Friday Facts #442 - Flip, Flow, and Fresh Paint

[Hurkyl]

Yeah you have no idea what you're talking about. The legendary : common ratio on throwing chunks into the recycler straight is ~1/3000, maybe?

2727.

And huh, a single common quality recycler processes 16 asteroids per second. I guess you probably need to recycle around 3636 times (2727 times 1 + 1/4 + 1/16 + ...) (it's probably not actually that simple but I imagine this is in the right ballpark) to get through that, so if you have unlimited asteroids that recycler can make 0.264 legendary asteroids per minute.

Times 43 for crushing with 50% productivity (0.7 asteroids make 30 ore), and another 1.5 for smelting in electric furnace, and that turns into 17 legendary iron plates per minute.

(that would be times 400 if crushing with 300% productivity)

[jackthesmack]

I'm registering an account specifically to say that I don't agree with the removal of space casinos.
They are the only way to make quality modules that isn't horribly tedious. Without space casinos I'm just going to not play with quality anymore (since you're making that an option now)

How is throwing quality modules in an assembler tedious? It's way less tedious than building a giant upcycling spaceship that then must be reprocessed on Vulcanus.

[coffee-factorio]

I mean, think critically about what I'm saying. This fundamentally doesn't work at small scales. You either have to have as part of a "voluntary required cleanup program" on all platforms that would normally be discarding material when you're making, I don't know whatever you get 150 hours in. Or you need to have it on a ship dedicated to it. I wouldn't venture to say you can do it simultaneously with promethium until you have an operation of truly ridiculous scale. Which a megabaser might take as "Tuesday."

If I had to bet five dollars, I would bet that a required function call to the boiler plate code in front of quality, then probably another function call to a PRNG, would hit the CPU worse than incrementing a counter on a particle and then deciding whether or not to draw it, because often people aren't looking at things they're throwing away on a space platform. It would be kind of like something you did to pad your supply of building materials because you had enough but a little more couldn't hurt.

It's a weird thing. Because optimization is a black box and context sensitive. So there's stuff that would have better rates of return but wouldn't be practical because it's like... 1800 random number generator calls on copper wires or ores. Abucmcnasty is going to have a long, long career. God bless him.

[Hurkyl]

I'm registering an account specifically to say that I don't agree with the removal of space casinos.
They are the only way to make quality modules that isn't horribly tedious. Without space casinos I'm just going to not play with quality anymore (since you're making that an option now)

How is throwing quality modules in an assembler tedious? It's way less tedious than building a giant upcycling spaceship that then must be reprocessed on Vulcanus.

I really think it's a misattribution. I think it's very likely that most of them don't actually find reprocessing loops any less 'tedious' than craft-recycle (or other) loops at all, and what they are experiencing is that they just have to do less of it because they can have one loop to rule them all and the large expansion factor when unpacking from asteroids to ores.

But people in general are really bad at rationalizing why they feel a certain way, and there's a meme for them to latch onto that short-circuits the process.

[Kalixt]

Thanks for nice update, fluid fix especially !

I like quality but when it comes to science packs it feels unfinished. I have server running since the release of Space Age, currently at 1.2M SPM, so I've tried to max-max everything the game has to offer. When it comes to science packs, its just feel wrong. It creates so much extra complexity and so many unnecessary setups that it's actually better to make non-quality science packs, and you end up saving UPS. So you're basically punished for trying to make things more efficiently.

In order to justify this, legendary quality packs would have to output more value. Also removing space casino does not help to make things better in this regard.

[Robosium]

I get about 1K prometheum per minute from my prometheum hauler. And that's low because it's early generation.

But you guys all realize that the recycle time on an asteroid is 0.03 seconds?

And that's not bragging. In the world of prometheum that's kid numbers.

And you don't get less huge asteroids than you get prometheum ones for a good long while.

Did you guys try looking at the production graph to get some perspective on what you could get?

Or as Hurkyl put it.

Playing as the house instead of the victim, because at say... 20 to 400 ores per asteroid times 1000/2727 the house wins in the end.
That's to the shattered planet of course.

