Factorio Forums

CyberCider wrote: Sun Aug 31, 2025 7:58 pm Blue underground belts, underground pipe casting, big power poles, foundry concrete… Surely with these, it’s easier to see how they have more substance than asteroid rerolling. Which, again, is basically the same as ore recycling but with a return rate of >40% instead of 25%.

I agree with the recipes you mentioned being interesting, but right now the game is ill-equipped for them. I've tried underground pipe casting on Vulcanus but in the end it's still inferior to ore upcycling plus importing the ore in terms of UPS / throughput. The random output keeps fighting stack inserters and things keep clogging, all of which eat up UPS to solve. Besides, in the end you need iron plates and the recycler tax cancels out most throughput advantage. The resulting puzzle is unrewarding and feels more repetitive than scaling asteroid cyclers.

BTW: the asteroid cycling return rate is about 33% and the ore cycling return rate is about 7.7%. You got both numbers wrong.

h.q.droid wrote: Mon Sep 01, 2025 4:37 am

CyberCider wrote: Sun Aug 31, 2025 7:58 pm Blue underground belts, underground pipe casting, big power poles, foundry concrete… Surely with these, it’s easier to see how they have more substance than asteroid rerolling. Which, again, is basically the same as ore recycling but with a return rate of >40% instead of 25%.

I agree with the recipes you mentioned being interesting, but right now the game is ill-equipped for them. I've tried underground pipe casting on Vulcanus but in the end it's still inferior to ore upcycling plus importing the ore in terms of UPS / throughput. The random output keeps fighting stack inserters and things keep clogging, all of which eat up UPS to solve. Besides, in the end you need iron plates and the recycler tax cancels out most throughput advantage. The resulting puzzle is unrewarding and feels more repetitive than scaling asteroid cyclers.

BTW: the asteroid cycling return rate is about 33% and the ore cycling return rate is about 7.7%. You got both numbers wrong.

As an aside, what I've done with quality in really fast buildings is that I do things like having one stack inserter filtered to normal, one stack inserter filtered to uncommon, and one bulk inserter for the rest.

And for the recycling tax, you can do things like recycling down to plates when you get to epic and use some other method to get to legendary. In the end, I think you're only losing on throughput due to this aspect when you have to recycle something legendary.

h.q.droid wrote: Mon Sep 01, 2025 4:37 am

CyberCider wrote: Sun Aug 31, 2025 7:58 pm Blue underground belts, underground pipe casting, big power poles, foundry concrete… Surely with these, it’s easier to see how they have more substance than asteroid rerolling. Which, again, is basically the same as ore recycling but with a return rate of >40% instead of 25%.

I agree with the recipes you mentioned being interesting, but right now the game is ill-equipped for them. I've tried underground pipe casting on Vulcanus but in the end it's still inferior to ore upcycling plus importing the ore in terms of UPS / throughput. The random output keeps fighting stack inserters and things keep clogging, all of which eat up UPS to solve. Besides, in the end you need iron plates and the recycler tax cancels out most throughput advantage. The resulting puzzle is unrewarding and feels more repetitive than scaling asteroid cyclers.

BTW: the asteroid cycling return rate is about 33% and the ore cycling return rate is about 7.7%. You got both numbers wrong.

You're right about the comment on the recycler tax, and the game just not handling the idea well. The repeated divisions by 4 hit an item which is clearly not being multiplied by 4, so you're always losing something. More often than not you're penalized for being creative.

It's incredibly frustrating to figure out how to do something novel like this.
And then realize the red belt cycle is essentially for aesthetics. And as you go through, you realize that abandoning the idea there is the correct choice because blue belts abuse your time the same way but also take lubricant.

And it's so consistent that yeah, if you have a showcase recipe I'm not surprised people are calling it a gimmick and asking for it to be banned instead.

coffee-factorio wrote: Mon Sep 01, 2025 7:40 am

h.q.droid wrote: Mon Sep 01, 2025 4:37 am

CyberCider wrote: Sun Aug 31, 2025 7:58 pm Blue underground belts, underground pipe casting, big power poles, foundry concrete… Surely with these, it’s easier to see how they have more substance than asteroid rerolling. Which, again, is basically the same as ore recycling but with a return rate of >40% instead of 25%.

