Friday Facts #375 - Quality

[Brathahn]

quality enjoyer.png (80.23 KiB) Viewed 4437 times

[morolo]

Nope.

I think I'll be skipping the expansion.

[Atraps003]

Cool feature, thanks.

[i-make-robots]

speaking of horizontal and vertical... Quality adds great breadth of play (more of the same) without adding any depth of play (new mechanics).

Will make the UI much more complicated, but only for DLC players, so no harm done.

Grinding for epic loot belongs in bad games that offer to sell a shortcut. Mod can be finished without using it? Great, will skip this feature like I skip flamethrowers.

[TheBuzzSaw]

Not a fan. Sorry.

Factorio largely centers around consistency, particularly for blueprints. I'm already having nightmares of accidentally capturing a "legendary power pole" in my design only to later realize the design will not work with normal power poles. The entire point is that higher tier items are different items entirely. Adding a new invisible-until-later tier is just unfun complexity.

Rather than tell people to "just don't engage with it", maybe relegate this work to a mod instead.

[DanGio]

Hehe so much feedback, FFF definitely live again
Just want to say the quality names didn't bother me at all at first. Now that they're mentioned by many here, I gave it a thought. I think I actually prefer the WoW reference as is because I think it's fun. Sure, it doesn't fit the industrial theme perfectly, but
wouldn't you love to get the *legendary burner inserter*??
I say : if you're implementing RPG elements into Factorio, go all out and take inspiration directly from WoW.
Getting to much funny reference would kill the game, but I don't mind a little. Remember that we sometimes launch fish into space

[Serenity]

I don't know. This seems like hell for making optimized designs. You can spend a lot of time designing sub factories that consume exactly the right amount (like X number of belts), with exactly the right amount of machines, and the right amount of inserters to move things. Throwing in some machines that have different speeds throws off all the calculations. It makes sense when everything of the same quality, but the random chances means that is not going to happen at the beginning. It's also not friendly for blueprinting. Whatever you blueprint will only work as designed with a specific mix of quality levels.

I like the recycling machines though. I've long played with a mod for that. Not for regular production, but just to get rid of these one time items like armor, and things you accidentally overproduce with uncapped chests and the like.

[Drizznarte]

Kovarex had added kovarex process to every item in the game .

[koliw_br]

Splitter 1 puts all "Epic" items to the right and sends them to splitter 2 which has the filter set to "Epic". How do items get into Assembly Machine 3 ? (They get in because they are in box 4)

[Faark]

Not super excited, but I cannot really explain why. First thought was "grindy", but it's factorio maybe it's just such a common game mechanic that it is below what I would want for factorio. But it makes a nice chance-based optimization challenge, sth base game could use some more. Maybe it's mostly naming, as others already mentioned.

[utetwo]

Anything to do with diverting resources and scarcity makes the factory feel more alive. The automatic recycling process creates a recursion that feels like a living organism.

I felt a bit dubious of this at first, but this post convinced me. I think it adds realism in the manufacturing process. Maybe change the identifiers to something less Fortnite-esque, though. Amazing work, devs! love the game

[Hi_ImKyle]

People clearly didn't read the whole thing. This entire system is optional. There's no one with a gun to your head saying you have to use it. Just don't unlock the quality modules and you play the game as if they were never there. Problem solved.

The names need changing 100%, a few suggested name sets are great but do not release this with the generic gacha tier names.

[pointa2b]

Splitter 1 puts all "Epic" items to the right and sends them to splitter 2 which has the filter set to "Epic". How do items get into Assembly Machine 3 ? (They get in because they are in box 4)

The filter splitter in #1 is likely set to ">= Epic" and #2 is set to "== Epic," meaning anything that isn't Epic quality is diverted to the other splitter output. Note the filtering in this image which gives you this functionality:

[MazorNoob]

I really, really, really don't like this update. I think it's explicitly taking the game in the wrong direction.