Inner system is more like 182 asteroids per minute. At that rate I make hmm. 0.06 legendaries. To put that in perspective compared to the shuffle. It makes about 0.5 asteroids per minute per kind on about 7/8ths of an asteroid per minute.

So when you look at it from that perspective, and you think about it. Yeah. That thing that was as effective as flying where you need a fusion reactor and railguns to see the same rate?

Yeah.

That was probably going to get balanced out of the game.

Yeah you have no idea what you're talking about. The legendary : common ratio on throwing chunks into the recycler straight is ~1/3000, maybe? I've just stuffed a turbo belt of chunks into recyclers in sandbox and it's actually hard to measure because it's so small it might as well not exist. I'm sure someone can calculate it mathematically but the point is that space casinos are completely gone with the change the devs made. They could nerf it in ways to still let it exist, but we'll see if they reconsider their approach or not.

I actually ran the numbers and by just using recyclers with legendary quality 3 modules it takes 2500 common items to get 1 legendary, with 54 chunks in a huge asteroid that is around 46,3 huge asteroids to get 1 legendary chunk just by recycling, used a Markov chain with 2 absorbing states, 1 for destroyed the other for legendary.

[factomight]

I can't add anything that hasn't already been said, but for what it's worth, I'm against killing Space Casinos too.
What do you care anyway? It's a very late-game mechanic that the vast majority of Factorio players will never reach. Those who do have already mastered all the other approaches to quality to get to that point. Why take this away?
It's not any more challenging for an experienced player to set up and spam a more resource-intensive and tedious upcycler just to get the desired number of legendary items. Your argument about this being too strongis kind of rich in a singleplayer game at the very end of its development cycle. You say you don't want to be fun killers, so how about just letting players have their fun with engaging space based builds instead of forcing them to play the way you want them to.

Id like to propose a simple solution: Make quality reshuffle an optional game mechanic, disable it by default and let the the community decide.

[Hurkyl]

Id like to propose a simple solution: Make quality reshuffle an optional game mechanic, disable it by default and let the the community decide.

They aren't changing the game mechanics. They're updating the recipe definitions in the Space Age mod files: presumably adding a single line allow_quality=false, to the definitions of each of the asteroid reprocessing entities in {your game install}/data/space-age/prototypes/recipe.lua.

And the "option" to remove that line would also be a mod. Or, I suppose, manually editing the files.

And I'm told the mod for that already exists.

What you want is already done, in the way (AFAIK) that the devs want 'options' like this to be done.

[factomight]

Sure whatever. Add an official 'Quality-SpaceCasino' and 'Quality+SpaceCasino' and everyone is happy.

[Cubickman]

Isn’t there a problem with the graphics of the windows on the pipes ?
Shouldn’t the bubbles be moving horizontally rather than vertically ?

[KillingTimeItself]

Personally i dont mind space gambling going the way of the dodo, it was kind of an odd mechanic from the get go, but nobody cares about that anyway, what we all want is quality repair packs that influence the repair speed

Now, i am very impatient and i hate waiting on repairing stuff like my spidertron for instance, but it would also be a nice mid/late game bump for robots so you don't necessarily have to grind one morbillion quality bots just to hope they don't die to flamethrowers quite as easily.

[AvengerStar]

I knew that when this Friday Fact mentioned the effective removal of the space casino strategy, that it'd be the prevailing subject on the associated forum thread, and over a dozen pages later, that still seems to be the case. Suppose I'll throw my hat in the ring, as well.

The problem is quality as presently implemented is boring. This was a problem I foresaw all the way back to when this mechanic was first unveiled. One of my major complaints about the idea at the time was that the solution on how to get one particular item to a higher quality is the same as it is to any other item: upcycle until you get what you want, and optionally augment your resource supply chain such that every step is lined with quality modules and sort out the mixed quality results somehow. The latter seems like a desirable thing to build until you try and plan it out, to which you come across having to exponentially expand your manufacturing and especially your logistics to accommodate. Look up Bigfoot's "Factorio - Quality" video on YouTube that describes this grand undertaking, which he dubs the "Full Monty". There are some people that might find this approach enjoyable, though to what degree the developers at Wube themselves envisioned this approach as practical is debatable (the FFF introducing quality mentioned this strategy, but seemed more inclined to highlight upcycling).