I agree with the recipes you mentioned being interesting, but right now the game is ill-equipped for them. I've tried underground pipe casting on Vulcanus but in the end it's still inferior to ore upcycling plus importing the ore in terms of UPS / throughput. The random output keeps fighting stack inserters and things keep clogging, all of which eat up UPS to solve. Besides, in the end you need iron plates and the recycler tax cancels out most throughput advantage. The resulting puzzle is unrewarding and feels more repetitive than scaling asteroid cyclers.

BTW: the asteroid cycling return rate is about 33% and the ore cycling return rate is about 7.7%. You got both numbers wrong.

You're right about the comment on the recycler tax, and the game just not handling the idea well. The repeated divisions by 4 hit an item which is clearly not being multiplied by 4, so you're always losing something. More often than not you're penalized for being creative.

The divisions by four are balanced by the fact you have a higher quality% in the process.

IIRC, in terms of the number of inputs you need to produce a legendary output, the asteroid reprocessing loop is only slightly better than if you were using an asm3+recycler loop. And if you have something better -- e.g. extra productivity bonuses or more module slots -- it's significantly better than the asteroid reprocessing loop.

(the basic processing with high productivity research is a different story, of course, since the loop spits out lots of bonus ores)

coffee-factorio wrote: Mon Sep 01, 2025 7:40 am The repeated divisions by 4 hit an item which is clearly not being multiplied by 4, so you're always losing something. More often than not you're penalized for being creative.

No, it's easy to test with a monte carlo simulation, as you call them

Hurkyl wrote: Mon Sep 01, 2025 8:43 am The divisions by four are balanced by the fact you have a higher quality% in the process.

What ? It seems to me that the division by 4 are balanced by the fact that you have 25% chance of getting something as a result if the original value was 1. 1/4 = 25%

Shulmeister wrote: Mon Sep 01, 2025 9:34 am

Hurkyl wrote: Mon Sep 01, 2025 8:43 am The divisions by four are balanced by the fact you have a higher quality% in the process.

What ? It seems to me that the division by 4 are balanced by the fact that you have 25% chance of getting something as a result if the original value was 1. 1/4 = 25%

I... really don't understand your objection here.

But anyways, I've seen a lot of people make the mistake dismissing craft-recycle loops with the thought "reprocessing gives you 80% of your inputs back, and the recycler gives 25% back" and fail to realize the other aspects make up the difference.

It sounded like the person I was replying to was making that mistake. With additional confirmation in the fact he was dismissing a foundry-based loop.

Hurkyl wrote: Mon Sep 01, 2025 10:37 am But anyways, I've seen a lot of people make the mistake dismissing craft-recycle loops with the thought "reprocessing gives you 80% of your inputs back, and the recycler gives 25% back" and fail to realize the other aspects make up the difference.

This is the primary misconception. For anyone who doesn't want to go looking for the reason why it isn't true I'll say it here.

Asteroid reprocessing has only two modules compared to other machines which have 3+ (usually 4), so the reprocessing is not 80% but more comparable to 64% (0.8*0.8). In addition, there is no crafting where productivity or more quality could be added (100% productivity would make the comparison 64% to 50% which is quite comparable and 200% productivity makes it 64% to 75%). (these numbers aren't exactly right but they demonstrate the idea well enough).

h.q.droid wrote: Mon Sep 01, 2025 4:37 am I agree with the recipes you mentioned being interesting, but right now the game is ill-equipped for them. I've tried underground pipe casting on Vulcanus but in the end it's still inferior to ore upcycling plus importing the ore in terms of UPS / throughput.

This doesn’t sound right. At which levels of mining productivity does this occur? I can maybe see underground pipe casting getting surpassed by ore recycling after several hundred levels, but I doubt blue undergrounds would be surpassed until at least level 1000 or something. Otherwise, upcyclers have superior throughput. Also, big power poles for steel and LDS for copper plates seem to be in absolutely no danger of being surpassed by ore recycling.