See, the early and midgame in Factorio has two components to it: the grind and the automation. The grind is meticulously setting up more and more stuff to produce more and more things. It's unlocking new recipes, new and less involved ways to do things, faster belts and machines, more efficient logistics, more powerful modules, faster movement and so on. It's the "horizontal" expansion.
The second, however, is automation. From doing things by hand to setting up ghost layouts, designing the initial factory, then getting access to roboports. Then, setting up circuit networks to more efficiently manage trains, covering your base in roboports for easier ware management, automated base defense, sending out spidertrons to do stuff, and so on. It's the "vertical" expansion where you unlock and design more and more complex solutions in order to overcome the bottleneck that is the player avatar.

I noticed that when talking about the lategame in Factorio, the focus is firmly on the "grind" side. The "established" meta for endgame production is 12-beaconing everything with top tier modules, because that's the best and the fastest. You should avoid lane balancers and nuclear power because it's bad for UPS. IIRC you should also avoid excessive radars and logistic networks, because that also hurts UPS. You should have more rail lanes, longer trains, direct insertion, yadda yadda, everything in service of making the factory bigger, more efficient, faster, and so on.

Except I don't care. The most fun I had in Factorio is when I gave myself the goal of building a big city block factory that could be managed entirely by placing down and deleting blueprints, without ever having to move anywhere again. This is where I find the best parts of factorio gameplay are. I designed a city block setup that could expand completely on its own, had a network of local logistic networks that traded items via wired up chests and inserters, could deliver basic wares like rails and fuel within its network, could place down train stations for transporting a variety of things from factory products to construction trains to trash trains, and in theory could even semi-automatically clean up biter nests. This was the run where I theoretically "beat" Factorio: I was able to manage all aspects of my factory from the map, no longer having to use my physical body. The UPS wasn't great, but I didn't care, it's the theory that mattered.

When you look at this update, you see content that caters only to the "grind" side of Factorio lategame. Build a lot of stuff to unlock more efficient stuff to replace your old stuff so you can build more efficient stuff, ad infinitum. It's no engineering where you have do design new clever stuff, it's labor. I mean FFS, the post literally gave us the "bottom to top tier component for any product ever" design in a single picture. For grinding players, it's a dream come true. For me, it's worthless. I'll just add some new ware types, a new thingy for my city block and my city block solution will go on exactly the same. I'll just need a bit more time for things to grind, but on a theoretical level absolutely nothing changed.

I really wish that instead of this, we got more toys to improve automation. Say, a simple CPU a la Shenzhen I/O as a high tier product that supplements circuits. Or radio. Or capacity to automatically place down and remove blueprints outside of mods. Or removal of global 500 entity simultaneous building placement limit that screws over any serious semi-auto blueprint setups. And maybe some new automation challenges like recipes that require controlled conditions to ensure good yield. Instead it seems we're going to the Factorio future of walking your spidertron for hours to replace square miles and square miles of your old machines with new 10% faster machines, again and again, for hours, forever.

[LiquidInsight]

You can actually see a little white >= symbol next to the four quality dots in the first splitter.

Untitled.png (108.42 KiB) Viewed 4213 times

On the FFF, I was initially dubious but I think this will be a cool addition to the game! It's an interesting system, that you can choose to engage with a later point. I don't mind the RNG element, it requires a bit of planning and adds some new puzzles to the game.

I wonder what the UPS implications are for belts that are mixing items of different qualities. I thought that the game treated stacks of compressed items on a belt as a single entity. Will mixing together the output into different qualities of item incur a performance hit?

[ShimmeringDroplet]

Is there any reason that quality production is done probabilistically instead of with a quota like with production modules? RNG is icky...

[ipeterov]

I feel bad for the reception of the feature, since you guys worked on this for such a long time.

I think it will be really fun from a certain perspective - I'm guessing it would make me feel a bit similar to Cookie Clicker (which I enjoy a great deal).

The mechanic is optional, you can just disable it, don't whine!
Not really. Of course it's possible to disable it, but the game has to be balanced around it. Disabling it would be like building a megabase without using modules - technically possible, but significantly harder.

Randomness is making the gameplay worse
I think it's the most fun part. This gives us a more complicated problem to solve - we'll have to feed the items back, and we'll have to more complicated ratio calculations, but that's what makes Factorio a good game in the first place.

