Friday Facts #305 - The Oil Changes

[IronCartographer]

Maybe a bit of a provocative question:
...
Or put in another way - why are you certain that adding solid fuel to the chemical science pack was the wrong approach?

The solid fuel was not a mistake, but sulfur fits the new paradigm of direct-to-gas better by not forcing use of the inefficient gas->solid recipe prior to other oil fractions becoming available.

It also crafts much faster and requires less gas than solid fuel.

Finally, it will inspire the distribution of solid sulfur to more places than before, with creative benefits. Solid fuel already had a role in extensive fuel distribution systems, along with being an ingredient in rocket fuel.

[Silfir]

There's no question that requiring sulfur is better than requiring solid fuel made from petroleum gas. But both are worse than requiring solid fuel made from light oil.

It makes blue science much worse to deal with post-AOP, since now you lack the solid fuel sink that blue science was supposed to be - for the entire experimental branch.

[Tricorius]

what happens when they do eventually research advanced oil processing, then realise they're going to have to completely redesign their base, because they didn't consider that you'd eventually have to add something to those 2 additional outputs on the refinery?

I think it is even worse than that. As mentioned by others, the most basic way to get oil to a line of refineries is by striping a line of pipes right above the input. So, it is covering the water input.

A new user is going to have to rearrange that set of inputs when they want to switch to advanced.

In addition, I’m guessing most newcomer setups would stripe a similar line of pipes across the bottom to get all the petroleum out.

The entire pipe system going to and from the refineries will have to be redesigned. Probably not a huge issue, but likely pretty annoying. (With the newer “no mixing” code, could you even change that refinery to advance oil? If not, is the error message going to be able to concisely relay the problem?)

For the “rebuilding is a way of life in Factorio” club...yes, that is one thing I love about Factorio. *Most* of the items in the game can just be picked back up and plopped right back down in a better arrangement. But fluids don’t work that way. Keep in mind that when this problem hits you will probably have a fairly saturated oil system, so there isn’t anywhere for the fluid to retreat to as you begin deconstruction, so you either lose a lot of resources or have to do some weird “pump everything out and around” method to rescue it. I was pretty pissed the first time I picked up a tank full of oil without really thinking through it, plopped it back down, and it was empty. (And I want to say, though my memory can be fuzzy, that this was a time when fluid tanks stored way more than they currently do. Even if not, losing 25k of an infinite, but deteriorating resource hits you in “the feels”...every time.)

I’m not super hot on the whole “flare stack” solution, but there *is* a built in flare stack graphic on the refinery itself. It seems to me that the refinery could just automatically burn off excess fluid when it builds up. (Maybe the flame color or smoke color changes? “Hmmm...that is different...why is it a different color?” one might ask.’)

If doing it automatically isn’t desired, then perhaps a UI option could be added to the refinery’s data pop up with a “burn excess” sort of toggle for each output. Maybe it only shows up if the refinery deadlocks. :: shrug :: I dunno. I’m not in love with that either. But it seems simpler than a fairly large restructuring of oil.

I’m not sure how to do this without it possibly being confusing, but I’d like to see some “foreshadowing” of the advanced recipe and the additional I/O. It might be interesting to be able to see a “ghosted” version of the advanced oil recipe in the recipe selection tab. When in “ghost recipe” mode it could show the current inputs/outputs in full color and apply the “ghosting” (alpha channel) to the advanced recipe inputs/outputs. At least it might then hint that you’ll eventually be able to do more with the refinery building than you currently can.

[dood]

If nothing else, basic oil processing should be way more efficient if you just want petroleum so there's a reason to build it at all later on.

It's way too attractive as it is. Many people will just produce all their plastic with BOP and skip most of the cracking. Yes you lose some product, but with the massive bonuses from mining productivity that doesn't matter.

Many people do inefficient, weird things, yes.
Your point?

[F_W]

There's no question that requiring sulfur is better than requiring solid fuel made from petroleum gas. But both are worse than requiring solid fuel made from light oil.

It makes blue science much worse to deal with post-AOP, since now you lack the solid fuel sink that blue science was supposed to be - for the entire experimental branch.

You've hit the nail on the head imo. Adding 10 units of light oil to the rocket fuel recipe is not going to balance its uselessness. People will be using BOP for PG (because it's simpler and more UPS friendly than AOP) and coal liquefaction for lube. There won't be any reason to use AOP. I don't understand what the devs are thinking, and it seems I'm not the only one. There have been many other better suggestions than what the dev's are proposing, but they are all being swept under the rug

[IronCartographer]

It makes blue science much worse to deal with post-AOP, since now you lack the solid fuel sink that blue science was supposed to be - for the entire experimental branch.

Sounds like you'd better get cracking, then.

Please pardon the pun!

[Roxor128]

I'm still not convinced.

A few things I think you should try BEFORE changing the oil recipe:

1. Add blockage indicators to the UI. Red background on the blocked output in the machine window, red ring around the icon for an output product in Alt-view, and a red lamp above the blocked output on the model. I think this is probably the most important one to address.