The random output keeps fighting stack inserters and things keep clogging, all of which eat up UPS to solve. Besides, in the end you need iron plates and the recycler tax cancels out most throughput advantage. The resulting puzzle is unrewarding and feels more repetitive than scaling asteroid cyclers.

It’s generally not advisable to use stack inserters with any process that has multiple outputs. And if you do, you should have multiple stack inserters per machine so that each is filtered to only one output. That’s the only way it will work reliably. Stack inserters aren’t universally preferable to bulk inserters, and the game itself actually acknowledges this. Fun fact for those that didn’t know, a default upgrade planner will not upgrade bulk inserters to stack, you have to specify this upgrade to show that you’re aware of the tradeoff. And this was added in a patch, deliberately, it’s not a bug.

Also, I think your concerns about UPS are somewhat unwarranted. Science production starts to impact UPS at gigantic scales, but quality production tends to not even approach these scales. Besides, quality production is turned off when enough supplies are constructed.

BTW: the asteroid cycling return rate is about 33% and the ore cycling return rate is about 7.7%. You got both numbers wrong.

Oh, thank you for correcting me, that’s most helpful. These numbers are even more in my favor.

CyberCider wrote: Mon Sep 01, 2025 1:27 pm This doesn’t sound right. At which levels of mining productivity does this occur? I can maybe see underground pipe casting getting surpassed by ore recycling after several hundred levels, but I doubt blue undergrounds would be surpassed until at least level 1000 or something. Otherwise, upcyclers have superior throughput. Also, big power poles for steel and LDS for copper plates seem to be in absolutely no danger of being surpassed by ore recycling.

The main difference is ore mining becomes superior to molten iron casting (in terms of throughput) at around 200 with big mining drills and speed modules. I'm running at level 1200 where each miner gives more throughput than rows of legendary foundries. The bigger deal is, miners don't need inserters for their throughput and by mining into cars and destroying unwanted ore in those cars, you can saturate stack inserters with uncommon ore or even get rare ore at descent bandwidth. Blue undergrounds are fast but you have to first put gears into those foundries (and somehow recycle them into something useful later), which will eventually bottleneck the setup. Of course, neither can hope to compete with asteroid cycling for iron ore.

If you look at the ratios, doing blue undergrounds gives you ~5% rares. Direct quality mining gives you ~2% rares or ~3.5% if you upcycle the uncommons. The yield difference isn't even that big here. There are also 0.02% by-chance legendaries. It may not look much, but if you mine 10k or 20k ores per second, that can be enough for infrastructure.

CyberCider wrote: Mon Sep 01, 2025 1:27 pm Also, I think your concerns about UPS are somewhat unwarranted. Science production starts to impact UPS at gigantic scales, but quality production tends to not even approach these scales. Besides, quality production is turned off when enough supplies are constructed.

My very point is doing quality science. That's where you need to seriously scale up quality. You can see the scale I'm working at here:

coffee-factorio wrote: Mon Sep 01, 2025 7:40 am You're right about the comment on the recycler tax, and the game just not handling the idea well. The repeated divisions by 4 hit an item which is clearly not being multiplied by 4, so you're always losing something. More often than not you're penalized for being creative.

Well, of course you lose things. That’s how quality achieves the effect of costing more with each level. Although calling the loss “repeated” when it typically only happen once or at most twice is a little excessive. Besides, with the right buildings, the inputs of the craft get multiplied by at least 1.5. And in the case of recipes like belts which don’t take productivity modules, you also get an additional quality upgrade. This increases the value of the material that survives the recycle rather than its quantity.

Also keep in mind that much of the quality iron you produce is used as either pipes or gears. Pumpjacks, oil refineries, chemical plants, pumps, engines (so by extension electric engines too), assemblers 1 and 2, mining drills and labs for big drills and biolabs, most turrets, recyclers… All don’t rely strictly on iron plates, they require gears and/or pipes. So the recycling loss isn’t even guaranteed to occur, you can keep a lot of the iron in the form of gears or pipes.

And then realize the red belt cycle is essentially for aesthetics. And as you go through, you realize that abandoning the idea there is the correct choice because blue belts abuse your time the same way but also take lubricant.

