Friday Facts #375 - Quality

[AirForce1]

I have no issue with this but there should not be so many qualities in vannila because its getting overcomplicated very fast. 3 qualities is more than enough to complicate this very much.

I can't disagree with you on the complexity part lol. Not sure if you're part of the Discord server, but since Friday, theres been endless discussion going over the numbers, making python scripts testing various configurations, different spreadsheets, etc.... Its a labyrinth if you want to get serious with it. But I guess depending on the person that is a good thing or a bad thing.

Discord is waste of time and lair of toxicity. Someone just kick me because I dont agree with something about this quality thing and someone just say that he dont agree with me and kick me... so yeah waste of time

Discord is bunch of henchmans of Wube and kick/ban everyone who dont agree with them. Its so great feedback to listen only someone who agree with you and supress everyone else.

[FuryoftheStars]

I have no issue with this but there should not be so many qualities in vannila because its getting overcomplicated very fast. 3 qualities is more than enough to complicate this very much.

I can't disagree with you on the complexity part lol. Not sure if you're part of the Discord server, but since Friday, theres been endless discussion going over the numbers, making python scripts testing various configurations, different spreadsheets, etc.... Its a labyrinth if you want to get serious with it. But I guess depending on the person that is a good thing or a bad thing.

Discord is waste of time and lair of toxicity. Someone just kick me because I dont agree with something about this quality thing and someone just say that he dont agree with me and kick me... so yeah waste of time

Discord is bunch of henchmans of Wube and kick/ban everyone who dont agree with them. Its so great feedback to listen only someone who agree with you and supress everyone else.

Well, if true, that would certainly explain why the discord survey someone was mentioning earlier was at 100% in favor of the feature.

[archont]

THIS SUCKS
Rationale:
1) No good way to visually communicate item quality
Whether it's dots or an item glow, the game needs to communicate that a Common item is different from a Legendary item. The current dots are UI leaking into the game world, which is a design hack.
I realize this is with overlay on. With overlay off, a Legendary inserter looks the same, but acts differently.
Cue PVP salt about legendary buggies with legendary bullets beating common spidertrons.
Cue "You should've hovered to see that's a LEGENDARY buggy and you have no chance"

2) Parallel mechanism to tiered machines
Machine efficiency is a two-dimensional matrix - machine tier vs modules. Now WUBE is adding a third axis. What's better, a legendary T2 assembler or Common T3 assembler?

Might be particularly frustrating with inserters.
Current behavior:
T0: Yellow inserter 10 u/m
T1: Blue inserter: 15 u/m

New behavior:
Yellow common inserter 10 u/m
Yellow rare inserter 12 u/m
Yellow epic inserter 14 u/m
Blue common inserter: 15 u/m
Yellow legendary inserter: 16 u/m
Blue rare inserter: 17 u/m

QUALITY does what ITEM TIERS do, but worse.
...

3) Possible optimization impact
Megafactories are possible because a compressed kilometer long belt of circuits is optimized into a single structure. On that compressed belt, the ordering doesn't matter - just item count. The freshly created circuit can magically teleport itself to the end of the belt and everything behaves the same. Not so if there's item quality.


4) Quality doesn't make sense for some items
Some items benefit greatly from quality, some do not. I'm expecting that 10-20% of items players will want in maximum quality, 80% don't care.
But quality applies to everything, it's a generic mechanism. Devs will be stuck trying to figure out what quality does to an item, potentially implementing a LOT of extra stuff.

5) Avoids reciple bloat, but not really
The design goal of item quality was to avoid recipe bloat, and that's checked on the surface level. Underneath though, if you account for quality in efficiency calculators, I expect the number of numerically unique recipes (ie. common plastic + epic red circuit + common wire + legendary wire + epic wire + epic wire) to become astromonical.
Factorio calc devs will have to do some advanced maths I feel, brute-forcing quality calculations may be computationally infeasible.

6) Breaking existing designs
Existing factory designs may rely on fine calibration and clever behavior. A higher tier version of an item might be faster, but that might result in breaking existing designs that are built for specific inserter speed, for example. Designs sensitive to this quality difference may demand specific quality components - which sucks, considering that quality, unlike existing item tiers, cannot be well communicated.
Cue "Finally found out why my uranium setup broke. My inserter was rare quality!"

