Friday Facts #376 - Research and Technology

[papiwill]

Is it possible to keep the old starting technologies or no? Like is that a toggle I can turn on and off or am I forced to do that?

[KuuLightwing]

To get a quality item you REQUIRE a QUALITY MODULE, thus no module no quality item.
Thus unless you are mix inputting an uncommon+ copper plate with normal plates that circuit design will not jam.

Problem is that theoretically by implementing quality at intermediate steps you maximize the chance of end product being of the high quality, so that's probably the most effective way to produce them.

This, however, pretty much forces you to implement them at every single step due to the way mixing qualities works, and you also have to use the recycler still, because this would produce different proportions of low quality/high quality items, because often times recipes mix items that went through different number of steps. The logistics of this just get exponentially more complex, unless you just use bots, then it's becomes trivial.

So I suspect this just naturally gravitates towards standard logistic net design of spamming assemblers and provider chests and recycling everything that's over a certain limit in logistics network.

[Anachrony]

The idea that i cant make oil refinerys before driving to and putting down a pumpjack is absurd.

You've got that exactly backwards. The idea that you can is absurd. We accept that level of absurdity, because it's a video game, but it's absolutely not a thing that makes sense if you think about it even a little.

[Ohz]

Dont hesitate to have a look on how Nullius mod does the research, it looks like your idea, and it is well implemented in my opinion.

[Pirate_Rance]

To get a quality item you REQUIRE a QUALITY MODULE, thus no module no quality item.
Thus unless you are mix inputting an uncommon+ copper plate with normal plates that circuit design will not jam.

Problem is that theoretically by implementing quality at intermediate steps you maximize the chance of end product being of the high quality, so that's probably the most effective way to produce them.

This, however, pretty much forces you to implement them at every single step due to the way mixing qualities works, and you also have to use the recycler still, because this would produce different proportions of low quality/high quality items, because often times recipes mix items that went through different number of steps. The logistics of this just get exponentially more complex, unless you just use bots, then it's becomes trivial.

So I suspect this just naturally gravitates towards standard logistic net design of spamming assemblers and provider chests and recycling everything that's over a certain limit in logistics network.

Agreed on that was part of my edit after I thought about how they may be using Quality modules

EDIT:
Exception would be putting quality modules in the Wire machines.
You can easily fix this via a single loop belt and additional inserters with a recycler.
I now see your point of belted Wire tho as the throughput of quality wire if using it would need to be insane.

Spamming of assemblers with LOGISTIC chests is kinda how lategame malls work atm anyway.
I do suspect that the changes will either make the game more LOG bot based like you said or depending on player it makes it so there will be entire lanes of the bus just sending to recyclers, Cause I dont think bots can handle that much throughput.

[KuuLightwing]

You've got that exactly backwards. The idea that you can is absurd. We accept that level of absurdity, because it's a video game, but it's absolutely not a thing that makes sense if you think about it even a little.

Whether its absurd or not only depends on your framing. Is it not absurd that you can engineer a machine that's designed to extract oil without ever seeing any oil patches?

Also when things being absurd actually mattered in factorio? Carrying locomotives in a pocket? Throwing conveyor belts and robotic arms into a magic assembler producing green potions that are fed into a disco ball that produces new technologies?

[KuuLightwing]

Spamming of assemblers with LOGISTIC chests is kinda how lategame malls work atm anyway.
I do suspect that the changes will either make the game more LOG bot based like you said or depending on player it makes it so there will be entire lanes of the bus just sending to recyclers, Cause I dont think bots can handle that much throughput.

Yes they are, because bots are kinda suited for that kind of stuff, but you don't use them for intermediates, just to assemble the final products. Still though, I'm pretty sure people do belt malls. Hell, I do them too, usually early game. They are kinda messy, but functional enough.

Like I said before I'm concerned that this expansion seems to have a theme of pushing towards very specific style of play overall. If you like building with proper ratios, there's now specific mechanics designed to just throw wrench into that playstyle, if you want to build refinery before organizing oil outpost, no, must build outpost first, and the whole quality thing seems to just so much easier to use with bots than with belts.

[YotaXP]

To hop onto the quality name suggestion dog pile, how about just ditching the adjective altogether and instead using the symbol? Our tabletop group uses symbols to indicate quality, (─,△,◻,☆,✶ instead of +1,+2,etc) and maybe that would suit here as well. Less to worry about for localization (some languages put adjectives on the other side of the noun) and it adds less length to the item name.

[JonGalt]

Also when things being absurd actually mattered in factorio? Carrying locomotives in a pocket? Throwing conveyor belts and robotic arms into a magic assembler producing green potions that are fed into a disco ball that produces new technologies?

This made me laugh more than is logical

[KuuLightwing]

This made me laugh more than is logical

I usually don't like to make this argument, because it could be a fallacy (using one example of something that's not realistic to justify some other thing being not realistic and/or fitting), but Factorio really doesn't go into detailed simulations, and mostly takes broad inspiration from real world.