A piece od advice I can give you is to not cast the input materials with quality, but with productivity and speed instead. The quality upgrades will be adequately provided by the cycle, which processes materials (and therefore rolls quality) at a much faster rate than the casting recipes do. Quality modules are best used in places where productivity modules are not allowed, as in most cases productivity will result in higher throughput.

Also, lubricant is easily sourced by coal liquefaction. And don’t forget Fulgora, the planet of free lube and plentiful gears.

Hurkyl wrote: Mon Sep 01, 2025 10:37 am

Shulmeister wrote: Mon Sep 01, 2025 9:34 am

Hurkyl wrote: Mon Sep 01, 2025 8:43 am The divisions by four are balanced by the fact you have a higher quality% in the process.

What ? It seems to me that the division by 4 are balanced by the fact that you have 25% chance of getting something as a result if the original value was 1. 1/4 = 25%

I... really don't understand your objection here.

But anyways, I've seen a lot of people make the mistake dismissing craft-recycle loops with the thought "reprocessing gives you 80% of your inputs back, and the recycler gives 25% back" and fail to realize the other aspects make up the difference.

It sounded like the person I was replying to was making that mistake. With additional confirmation in the fact he was dismissing a foundry-based loop.

Maybe there is a misunderstanding, from your words it seemed as if you were partially agreeing that the division by 4 could potentially be treated poorly by the game when not using quality.

There was nothing correct in the claim that the games treat poorly division by 4 in the recycler, be it with or without quality.

It appeared to me that the person you were replying was trolling.

h.q.droid wrote: Mon Sep 01, 2025 1:58 pm Blue undergrounds are fast but you have to first put gears into those foundries (and somehow recycle them into something useful later), which will eventually bottleneck the setup. Of course, neither can hope to compete with asteroid cycling for iron ore.

I find that kind of hard to believe. The foundry is a big building, and legendary stack inserters are fast. And the results I obtained were based on an unbeaconed upcycler, so there is plenty of room all around the building for very many input belts. And yes, I know that ore recycling is typically slightly beaconed, and I did test it with that. But I left the upcyclers unbeaconed. Perhaps if I also beaconed the upcyclers to match, they would have an even bigger advantage? I think I might explore this possibilty.

Also, a significant portion of quality iron is used as gears and pipes. Not all of them will be getting recycled.

My very point is doing quality science. That's where you need to seriously scale up quality. You can see the scale I'm working at here:

Well, that’s a special kind of challenge playstyle. You can’t realistically expect the game to cater to it if it comes at the expense of the default playstyle, which everything is designed and balanced around. There are so many unique playstyles and player-invented challanges, the game can’t possibly be expected to fully support each one. Especially so if your main concern is UPS. From my experience, but do correct me if I’m wrong, Factorio isn’t a game that sacrifices its game design for performance. Besides, remember that Factorio has excellent mod support. If your challenge is grounded in a ruleset that allows the exploits, you would only be using the mod to ensure that the challenge can continue within the rules it was intended for. I don’t know about you, but that sounds entirely fair and acceptable to me.

h.q.droid wrote: Mon Sep 01, 2025 4:20 am Well, plastics is an entirely different game than coal. My final solution for plastics (for legendary science) is to mine quality coal, throw away normal and uncommon, upcycle rare to epic, put them in a cryo with 8 quality, and destroy all epic plastic with recyclers. That's the only solution with tolerable UPS / bandwidth for me. LDS upcycling is too heavy inserting all the copper into recyclers. I can't even upcycle epic plastic bars due to quality recyclers being clunky. Even throwing them into lava was too slow due to lava not stacking.

On grenade upcycling, it's not just bad at resource efficiency if you don't use the final grenades. It's also slower than both upcycling coal directly and upcycling LDS in terms of UPS and bandwidth. Investing A to upcycle B may sound interesting but in the end it's usually impractical, unless you have a fast recipe like EM plant, foundry, or nuke.

For infrastructure, LDS is the way to go. Space coal / quality mining can help by starting at a higher quality.

"destroy" ?! you mean upcycling the epic plastic to legendary ?