The one thing I don't like about randomness (maybe I missed something), is that it looks like I might accidentally get some items that are better than intended when crafting. I don't want one of my smelters to be more productive than the others - that kind of stuff will make my factory unpredictable.

It would be nice if we could toggle this feature on a per-assembler basis. This way, we can decide where we want additional complexity in exchange for more productivity, and work on implementing this thing in our base gradually.

The names feel yucky and generic
Yep, have to agree with this one. Although I enjoy getting better color items in Diablo and games like that, I feel like it's possible to come up with names that fit better thematically.

[DarkShadow44]

This is a joke, right?

Right?

My thoughts exactly. No offence, but this reads more like an april fools than an actual FFF.

Assuming it is real:

I assume items with different quality don't stack. Sounds like a huge pain with inventory management and train/chest filter. Do I really have to dedicate 5 slots to each item to assure it can fit? Or get rid of the rare good items if I don't want to?

Quite frankly, I hate the random aspect the most. I love that you can calculate how much your factory produces, with the whole quality RNG this gets a lot harder. Heck, even productivity isn't random. I do prefer consistency, could we at least make it calculable instead of random?

[melind]

Overall I like it and it seems like a fun way to expand options. I have a couple concerns and suggestions:

- The 25% recycle rate is extremely lossy early game and putting a cap on prod seems like a weird fix. I agree recyclers will end up being a noob trap since it is easy to imagine someone chasing the cool upgraded stuff while not realizing how many resources are being throw away.

I'd suggest instead recyclers refund all or nearly all of the resources used and you keep track of how much to refund based on % of prod modules used. Eg, every time something is crafted you keep track of how much prod was used to create each intermediary and refund a % based on that.

There is definitely still a bit of a problem here is in that a single entity might need to know dozens of productivity rates if its in a complex enough resources chain, even assuming you are averaging productivity of components and subcomponents. If you instead recycle into base resources (iron, copper, oil, etc) you would only ever need to track those productivity of each of those resources. Regardless though I'd really like to see recycling done in a way that is usable mid game so if you want to build slower for quality you can.

Altogether I think the best setup may be to recycle to raw materials based on productivity used to create the item and set recycling high, 90% with tech to get to 99-100% or just 100% recycled to remove worries about loss.


- This is also going to generate a huge number of potential items in the game (up to 4x as many!). Which I think will make a couple pain points worse.

Updating modules in entities is pretty painful and mods that do this are becoming more mandatory to the point where I think it would be good for the native game to handle it. Between blueprints and blueprint upgraders I think we have basically everything needed. But you should be able to place a blueprint on top of a machine and have bots replace modules so it updates to match the blueprint. Also a checkbox on blueprints like "Allow lower quality" that determines if lower quality entities/modules can be used to build a blueprint or if you need to use the exact quality of each thing in the blueprint.

This is also going to clog up chests... like imagine a base where you are crafting miners and you only want one 1 stack. Today you would just limit the box to 1 stack but if you can produce quality you need 3 inventory slots to represent this. If you keep the box at 1 slot and you produce a quality item, it would never get moved to the box until you empty the box. If you increase the box to 3 slots you could very well get 3 stacks of miners. You could handle this with very simple circuitry, but forcing people to engage that system for something like this seems overbearing. A better solution might be to allow filtering of all boxes so players can limit and filter the slots to make sure they have better control over what space is reserved in a box for different qualities.

[Sopenas]

When i thought about it I think I would have implemented it in a bit different way:
* In every machine which builds stuff you have a setting - which quality is the lowest that this machine outputs. You can choose from common (default ) to legendary
* If machine is set to certain quality but using same algorithm as described it produces lower quality product, it automatically destroys it automatically adding salvaged parts to machines input slots. basically every machine would have recycling plant inside

This would avoid all this lower quality stuff that is produced which will require managing. Lower quality parts will be going round and round. Why?

But qaulity idea is interesting, just have doubts if it is implemented in right way.
Maybe some mods will do what i described.