I kinda see the point of having steel axe without producing any steel being weird, but at the same point that's kind of a band-aid solution, cause it doesn't feel any more logical to get a steel axe by producing steel but not actually making anything with it.

[Bluebrightfire1]

Look,

I don’t really have any expert advice on why it’s wrong, in my opinion, it just “feels” wrong.
What I’m trying to say is that the idea is great, but the implementation is not. I really feel like you are on to something, it’s just the the implementation just doesn’t make me want to play this part of the game.

I think that people are concerned with the complication it adds to the game, however we don’t get to play test the game (like you do) so we don’t have a first hand experience and this may change once we actually get to play so take everything I say with that in mind.

What I think might be a better way would be treating it like modules are treated and removing the randomness. For example instead of having a 1/100 chance or a 1/1000 chance simplify it to a recipe of 100 in = 1 higher quality level 1 and so on.

On top of that if you think about it “upgrading” something to a higher quality should take some “research” to do, so what do we already have in the game for research? Science packs. Using science packs as an ingredient to upgrading our factory with each planet unlocking a new science pack it could correspond to the increases in levels.

Another different idea that I had would be keeping the recycling system as it is but instead of having individual items that improve in quality, have a single “upgrade module” that when combined with a assembler (or other quality improve-able item) would give 100% chance of getting the upgrade of the same level as the “upgrade module” , BUT, to get the “upgrade module” you would have go through grind of R & D by creating one, recycling it, and repeating until you got the better ones which you could then choose to upgrade any machine you liked.

There is probably a million and 1 reasons why my suggestions are stupid, and I agree, however the current implementation is a good idea, but it’s just not clicking with me (and clearly with others too) I don’t think this is a case of going back to the drawing board and starting from scratch, but tweaking the fundamental aspect of the new feature and making it not just a “good” feature but a great one. That is after all, why the original game is so good.

Thanks for reading!

[QuackCow]

really loving this update I've never even played before. Sometimes when I feel stuck in a part of the game and feel like "ughhh i don't wanna start automating green science tonight, i'll do it tomorrow..". now i can just work on getting better qualities for items i have already. Also the start of the game researching is awesome! I really like that you have to do little things to unlock recipes that you normally start with. Makes you feel accomplished. Reading this FFF every week is fun because so much new things are getting announced and i like them all!

[Neelost]

Like I said before I'm concerned that this expansion seems to have a theme of pushing towards very specific style of play overall. If you like building with proper ratios, there's now specific mechanics designed to just throw wrench into that playstyle, if you want to build refinery before organizing oil outpost, no, must build outpost first, and the whole quality thing seems to just so much easier to use with bots than with belts.

Well I even went to check and there hasn't been any info released about any quality bonus for science

So... no ?
Most of your factory is directed towards building science so for most of your factory you are not pushed toward quality at all.
Better yet, it encourages you to think about a design for the buildings that you want to have at a good quality level instead of just slaping 2 logistic chests around an assembler 3
The only part of your factory where I could see someone bothering with maxing the quality of intermediary items would be for max quality modules but tbh they seem quite overkill outside of very long & costly recipes like blue circuits

Also if really your OCD forces you to max quality of everything for no logical reason, well just spend billions of ore for making perfect plates and use the recipes that use maxed quality ingredients for 100% chance of maxed quality products and this will be at ratio with no sorting whatsoever.

[agmike]

A better approach would be

This is a very strong wording without providing any substantive evidence backing it up. It might be better in your opinion, but for me it's not. For a lot of other people in discussion it isn't.

Random recipes work well in Factorio. No one is saying they are afraid that they will get unlucky and never get 40 U-235 to start kovarex. Heck, it's possible to have real nuclear power using pretty small processing set up, just by relying on that quite small 0.7% chance.

For quality chance to get next tier is a magnitude higher. If people have no problems with Uranium, I just don't see how they will have problems getting one quality tier up.

That is, of course, unless they overcommit into higher tier without having necessary capacity to support it. But for all intents and purposes this can be a mistake with consequences. In the end, you are going to wait one real day for top quality item or something, regardless whether quality mechanic is random or not. Is this a mistake even, a build didn't work in practice as expected, deconstruct and move on, just normal Factorio things.


The fact that quality works in a different way, requires different approaches to handle is interesting adds new flavors to gameplay. I am eager to try it out in the game and certainly won't be rejecting it based on vague gut feelings.

[KuuLightwing]

Well I even went to check and there hasn't been any info released about any quality bonus for science

So... no ?
Most of your factory is directed towards building science so for most of your factory you are not pushed toward quality at all.
Better yet, it encourages you to think about a design for the buildings that you want to have at a good quality level instead of just slaping 2 logistic chests around an assembler 3
The only part of your factory where I could see someone bothering with maxing the quality of intermediary items would be for max quality modules but tbh they seem quite overkill outside of very long & costly recipes like blue circuits

Also if really your OCD forces you to max quality of everything for no logical reason, well just spend billions of ore for making perfect plates and use the recipes that use maxed quality ingredients for 100% chance of maxed quality products and this will be at ratio with no sorting whatsoever.