2. Add a tutorial for the current oil system. This combined with number 1 might be enough to solve the problem of the Oil Spike.

3. Require the player to make something with a fluid input before oil, not just the power plant. The obvious candidate being concrete, which could be required to make the refinery or chemical plants. Very little actually needs it right now, to the point where I forget it's even a thing until I need it for the rocket silo or I decide to use nuclear power. A few suggestions in the previous thread built on this idea and had basic and advanced concrete, with the latter needing lubricant to make and being needed for the silo or nuclear power.

4. Split oil extraction and oil refining into two technologies. Let the player work on figuring out pumping and transporting crude oil (along with something that can use it directly) while they work on researching how to refine it into something more useful. Crude oil is supposed to be sticky, so maybe it could be used for traps to slow down the enemies, which could then be ignited by burning streams from a flamethrower.

If after implementing all this, players are STILL having trouble with oil refining, THEN go and change the recipe.

[vampiricdust]

“Pushing robots back is not fun.”
This was pretty much my first response when I first heard about the whole idea, but now I’d say it’s really not that huge difference between late logistic and early blue science, especially with the basic refinery being quicker to set up. I will be observing this very closely what exact effects will it have. We were considering to add burner powered construction robots a few years ago, but I can’t currently see how and when would those appear.


“Petroleum gas has too many uses.”
Anything that is not from petroleum gas could introduce a need for reverse-cracking or forcing to make for example solid fuel from petroleum gas in order to save light oil for something else and keep the refineries running. I think it’s ok.


Thank you very much for all of your replies. All of them.

V

<3 you all, thanks for the game.

First, if you add in burner or perhaps slower construction bots that are not so expensive, I'll be happy. Construction bots take so much of the frustration of oil out as you can mass deconstruct things. It makes redoing setups to find a good build so much less frustrating.

There should be sinks for light and heavy oil before cracking so players are not frustrated by having product that can't use. You could make solid fuel, but you don't use that much solid fuel until late game. Player will not need that much fuel for power generation, especially since they might be going into solar instead anyways. So players have been stuck with 2 fluids that stop the 1 fluid they need and no way to have it consumed without destroying it or storing tons of solid fuel.

Trying to move the problem is only going to cause a massive rebalance which this FFF has addressed. All you guys needed to do was add a resource sink for light & heavy oil such that new players can easily find without being so much that it out paces PG. I personally would add a special furnace that gets faster smelting times when provided with oil. Each oil giving it different speed benefits, but taking any of the 3 oils so it gets consumed. This solution would help with the mid game resource increases as well as giving an optional sink for players instead of destroying liquids or restructuring the game.

[Reika]

I personally would add a special furnace that gets faster smelting times when provided with oil. Each oil giving it different speed benefits, but taking any of the 3 oils so it gets consumed. This solution would help with the mid game resource increases as well as giving an optional sink for players instead of destroying liquids or restructuring the game.

This is an interesting idea, though i will counter it with the fact that I rather doubt most smelting setups are built to accommodate piping. It also feels far more modded than anything else so far suggested.

[vampiricdust]

This is an interesting idea, though i will counter it with the fact that I rather doubt most smelting setups are built to accommodate piping. It also feels far more modded than anything else so far suggested.

It's a new puzzle! I would make the furnace a 3x3 furnace and so piping would not be anything different than dealing with say plastic making. You need input & output belt with a pipe.

I suppose it would need an intermediate product to simplify the liquid mixing. It would also serve as a foreshadowing of module benefits and the eventual electric furnaces.

[gamestefan]

*snip*
As for oil products, my sugesstion is to keep current science, flamer fuel and rocket fuel recipes as is, but change following:
1) Change Petroleum Gas input in Sulfur recipe for Light Oil. This, along with Solid Fuel, gives Light Oil a major resource sink, so backing up should be less of a problem. This way, Plastic uses Petroleum, Acid uses Light, and Lubricant uses Heavy Oil.
*snip*

Having read every word on the posts about the oil changes (FFF#304 and FFF#305) I came to a similar conclusion to this as well. At the very least, have the refinery spit out at least 2 items - definitely not 1 - and the player is more likely to have some room to add a third output. I actually think HO and PG (leftmost and rightmost) outputs are optimal, as the player is likely to be hinted that something will come out the middle, especially if they've looked any way down the tree to see that LO has a use.

Having nearly 2000 hours in game (since version 0.14.x), I have to admit I don't remember my first oil build well, but I know I launched a rocket in my first game (eventually!!) and went to youtube to get some ideas on how to optimise it.

Exactly this, I am not in favor of oil changes as described in the FFF. The proposal above sounds much better to me. If the player would get at leat 2 fluids the player will have to deal with multiple outputs earlier. I like giving HO and PG in the basic recipe and moving sulpher to light oil. Sulpher and anything made from it can then be moved behind advanced oil processing. This will also solve the issue of having too much recipes to deal with at once.