That seem to be the general consensus that asteroid shuffling is bad for UPS compared to ore mining when you have large level of productivity, which overall makes the game progression interesting i believe, because there is several alternate ways to make quality that unlock over the course of the progression.

I see no problem in having methods that only works in late game be "very strong", like defending the base becomes easier with new weapons, expanding the factory becomes easier with faster bots, making quality also becomes easier, first with some very basic upcycler for one item , like EM plant , then temporary build like asteroid shuffling, that are situationnal, because of the trade-off throughput /UPS, they don't stay viable into the latest stages of the game where the high mining productivity allows "simpler build", like it was the case already in factorio for a long time, before space age, megabases reached a point where mining drills were used to fill the wagon directly when mining productivity was researched enough that it was no longer necessary to use belts and group the ouput of several mining drill.

Of course there will always be some fake purist asking for the game to be much harder than it is, sometimes they already make the game harder for themselves by using bad methods and advocating the others use them too. There are plenty of mods that actually make the game harder, but for some reason no-one created a mod that removes quality module in crushers. Maybe i haven't found one, but my guess is more that this is terrible idea to just remove something that many players find interesting without proposing something to make up for it. ( judging by the amount of views and interaction people have with space casinos, spreadsheet, discussions and so on).

mmmPI wrote: Mon Sep 01, 2025 9:34 pm "destroy" ?! you mean upcycling the epic plastic to legendary ?

I meant destroying them for nothing It's like a ~4% yield difference and not worth the logistics hassle or the UPS. I tried throwing them into lava too but that was actually slower than recyclers. To think I mourned every rotten fruit at the start...

mmmPI wrote: Mon Sep 01, 2025 9:34 pm That seem to be the general consensus that asteroid shuffling is bad for UPS compared to ore mining when you have large level of productivity, which overall makes the game progression interesting i believe, because there is several alternate ways to make quality that unlock over the course of the progression.

They buffed the UPS significantly in some recent update. Also when used for legendary iron ore specifically, they were better than ore mining before the buff considering that you get ~5 legendary ore per asteroid. They are indeed worse than ore mining for coal though. The key trick is don't get too greedy with asteroid density. The cost of turret aiming is at least the total number of asteroids / in-range turret pairs, which is effectively quadratic. So a fleet between Vulcanus and Nauvis can be better than a single big ship to shattered planet, or even Aquillo.

Asteroid shuffling also has a subtle benefit of producing iron in the easiest-to-transport form: legendary asteroids. Each rocket of that gives you 6k legendary iron plates which you can use on smaller ships.

mmmPI wrote: Mon Sep 01, 2025 9:34 pm I see no problem in having methods that only works in late game be "very strong", like defending the base becomes easier with new weapons, expanding the factory becomes easier with faster bots, making quality also becomes easier, first with some very basic upcycler for one item , like EM plant , then temporary build like asteroid shuffling, that are situationnal, because of the trade-off throughput /UPS, they don't stay viable into the latest stages of the game where the high mining productivity allows "simpler build", like it was the case already in factorio for a long time, before space age, megabases reached a point where mining drills were used to fill the wagon directly when mining productivity was researched enough that it was no longer necessary to use belts and group the ouput of several mining drill.

The same! Asteroid shuffling is an amazing design that necessitates multiple iterations of the optimal strategy:

- Transition from reprocessing to basic processing at asteroid productivity 26
- Transition from upcycling rare / epic byproducts to throwing them away at UPS / clogging bottleneck
- Transition into full-quality plastic making at plastics productivity 30
- Transition to coal mining at enough mining productivity
- An envisioned future transition to calcite mining

Ore mining also has much more depth than simple upcycling:

- Transition to mining into rocket silo, then into train-crushed-cars
- More efficient crusher train design at rocket fuel productivity 30 where legendary fuel can be mass-produced locally
- An envisioned future transition: mining into railgun-shot cars / chests (you inspired me that overgrowth soil can make eggs with predictable spoiling progress on Vulcanus with high-enough throughput)

I briefly mined into train wagons on Fulgora but at that time it was already too slow for me

Upcycling EM plants though, is one of the best recipes end-game. It actually has better yield rate than asteroid shuffling, though lower throughput due to requiring holmium and lower "density". Kinda sad I haven't found a use of the byproducts yet due to the low throughput and the high transportation cost.

