. Of course, productivity modules are a better choice than no modules at all. Even efficiency modules feel a bit unnecessary on Gleba since processing Bioflux creates batches of 60 nutrients (with the biochamber 50% bonus) without productivity modules, and at least it does something with minimum effort. Speed modules do improve the throughput at the cost of consuming a little extra nutrients.


We can compare different modules in biochambers or machines on Gleba all day; however, this thread is about quality mechanics and quality modules breaking factories.

[coffee-factorio]

It is not about whether something is a "turn off". It is about looking at different design choices and picking the better one that gives me, the player, the most return on my time and investment.

Quality fell a bit short compared to module-less designs on Gleba. Of course, productivity modules are a better choice than no modules at all. Even efficiency modules feel a bit unnecessary on Gleba since processing Bioflux creates batches of 60 nutrients (with the biochamber 50% bonus) without productivity modules, and at least it does something with minimum effort. Speed modules do improve the throughput at the cost of consuming a little extra nutrients.


We can compare different modules in biochambers or machines on Gleba all day; however, this thread is about quality mechanics and quality modules breaking factories.

If it is how I said it, I do not mean offense.
I feel I have to clarify, compared to productivity quality returns less science and seeds. In my facility this makes problems because I use those to make legendary overgrowth soil. I also was running a stealth build until 2.1, now I'm running a few more spidertrons. I do not disagree with your point of view. I see it differently, but arrive at a similar conclusion.

As for breaking things. There's a reason why I said you want to analyze upgrades at rare. A normal to legendary upgrade is going to do some very odd things to a conventional city block because of how they both cause the building count to shrink and the power use to drop.

[XT-248]

If it is how I said it, I do not mean offense.

Nothing wrong with what you said. It is just that the topic is about quality mechanics, and I don't want to drift too far off-topic.

I feel I have to clarify, compared to productivity quality returns less science and seeds. In my facility this makes problems because I use those to make legendary overgrowth soil. I also was running a stealth build until 2.1, now I'm running a few more spidertrons. I do not disagree with your point of view. I see it differently, but arrive at a similar conclusion.

No disagreement on quality modules yielding less science value per harvest than without quality modules. I explicitly mentioned in my previous post that having no modules is better than using quality modules for getting more science value.

I was curious if there was any extra bonus or effect from producing higher-quality overgrowth soils. I couldn't find any. Can you explain why you are making higher-quality overgrowth soils? It might help me understand what you are trying to do better.

It is less of a "stealth" build. It is more about getting more done while using less raw material. It happens to be good at being "stealthy," which might be the primary goal for some people, but not always.

I like finding unusual, more efficient designs that people might not think of.

As for breaking things. There's a reason why I said you want to analyze upgrades at rare. A normal to legendary upgrade is going to do some very odd things to a conventional city block because of how they both cause the building count to shrink and the power use to drop.

I will reiterate my point in a shorter form.

• No-quality build: Harvest enough material for 1k Gleba science packs to make 1k Gleba Science value worth of science packs.
• quality-2 build: Harvest enough material for 4k Gleba science packs to make 1k Gleba Science Value worth of science packs.

Even at an absolute minimum, moving to the next tier requires 4 times as much material. This is before analyzing quality-3 Gleba science packs, as you suggested I do.

What do I get in return for breaking the Gleba factory by going to the next tier? Fewer rocket launches? Meh. Rocket launches are very cheap. It is still better not to use quality modules, given the quality-2 analysis.


As an aside, I am interested in inspecting city-block designs, if you can find them, on Gleba because I don't think I've seen any.