7) Bethesda problem of Lesser+ Potion of Fire Resistance
Why do I have my inventory full of potions of fire resistance of varying qualities, giving anything from +15 to +60% fire resistance? Does having 4 potions of fire resistance of different qualities really add to the game?

8) Quality tier naming
That is, most obviously a duck.
https://www.simplethread.com/looks-great-lose-the-duck/
Unless WUBE plans to genuinely get a LLM to write a unique short legend for each Legendary Iron Gear Wheel.

RECOMMENDATION
REVERT NOW BEFORE IT'S TOO LATE
I'm fine with quality modules. I'm fine with special versions of items, but they should be unique items, with unique graphics.
4 tiers is too many. 1-2 should be enough, but quality version count ought to be item-specific.
Only select items, for which quality makes sense, should have high quality versions.

[dog80]

idk why so many people are so upset it enables a bunch of new layouts and calculations that have to be made + it increases scenario gameplay, hype for factorio++

[mcdjfp]

When I first read the Friday Facts, I was panicking about my ratios all getting messed up. Then I thought about it for a bit.

After a bit of though, I really like the concept. Quality is all over real life, and this seems like a fairly good way to bring it into the game. If you don't want to mess with it (either permenantly or for the moment), run everything at standard performance levels. Want to min/max? Sort out the high-quality components.

The names absolutely have to change I would back either no name or "Standard" for the base level and maybe "Flawless" for the top.

[solidzaku]

I can't support this. Let's say you're making copper wire. My trains pick up a specific amount because I have them keyed to certain ratios. Now all of that is going to be thrown out the window if I enable this. People are going to have to babysit their boxes at all times until they unlock logistics networks, constantly worrying about missing that 'legendary' wire. Fix the names, by the way, you're Wube, not EA/Bethesda...

This just makes things tedious in an un-fun way (and yes, I've played SE/Kras/Pyanodon). Yes, I know it's 'optional', but it's sounding like a lot of the new space content is reliant on this process, or was designed around it. Almost as if you're punishing players who play 'normally'. Oh, and RIP speedrunners. The new Factorio will become a temple to RNGsus.

It's not that I have a problem with the 'tier' system inherently. The fact that it's THAT random and effects every element of the supply chain, I'd just be completely turned off by it. Perhaps if it was a late/end game module that guaranteed certain qualities, perhaps. Just please have the space content tested in 'vanilla' mode. This just isn't a feature I see ever turning on, and I'd really rather not be penalized for that.

[morsk]

I can't support this. Let's say you're making copper wire. My trains pick up a specific amount because I have them keyed to certain ratios. Now all of that is going to be thrown out the window if I enable this. People are going to have to babysit their boxes at all times until they unlock logistics networks, constantly worrying about missing that 'legendary' wire. Fix the names, by the way, you're Wube, not EA/Bethesda...

We aren't supposed to use it that way, but a lot of people get the idea. I think Wube should not have used Electronic Circuit as the example in the video. It makes it look like quality spam everywhere across the factory. In a real mall it would be a Gun Turret or some Power Armor, and the random parts would only be on some loops at the mall.

It's actually kind of boring. I like the option it's giving, of upgrading a few things of my choice by making loops for them, and having a few extra "good" turrets to put in nice places. But it won't change much outside the mall. And the module-maker build; that might be the one where I let myself try something complicated. Unless there are more things about it we don't know yet.

[Loewchen]

We aren't supposed to use it that way, but a lot of people get the idea. I think Wube should not have used Electronic Circuit as the example in the video. It makes it look like quality spam everywhere across the factory. In a real mall it would be a Gun Turret or some Power Armor, and the random parts would only be on some loops at the mall.

It does make sense to go for quality modules as early as possible in the value chain though, better to have the recycling loop (and therefore the 75% loss) at green circuits than at processing units or later or all the cost added in between is lost as well.
So in gameplay early on I'd expect people to insert QM for low level intermediates and produce high level intermediates and entities from those without any risk of loss.

[morsk]

It does make sense to go for quality modules as early as possible in the value chain though, better to have the recycling loop (and therefore the 75% loss) at green circuits than at processing units or later or all the cost added in between is lost as well.
So in gameplay early on I'd expect people to insert QM for low level intermediates and produce high level intermediates and entities from those without any risk of loss.