Since not many players seem to be privy to enacting such a grand construction project, naturally most people sought out simpler or at least less tedious methods. Enter the likes of the space casino or the LDS shuffle, the general strategy being to move the upcycling process over to the fewest base ingredients necessary to viably build out most everything that needs it. This effectively accomplishes the results of the "Full Monty" without needing a massive sprawling supply chain and eliminates the need for upcycling down the road. There have long been whispers of the devs disapproving of at the very least the space casino strategy (perhaps others as well, but I don't recall them feeling nearly as strongly about such alternatives).

To which I ask, what then even is Wube's vision for quality? There was an old argument passed around during heavier discussion for it that boils down to "since it's optional, if you don't find it fun, don't engage with it". It being optional is evidentially not entirely true; while it is technically possible to progress through the entirety of Space Age without a single quality module, by the devs own admission in this very FFF, it makes ship building, especially in the early game, significantly worse without it. Not to mention that the recycler being tied to the Quality mod while also a hard requirement for progressing Fulgora.

I guess someone at Wube finds upcycling fun and engaging gameplay? To the point that attempting to bypass it with tools only reasonably available in the late game is worth nerfing into impossibility? I'm all for respecting the developers' vision, but not when it clearly doesn't manifest properly. Quality as it is currently implemented always felt to me a hackneyed solution to the problem of implementing vertical progression, where the alternative is having to make more content consisting of new supply chains, possibly new items and buildings, and having to balance all of that. Understandable to not want to go that route, but I would hope that the couple of years that quality's been in the game would at least inspire someone on the team with a clever compromise, or at the very least understand that creative problem solving around boring gameplay isn't necessarily a bad thing.

If Wube's going to can the space casino approach, they at the very least should address why people even wanted to use that strategy in the first place, and it isn't just because it's overpowered (though that certainly did help its case). Quality's boring, and space casinos were one way to make it not boring. The replacement doesn't have to be nearly as powerful, but it should, at the very least, be as fun to build out.

[GregoriusT]

Quality's boring, and space casinos were one way to make it not boring.

Quality is nice in that it adds progression purely by making you build different Setups with the same Machines, and improving one significant stat about your crafted Machines, leading to more interesting factory designs during the time before Legendary.
That is the fun part about Quality.

It is much less boring than having the same copy-paste processing chains that you can solve with the same basic concepts that you used in all the previous Setups, like in modded pre-Space-Age Factorio.

Quality is a different way of doing things with lots of options to try out. That is what most people find exciting about it.
Now if you prefer designing the same thing over and over, then a mechanically different system like Quality is quite boring.

The original 2.0 Space Casinos were using a mechanic that looked a hell of a lot like an exploit because of how absurdly good they were.
While the new upcoming 2.1 Space Casino designs seem about as reasonable as the LDS Shuffle or Blue Chip Upcycling.

[coffee-factorio]

It's a weird one. The variety is there when you look for it. But it's something where you kind of have to go looking for it. I don't think two people will do it the same. I'm kind of a shark with it. Bite and spin.

[Hurkyl]

It's a weird one. The variety is there when you look for it. But it's something where you kind of have to go looking for it. I don't think two people will do it the same. I'm kind of a shark with it. Bite and spin.

I wonder if that's kind of the main point. (once you set aside outliers like asteroids, fluid-to-solid, and high productivity researches)

Classically you have fairly well-defined production subgoals. Like, if you're making yellow science, you have to make flying robot frames, processing units, and LDS. And if you're making flying robot frames, you have to make the electric engines, batteries, green circuits, and steel. And so forth; requirements flow like a waterfall through clear "to make this you have to make that".