mmmPI wrote: Mon Sep 01, 2025 9:34 pm Of course there will always be some fake purist asking for the game to be much harder than it is, sometimes they already make the game harder for themselves by using bad methods and advocating the others use them too. There are plenty of mods that actually make the game harder, but for some reason no-one created a mod that removes quality module in crushers. Maybe i haven't found one, but my guess is more that this is terrible idea to just remove something that many players find interesting without proposing something to make up for it. ( judging by the amount of views and interaction people have with space casinos, spreadsheet, discussions and so on).

Agreed. Legendary science is like Chekhov's gun. If it exists, it should find a use. It's more canon than recycling grenades or blue undergrounds

h.q.droid wrote: Tue Sep 02, 2025 3:23 am Agreed. Legendary science is like Chekhov's gun. If it exists, it should find a use. It's more canon than recycling grenades or blue undergrounds

For me it's a bit deeper. There's a logical fallacy in assuming you have an infinite amount of resources and the law of large numbers will somehow "make everything pan out". In practice you have a weekend. You always finite time.

If I where to seriously say to an uninformed third party, like say the bosses I met who shipped product "there's a research bonus of 6 times that I buy for a deficit of parts". They would transfer me to the "lift-heavy-objects-only" section of the warehouse.

If there's recipes that show geometric scaling around, I can make a nuanced argument that "there's research bonuses that are affected by this scaling, and research bonuses that do not." In order to make that argument I would have to collect data and make evidence based reasoning. But at the end of the day, if the longevity bonus gives: ( 6/x < 1 ) the game just won't survive social criticism. It's not that "the myth of progress" is a myth.

And if some weird quirks of computation granted enhanced UPS, it's part of the meta-game to say "This dev, who I've selected as a strawman because... reasons...
(opinion that may also get me sent to the lift-heavy-object-only section). (Obligatory citation to Knuth's Art of Computer Programming)." Carry on. I'll succeed and fail on the basis of my arguments.

You'll note I made a nuanced argument, if this can demonstrate harm because a player is chasing down a bad lead. The thing is if the player chased down a good idea, I have room to make note of it and adjust. There's six different rates for rerolling and when you look at it on a machine by machine basis... iron works. And less enthusiastic about sulfur because... I've been able to chase down 1 hypothetical method. And it's so technically complicated it's more of a joke than something to be examined.

The "nerf" is just to come up with better recipe chains for those items which don't abuse a players time.

h.q.droid wrote: Tue Sep 02, 2025 3:23 am The same! Asteroid shuffling is an amazing design that necessitates multiple iterations of the optimal strategy:

I admire and respect the amount of effort and thought you’ve put into your run, I really do. But surely you understand that the vast majority of “space casino” users never use it at this scale or make these kinds of optimizations. They build a simple, unoptimized ship and use it at low asteroid/LDS/plastic productivity. And it still produces an appreciable output for how little effort went into designing it. Your very well thought out and high-effort quality science run isn’t the target of this balance proposal, but rather it’s the masses of average players who just heard “space casino is good” online or from a youtuber. And they use it in their games to essentially erase the complexity and depth of the quality mechanic at no cost.

And while ore recycling might be the next go-to “simple” quality method, it actually has consequences. Unstrategic use of ore recycling will drain resource patches, actually punishing players for avoiding complexity. And at the very least players need to occupy ore patches, while space platforms can be built and operated indefinitely without any kind of territorial expansion or depletion over time.

mmmPI wrote: Mon Sep 01, 2025 9:34 pm but for some reason no-one created a mod that removes quality module in crushers. Maybe i haven't found one

I know that there are at least two. There is a mod for any kind of factorio player, and it’s completely certain that “easy quality” mods would be instantly created if vanilla quality was ever fixed. Like I said before, nobody would lose anything. People who want to obtain quality items quickly and with little effort will simply install these mods and happily play as if nothing has changed. And of course, so will the rare few like h.q.droid who have managed to find some real depth in the mechanic.

coffee-factorio wrote: Tue Sep 02, 2025 1:43 pm For me it's a bit deeper. There's a logical fallacy in assuming you have an infinite amount of resources and the law of large numbers will somehow "make everything pan out"

Not everything, but it's fairly easy to math out when you can expect stabilization and with which % of accuracy. The logical fallacy is there for you only if you don't understand probabilites and law of large numbers.