So it makes later parts of the chain smaller, the opposite of how productivity makes earlier parts smaller? This is going to make the mall more interesting than I thought, but I don't think I would do it outside a mall, also I don't trust myself to reason about it without a spreadsheet anymore. The example is fine then, but I have no idea how to communicate that mixed quality is not supposed to spew everywhere unwanted.

[bgeniij]

I was initially a bit skeptical about this, but I am really starting to like the idea. (This is way longer than I planned. Apologies in advance.)

The bonuses granted by higher qualities are significant, but don't seem so powerful that they necessitate using this mechanic, so I should think that it would still be viable to simply scale up normal production instead of focusing on high-quality items. In most cases, it would probably be much cheaper and easier to build 5 normal machines instead of two legendary ones, but in the later game when building tall it might be worth it to create a loop to make high quality machines to save space.

The equipment grid boosts to armor and the spidertron would be much more "essential" since simply scaling up isn't possible, but this incentivizes the player to actually create a factory setup for armor instead of just manually crafting it a couple of times and then forgetting about it. (As a side note, it would be cool if there was some other use for armor or weapons so that the player doesn't just craft one or two the whole game. For example, maybe adding a humanoid sentry robot that can equip armor like the player (or that simply has armor as a crafting component), so there is a reason to keep producing armor, weapons, and equipment modules late in the game.)

I have some concerns over the logistics of sorting and transporting several different tiers of items, since I assume they wouldn't stack. Also, I can see how the quality system could really complicate mall design. If every single mall item had to have a dedicated setup like the one shown for green circuits, that would be incredibly annoying. However, based on the probability table, it looks like it is possible to guarantee even tier 5 items by using tier 5 ingredients and four quality modules (probably less if they were higher-quality quality modules), so creating setups to produce high-quality inputs seems like it would allow a normal, compact mall that still produces consistent high-quality products without requiring every product to have its own setup for filtering and reprocessing.

For example, the player could have a relatively normal setup, but create production loops for a few specific items that they really want to ensure are high quality. Alternatively, the player could create a massive setup to make all their most basic products high-quality to ensure that everything in their factory is made at the highest quality while sacrificing efficiency.

Honestly, just thinking about the different ways the player could do this makes me feel pretty excited. Is it worth it to craft legendary iron plates, or is it better to leave them at a lower tier and allow the quality bonus to accumulate as they are further processed into green and red circuits? Do I really care about making high-quality power poles or inserters, or do I only want high-quality radars and assemblers? Maybe I don't want to specifically create high-quality versions of every type of machine, so instead I can create a setup for high-quality modules that can affect whatever buildings I want. Do I create production loops for each product to ensure maximum quality, or do I create a centralized recycling and sorting system that reprocesses inferior items back to my main bus?

The possibilities are endless, and the quality system could open up tons of new possibilities when it comes to factory design. Of course, it also creates openings for unnecessary tedium. For example, a setup might jam after receiving the wrong quality product. It is already problematic when you accidentally have the wrong item get on a line because you accidentally rotated a belt or something. It can already be incredibly annoying to find the jam, and I imagine that it would be even harder if the offending item was identical to everything else around it except for a couple of small dots next to it. It is hard enough finding the one iron ore that snuck onto a belt carrying iron plates, imagine finding a tier two iron plate on a belt of tier three iron plates. This might not cause a jam by itself, but the random nature of quality means that now your later products have a 0.3% chance to be made at a lower quality, and you only created filters for high-tier items since you only anticipated tier 4-5 products, and now everything grinds to a halt. This could be prevented with filters, but then it could get tedious to meticulously filter every branch in your factory, especially since you potentially have five different quality levels to sort through.

I can definitely see how the quality system could be annoying for people, and I hope that it will still be viable to complete the game without using it (there could even be a new achievement for doing so: "Beat the game only using normal quality items"). I have some concerns, but I think there is a lot of potential for this system, and I am excited to experiment with it. The developers seem to like it in their playtesting, and I think they have a good sense of game design, considering they made a game as amazing as Factorio.

Final note: I don't necessarily hate the current quality names, but the ones suggested by everyone else are so much better. That, or just copy the quality names from Dwarf Fortress.

[NotRexButCaesar]

At first I thought this was going to be a bad idea, but now that I have thought about it more, I think it will probably be fine.