But if you have a subgoal of legendary iron plates, the design space is much larger. The game doesn't tell you if you should toss plates in lava or recyclers, craft/recycle iron chests, craft/recycle some sort of belt in the foundry, a mix of those to try to get the best of several worlds, or something else entirely.

So it's interesting if you like looking through the space of options to find something that works well for your purposes.

But maybe not so much if you just want a clear defined goal so you know what to put together. Especially if you default to an option that is poorly optimized for your preferences.

[Bloodred217]

But if you have a subgoal of legendary iron plates, the design space is much larger.

Is it really, though, in a meaningful way? Once you set aside those outliers you mentioned, do you actually have a meaningful variety of options to make legendary iron plates? When I played the only real options - once the outliers are excluded, again - were to make upcycler loops. You can choose the recipe used in the upcycler loop, sure enough some may be better suited than others, but the fundamental thing you build is still going to be an upcycler loop. This does not register as variety in any meaningful way to me, I'm still making the upcycler loop and not some sort of different design. You can claim the design space is large because you can pick out of multiple recipes which will result in iron plate quality upcycling but this doesn't actually result in any real variety in how the game plays.

The only real variety to how you make quality items comes precisely from those outliers, other than that it's very repetitive and to me at least the repetition makes it tedious. Even with the outliers included, you will be making plenty of upcycler loops for items that the outliers are not applicable to.

[Tertius]

the fundamental thing you build is still going to be an upcycler loop. This does not register as variety in any meaningful way to me, I'm still making the upcycler loop and not some sort of different design. You can claim the design space is large because you can pick out of multiple recipes which will result in iron plate quality upcycling but this doesn't actually result in any real variety in how the game plays.

The only real variety to how you make quality items comes precisely from those outliers, other than that it's very repetitive and to me at least the repetition makes it tedious. Even with the outliers included, you will be making plenty of upcycler loops for items that the outliers are not applicable to.

As far as I see, there are many possible implementations for an upcycler loop. You can do it with dedicated belts, with sushi belts, with bots, hybrid, circuit controlled, one machine, multiple machines, recipe changing setup for single machines, recipe changing for a whole upcycler. There is a large variety. I did all of them. I preferred the recipe changing variant for the whole upcycler, because it's generic for a given production machine (assembler, emp, biochamber, foundry, cryogenic plant). I use this for the planet specific buildings. However to replace a space casino, it's best to build a custom bigger upcycler to get scaled up output volume and profit from the special productivity of processors, lds and plastic.

You can call this tedious, or you can call this many interesting challenges. I was overwhelmed at first, but found a way to proceed and get something. My criticism is that the game doesn't suggest or hints at acceptable solutions, so I had to do much trial & error to avoid resource wasting and being overflowed by intermediate quality items nobody will ever use.

The principle is the same for both, upcycling and space casino.
Upcycling is craft, then recycle if not match quality, then feed back recycling result.
Space casino is the same loop, just one part less: reprocess, then feed back if not match quality. It can be better tiled and scaled due to its simplicity. I guess it's that what the players seek: simplicity. This is also my criticism ("lack of direction"), but I don't blame it on space casino alone, it's a general issue with quality.

[Hurkyl]

But if you have a subgoal of legendary iron plates, the design space is much larger.

Is it really, though, in a meaningful way? Once you set aside those outliers you mentioned, do you actually have a meaningful variety of options to make legendary iron plates? When I played the only real options - once the outliers are excluded, again - were to make upcycler loops. You can choose the recipe used in the upcycler loop, sure enough some may be better suited than others, but the fundamental thing you build is still going to be an upcycler loop. This does not register as variety in any meaningful way to me, I'm still making the upcycler loop and not some sort of different design.