Quality bonus for science I think has been shared on discord (IIRC it's like pack lasting like +100% longer per quality level in lab or something) and from what I understand it's not worth it for science at all (though if so, when why make it a thing to begin with? Adding trap options?), but the point wasn't about science or no science.

Also it's interesting that you say things like "slapping logistic chests around assembler" when quality-adjacent mechanics are actually encouraging that playstyle more if anything, considering how much easier it is to sort items by quality with bots if you want to utilize it in intermediates and not just in final products. As for "worth it/not worth it" well, mathematically if you want to produce a quality item, then you very likely will spend less resources by trying to implement it on intermediate steps, rather than just recycling the final product on repeat.

Also "slapping two chests around assembler" style is also going to be able to utilize an odd higher quality assembler here and there better than belt designs too, so again I think that's where quality encourages this playstyle rather than discourage, as you imply.

As for whether it is overkill on modules or not, that depends entirely on your goals, and at the megabase stage it will definitely not be an overkill, but a requirement due to how insanely beneficial productivity is for megabases.

[agmike]

if you want to build refinery before organizing oil outpost, no, must build outpost first

You just need to touch grass oil. I picture an engineer arriving at unknown world, having no clue what's needed and what's not. Stumbles upon oil — my goodness, time to remember how to build that oil processing stuff they taught us in the academy! Seems pretty natural to me.

and the whole quality thing seems to just so much easier to use with bots than with belts.

Your point is not helped by the fact all quality designs shown in FFF are belt based

[draslin]

I'm still not sold on the implementation of quality. Primarily, in what world is a 90% failure rate acceptable when manufacturing a product? Suppose we could build end game rockets of quality, would it really make sense to build 10 rockets just to "ensure" we get one Uncommon Rocket to launch into space?

I think a better way of doing it, is to do so at the bottom of the food chain.

It makes a little more sense to reprocess the raw materials in a way similar to uranium, to produce higher quality equivalent base materials. For example; Iron Ore becomes Steel Plate, Steel Plate becomes Uncommon Steel Plate, which becomes Rare Steel Plate and so on. Failure rates still apply, but you recycle that waste product and get some of the last tier plate back. All this at the source. (This would, I think, also address some of the consequences of net positive output from the recycler).

Then on the assemblers, have a drop down or radio buttons to select the quality you want. It only accepts materials of that quality. And only produces items of that quality. No new recipes. No waste Spidertrons to recycle.

Higher quality items are still gated by randomness, you don't have to make entirely new factory designs, and nothing is being ground into scrap simply because the engineer chose this moment to start huffing paint instead of figuring out why his assemblers aren't operating to spec.

[MEOWMI]

I can only say, I'm very excited.

Regarding the quality naming, I do hope you find some words that fit better eventually! I've liked "exceptional" and "perfected".

[agmike]

I'm still not sold on the implementation of quality. Primarily, in what world is a 90% failure rate acceptable when manufacturing a product? Suppose we could build end game rockets of quality, would it really make sense to build 10 rockets just to "ensure" we get one Uncommon Rocket to launch into space?

I think a better way of doing it, is to do so at the bottom of the food chain.

It makes a little more sense to reprocess the raw materials in a way similar to uranium, to produce higher quality equivalent base materials. For example; Iron Ore becomes Steel Plate, Steel Plate becomes Uncommon Steel Plate, which becomes Rare Steel Plate and so on. Failure rates still apply, but you recycle that waste product and get some of the last tier plate back. All this at the source. (This would, I think, also address some of the consequences of net positive output from the recycler).

Then on the assemblers, have a drop down or radio buttons to select the quality you want. It only accepts materials of that quality. And only produces items of that quality. No new recipes. No waste Spidertrons to recycle.

Higher quality items are still gated by randomness, you don't have to make entirely new factory designs, and nothing is being ground into scrap simply because the engineer chose this moment to start huffing paint instead of figuring out why his assemblers aren't operating to spec.

I like how your suggestion does not fix the main issue you have with the implementation

Regardless, what you described looks like a subset of quality system as implemented. I believe one can use recipes on the bottom of the chains such as iron gears or copper wires to boost quality of base materials.

[Anachrony]

Quality bonus for science I think has been shared on discord (IIRC it's like pack lasting like +100% longer per quality level in lab or something)

Hopefully that's exponential (Q2 2x, Q3 4x, Q4 8x), not linear (Q2 2x, Q3 3x, Q4 4x). Still not necessarily worth it given the numbers we've seen, but at least it's something. Given that the qualities are exponentially more expensive, and extra research pack durability is solely a resource issue, I'd hope the scaling is at least in the same ballpark as how the cost scales. If Q4 is +300% and Q5 is +400%, then it's something like 25% longer lasting for 10x the cost, which is extra not worth it.