I was originally worried that having so many different kinds of items would cause me some frustration at not being able to make things uniform, but I think it will probably not be much of an issue because it seems like it will be somewhat easy to separate out quality items from normal items—my power poles will not all be different if I have a blueprint with normal power poles, and the better poles can be used in a few helpful areas. I hope that there is some way to let higher quality items stack with lower quality and just reduce the random higher quality items when it is irrelevant. I want to be able to differentiate situations where I care about separating quality and where I do not care.

Like almost everyone else, I do not like the "rarity" names for quality—I suggest the Dwarf fortress names: Well-crafted, Finely-crafted, Superior quality, Exceptional, Masterful. Because Factorio will only have five levels, they could be shortened to (normal), Fine, Superior, Exceptional, Masterful.

I am also not sure I particularly like the randomness factor, but I cannot think of a much better solution—it was stated that some other mechanics will interact with quality, so I am not greatly worried about it now. Recycling seems like it might be somewhat annoying, but I have not played with it, so it is possible that not much of a difference is made.

[Splitframe]

The upgrade chances don't look quite right to me for how it was explained in the article for a 1% quality boost

out of the 1000 gears, 9 would be uncommon and 1 rare.

I believe these are the correct chances instead for the 10% quality boost of 4 regular tier 3 quality modules:

This bothered me as well, omitting the 0% and ~90% chances plus the 5th row for legendary iron plates surely confused a lot of people.
I also contemplated to adjust the graphic like you did.

[AirForce1]

When I first read the Friday Facts, I was panicking about my ratios all getting messed up. Then I thought about it for a bit.

After a bit of though, I really like the concept. Quality is all over real life, and this seems like a fairly good way to bring it into the game. If you don't want to mess with it (either permenantly or for the moment), run everything at standard performance levels. Want to min/max? Sort out the high-quality components.

The names absolutely have to change I would back either no name or "Standard" for the base level and maybe "Flawless" for the top.

Yeah quality is all around us but real factories dont spit out random quality product in very low probability. Real factories have quality control and output production meet desired quality standard in very high chances that is about 95%. so in average on general basis its about 95% of output products meet desired quality standard.
Soooo.... nothing like chance of high quality item in 0,01%

[AirForce1]

I can't support this. Let's say you're making copper wire. My trains pick up a specific amount because I have them keyed to certain ratios. Now all of that is going to be thrown out the window if I enable this. People are going to have to babysit their boxes at all times until they unlock logistics networks, constantly worrying about missing that 'legendary' wire. Fix the names, by the way, you're Wube, not EA/Bethesda...

We aren't supposed to use it that way, but a lot of people get the idea. I think Wube should not have used Electronic Circuit as the example in the video. It makes it look like quality spam everywhere across the factory. In a real mall it would be a Gun Turret or some Power Armor, and the random parts would only be on some loops at the mall.

It's actually kind of boring. I like the option it's giving, of upgrading a few things of my choice by making loops for them, and having a few extra "good" turrets to put in nice places. But it won't change much outside the mall. And the module-maker build; that might be the one where I let myself try something complicated. Unless there are more things about it we don't know yet.

You will have tu build whole new extra production lines for almost everything and hold very high amout of epic items of all sorts in storage and I mean about tens of thousands because if you suddenly build new big block of late game base it could easily consume very high amout of lots of items. So good luck to have like 2000 epic inserters, 1000 assemblers and the king of all few thousands T3 epic production module - you would AFK hours to build them even with quite big factories.
And imagine if the alter for example belts so betls would have some extra properties so good luck tu build like few thousands of epic belts, undegroud belts and spliters - You would need like hundreds of thousands of epic gears!

[Chrisdasdasd]

Don't get your hopes up, most of the most popular mods introduce RNG in some of the recipes..

Luckily, I can write one myself (as long as API will allow it) so maybe that's some hope for you?

yes, your ups will potentially suffer if you mix quality items on your bus in your megabase. [...]
(the whole science per ups argument is imo silly because any attempt at making the game more complex, and therefore in my opinion interesting, will tend to hurt the UPS simply due to having to build more stuff for the same amount of science. But I digress...)