CyberCider wrote: Tue Sep 02, 2025 1:48 pm And while ore recycling might be the next go-to “simple” quality method, it actually has consequences. Unstrategic use of ore recycling will drain resource patches, actually punishing players for avoiding complexity. And at the very least players need to occupy ore patches, while space platforms can be built and operated indefinitely without any kind of territorial expansion or depletion over time.

No, ore recycling when it becomes more viable in UPS than space platform do not deplete ore patch; At this stage of the game it's difficult to keep a mining drill running full time.

CyberCider wrote: Mon Sep 01, 2025 5:48 pm

h.q.droid wrote: Mon Sep 01, 2025 1:58 pm Blue undergrounds are fast but you have to first put gears into those foundries (and somehow recycle them into something useful later), which will eventually bottleneck the setup. Of course, neither can hope to compete with asteroid cycling for iron ore.

I find that kind of hard to believe. The foundry is a big building, and legendary stack inserters are fast. And the results I obtained were based on an unbeaconed upcycler, so there is plenty of room all around the building for very many input belts. And yes, I know that ore recycling is typically slightly beaconed, and I did test it with that. But I left the upcyclers unbeaconed. Perhaps if I also beaconed the upcyclers to match, they would have an even bigger advantage? I think I might explore this possibilty.

It's fairly easy to test for yourself though, you don't have to "believe", you can just verify

CyberCider wrote: Tue Sep 02, 2025 1:48 pm
I admire and respect the amount of effort and thought you’ve put into your run, I really do. But surely you understand that the vast majority of “space casino” users never use it at this scale or make these kinds of optimizations. They build a simple, unoptimized ship and use it at low asteroid/LDS/plastic productivity. And it still produces an appreciable output for how little effort went into designing it. Your very well thought out and high-effort quality science run isn’t the target of this balance proposal, but rather it’s the masses of average players who just heard “space casino is good” online or from a youtuber. And they use it in their games to essentially erase the complexity and depth of the quality mechanic at no cost.

I'm of the opinion that one should not need much effort to get some infrastructure after legendary. Vanilla Factorio's post-victory game is very grindy and repetitive. Quality seems to be intended as an extension of that. As a result, many players simply stay away from it because how unrewarding the original design is. You can deduce that from the many screenshots without a single quality building. If one wants to grind, why do simple upcycling instead of a vanilla megabase? That logistics puzzle is more rewarding and less annoying. Fixing recycler lines clogging is not fun at all as any player getting to that point has already experienced it in Fulgora.

The simple, unoptimized ship giving you legendary infrastructure is a very nice departure from that. It gives fast reward for the 5000 science unlocking legendary. I'd have preferred a whole legendary ship to be towed back to Nauvis from Solar's Edge which would reward the "victory" too.

More importantly, the instant legendary infrastructure allows spaghetti play styles to survive smoothly. With legendary everything you can get to ~10k SPM with spaghetti only. Not having to modularize, expand or grind in Space Age makes it fun for a broader player base, including me. I would never have built the legendary science base without my 6-crusher legendary platform. My legendary science base used to look like this: https://factorio.com/galaxy/Iron%20I:%2 ... X3/planets.

h.q.droid wrote: Tue Sep 02, 2025 3:37 pm My legendary science base used to look like this: https://factorio.com/galaxy/Iron%20I:%2 ... X3/planets.

This base doesn't appear to have created a legendary item at all

Shulmeister wrote: Tue Sep 02, 2025 3:57 pm

h.q.droid wrote: Tue Sep 02, 2025 3:37 pm My legendary science base used to look like this: https://factorio.com/galaxy/Iron%20I:%2 ... X3/planets.

This base doesn't appear to have created a legendary item at all

Exactly, that's about the state at which I unlocked legendary. My point is, grinding for enough infrastructure without asteroid reprocessing from that state won't be fun.