If I assume I'm doing lava-based industry, variations I would have to consider are:

• Recycling iron plates to themselves. Actually I've pre-emptively ruled this out, but I've seen people propose it seemingly seriously, and I should probably double check that the simplicity of design isn't significant
• How do I process the lava?
  • Do I cast to plates, or do I cast to sticks an direct insert into a recycler?
  • Do I do a large casting operation with quality modules, or a tiny casting operation with productivity and beacons?
• Craft-recycle on iron chest has the advantage of being small and fast. Lots of direct insertion. But requires a larger casting operation due to low efficiency.
• Craft-recycle on transport belts is a more complex two-ingredient design so in addition to being larger it takes more effort to lay out, and for reliability I also need some side operation to balance the ratio of gears and belts.
• Do I mix the two? Use iron chests to churn through the masses of common and uncommon products, then the more efficient transport belts for rare and up?
• Craft-recycle on underground belts would allow for greater throughput and smaller builds. But it's a lot more complex:
  • I probably want to craft the transport belts with quality
  • Balancing the ratios of iron plates to transport belts will take some analysis and effort
  • Surely I want to back out of this at epic quality to avoid slashing my throughput with unnecessary recycle steps. Do I need to do so earlier?
• Craft-recycle on green circuits profits from having an extra module slots, for a big increase to resource efficiency and a smaller casting operation. But now I have to juggle my iron and copper quality efforts.

The analysis alone is, IMO, an interesting thing to consider, but the logistics of actually laying all of these out is quite different too.

Other planets would replace the casting step with whichever of iron ore/gears you start with naturally. Gleba is the most interesting variation with being able to set up your farm factory to churn out middle-quality (or even high-quality) iron ore.

[Hurkyl]

And as an addendum, I'd like to point out that using craft-recycle on underground pipes would also be a good workhorse to consider for legendary iron production... but because of the liquid-to-solid exploit, you can't use it unless you're willing to do the exploit.

[Bloodred217]

As far as I see, there are many possible implementations for an upcycler loop. You can do it with dedicated belts, with sushi belts, with bots, hybrid, circuit controlled, one machine, multiple machines, recipe changing setup for single machines, recipe changing for a whole upcycler.

You are describing general Factorio mechanics which have absolutely nothing to do with the quality system or the variety this system provides, they can be applied to any other thing you do in the game and don't even need quality to be turned on. These are not strong points of the quality system, they are strong points of the other systems the game has.

The principle is the same for both, upcycling and space casino.

The principle is the same, but the "casino" requires designing an entire end-game spaceship around it in order to to exploit an opportunity that only becomes available at high tech levels and rewards your efforts with better yields that result from the emergent properties of several other mechanics interacting (as was described in the FFF itself). To me this is far more interesting and more exciting to pursue as an end-game goal over making more conventional upcyclers.

In my first playthrough of Space Age I was very excited when I realized what could be done and found that it actually works after testing it (I didn't spoil my first playthrough by looking at things online, so I had no idea). Even on subsequent playthroughs this is an end-game goal to strive towards. I think it's clear now that even Wube failed to notice this before launch, but these sorts of interactions between different game systems that you can use for optimization are exactly what make Factorio interesting and appealing.

I do agree that it's kind of "the same" in the end, but ultimately this just strengthens my argument that the quality system needs more options in general, not less. The system is already quite narrow in the fundamental approaches it allows, so removing approaches entirely while adding nothing new to enrich it is going to be bad for the game.

If I assume I'm doing lava-based industry, variations I would have to consider are

I don't mean to sound reductive, but I feel like you've written a lot of text that says nothing other than the fact that you need to pick which recipes you use for your loops, which I have already mentioned in my post as well. Yes, you do indeed need to think about the recipes and decide which one you think is best. Similarly to what I have said to Tertius above, you also mix in general game systems into the topic. Whether you use beacons on your baseline resource input, or you arrange for direct insertion and so on are virtues of Factorio's core mechanics and not virtues of the quality system, which in and of itself remains narrow and will become even narrower as of 2.1.

I'm not sure if I'm getting my point across, basically the way I see it the fact that you also need to consider all the things you consider for any build when making a quality upcycler doesn't add variety or depth to the quality system. That's just the depth of the core mechanics, the quality upcycler just goes on top and remains repetitive to me.