I have to agree with you.
It's just the way they want to implement it...
I see myself juggling these over and over again in recycler until I get perfect ones and I hate the idea

[FuryoftheStars]

I see myself juggling these over and over again in recycler until I get perfect ones and I hate the idea

Same. I'd only use the in-between qualities as far as what was required to produce tier 3 Quality modules. The rest I wouldn't waste my time with as I have no desire to upgrade part of the base with Q2 or Q3 items to then start replacing those with Q4 to then yet again replace with Q5. :/

[pleegwat]

We aren't supposed to use it that way, but a lot of people get the idea. I think Wube should not have used Electronic Circuit as the example in the video. It makes it look like quality spam everywhere across the factory. In a real mall it would be a Gun Turret or some Power Armor, and the random parts would only be on some loops at the mall.

It does make sense to go for quality modules as early as possible in the value chain though, better to have the recycling loop (and therefore the 75% loss) at green circuits than at processing units or later or all the cost added in between is lost as well.
So in gameplay early on I'd expect people to insert QM for low level intermediates and produce high level intermediates and entities from those without any risk of loss.

Not everything needs quality. Rather than recycling low-quality green circuits, it may be more effective to route them to science production while the high-quality ones go to producing factory components.

[AirForce1]

We aren't supposed to use it that way, but a lot of people get the idea. I think Wube should not have used Electronic Circuit as the example in the video. It makes it look like quality spam everywhere across the factory. In a real mall it would be a Gun Turret or some Power Armor, and the random parts would only be on some loops at the mall.

It does make sense to go for quality modules as early as possible in the value chain though, better to have the recycling loop (and therefore the 75% loss) at green circuits than at processing units or later or all the cost added in between is lost as well.
So in gameplay early on I'd expect people to insert QM for low level intermediates and produce high level intermediates and entities from those without any risk of loss.

Not everything needs quality. Rather than recycling low-quality green circuits, it may be more effective to route them to science production while the high-quality ones go to producing factory components.

Absolutely terrible idea. Making low level items like you suggest would be nightmare because in very late game you have to produce this items in very large scale in tens or hundreds of thousand (or even milions) per minute and in this scale your factory have to be compact because you would have thousands of factories make just one low level product.
So if you want to have your factories in reasonable size you cant increase size by lots of extra belts and inserters for quality byproduct and sorting stuff. Also in very late game scale there is no other reasonable way then use trains for tranport everything everywhere and now you would have to have everywhere at least 5 times more trains, 5 times more input/output train staions and your base would be 5 times more bigger.

It would be much easier and more compact to do separate production lines in small scale for everything for Q5 production with that stupid RNG loops lines. Because like Wube suggest some items have no reason to be in better quality and lots of desired Q5 product would be thing you would stock in warehouse in few thousands for sudden base expansion and some of them would be in constand rates like mining machines that will be deconstructend after ore is depleted and used again somewhere else.

[Jon8RFC]

I like the idea in principle, but the randomness of the quality seems contradictory to the nature of engineering. IRL, Nobody would tolerate building a car, which randomly attains the desired quality, and subsequently recycling multiple runs of it until it comes out perfect by chance.

Hilariously, that is/was the method of American car production.

I recall reading an article about American engineers going to a Japanese vehicle production plant for a tour and asking where their bone yard was. The Japanese were confused and everyone thought there was a translation issue.

It turned out that the American engineers would just toss numerous bad parts/cars off to the side. The Japanese didn't have such a situation. They said "when we produce something bad, we examine it to find out why it was bad and make corrections so that it doesn't happen again".

[Roughwaves]

I guess I'm late to the party.

I'm not excited about this new feature to begin with, I'm concerned about inventory management which is already pretty cluttered for me. I can only imagine having a box with 15 types of assemblers. I think adding quality to intermediate items could be bulky and frustrating that it would only provide a bonus chance. We will be able to filter by quality, but does that mean we carry 5 times more stuff, or that we'll have to flip to 'all qualities' filter to see how much inventory space we actually have? Like I said, concerns.


But I'm really not impressed by the names and colours, it just feels uncreative and un-factorio, and the random recipe outcomes definitely seem chaotic. I understood that you didn't want to add the loop recipes and complexity of Bobs/py/etc, but now you seem to have added a similar complexity of loop recipes without any of the creativity. It'll be the same quality sorter stack with recycler looping back to the inputs, for literally everything. Is that part adding as much as you think it is?

I'm sorry to be so critical, my friends.
I won't offer any solutions, this is your game, all I offer is my reaction.

Good luck