It would also be nice to add a bit more tutorials. For example one could be simply telling the player that any of the oils can be "burned" when converted to solid fuel. I remember solving oil based on some very nice tutorials from youtube.

Factorio is a great game especially for it's price, so thanks devs and keep up the good work

[Ryba666]

This change is bad. Problem was never for new playars a fluid output. I watched a several lets play youtubers who was new for factorio, and when they have a bottleneck of fluid, they simply add storage tank. When they reached advanced oil they only add water. In this version, new players would shock when they see additional fluids from rafinery and necessity to rebuild rafinery complex without robots, rebulding in this way is much more frustrating than adding a water.

So this idea is stupid like removing oil , no oil, no problem for new players

[silenced]

There is a rather simple solution to all the confusions that may arise. Just have fixed refinery inputs and outputs for everything. When having only basic oil researched, have only crude oil input on one side allowed, instead of allowing both inputs get crude oil. Same for the output, fix the output for basic oil according to how it may be with advanced oil, so, instead of 3 times petroleum gas outputs for basic oil have only one output where it also will be once advanced oil is researched.

This also prevents the ever evident problem with not being able to change recipes due to 'cannot mix fluids'. Which are in my eyes the biggest causes for confusion and problems.


Help the player, especially the new one, by setting up his oil somewhat right from the beginning, without actually telling him. A new player won't be mad about having only one input for crude oil, instead of two, or have only one output for the gas with basic oil, he won't even realize that the refinery usually has two inputs and three outputs. He will realize that once he goes into advanced oil and he will be like: 'Oh, that's practical! I only need to add water there and build more pipes.', and, he can change the recipe without issues due to fluid mix problems.

Life can be so easy.

[Yandersen]

...

Thank you for a nice well-argumented response! I tend to agree with many of your points, but I still hardly see how some major issues got solved with proposed changes to oil processing.

The major IMO, is the difficulty spike related to balancing multiple output problem that will happen sooner or later. As blue science can be achieved based on BOP now, it is right to expect that the beginners' base will get designed around it. Considering the relative base' size increase at that stage, future redesign required to move to mandatory AOP will feel more like a frustrating rage-quitting spike rather than interesting optimization process or fun puzzle solving.

I see the way to avoid it. And in the same time, force the player to learn the multiple output blockage concept from the very beginning. Postpone the refinery, and replace basic oil processing with "Petroleum gas extraction" (PGE). This is a recipe for the chem plant, it takes crude oil and outputs petroleum gas and, again, crude oil itself. For example, like this:

5 Crude Oil -> 4 Crude Oil + 1 Petroleum Gas (0.5 sec)

This will force the player to immediately face the problem of multiple fluid output problem and solve it by usage of pumps and circuit logic - the knowledge he will need in time when the Oil Processing in the refinery will inevitably strike down.

This will actually break down the learning curve spike and smoother it, not just postpone for later time, when it will be even harder to deal with (all at once, huge base redesign...).

As for BOP and AOP in this case - leave AOP only (then it can simply be called "Oil Processing"). This tech should unlock the refinery building. The outputs should be changed/rebalanced then - either light and heavy oil only, or all three with PG output produced in a relatively VERY small amount.

As for blue science pack requirement, I strongly agree now with the change of solid fuel requirement to sulfur. But I think that also changing red circuits to plastic may further smooth down the difficulty spike. Possibly wrong idea, not sure...

[Venrob]

This is clearly the worst change factorio has made. Putting the complexity off is NOT the answer. That just means people will set up a refinery in a way that is wrong, but it will work PERFECTLY.... until they unlock a new tech, and suddenly realize "Crap, all that time was a waste, gotta tear this ALL down and rebuild it better". It's better to simply give the challenge up front, so that you don't get the false positive that this new system would give you. I don't like a single one of the recipe changes coming along with this either; requiring fluid for rocket fuel? Removing solid fuel from blue science, which, was a very GOOD addition, now just being undone? GET YOUR CRAP TOGETHER.

[FuryoftheStars]

“The gui should explain it better.”
When we were the testing the NPE/tutorial/introduction, we added the “Status” in tooltips and the yellow/red icon backgrounds to indicate why the entity is not working. If I remember correctly the coming entity GUI redesign will integrate mainly the status to make it much easier to see. Good point regardless of which version of oil processing we would have.

Many people unfortunately don't seem to notice "subtle" hints like the flames stopping on the refinery's stack, the status text, or a highlighted background of the output icon in the gui. You need things to flash at them to pull their attention.

“Flare stacks.”
“Add 3 separate recipes with one output each.”

Agree, don't add these. Leave those for mods.

“It has been working fine for years.”
Has it? Sure, most of us here figured it fine more or less. But also many of us report that oil is tedious, and some of us aren’t here because they have already quit because of it.

My only problem with this reasoning is that there are actually a world of reasons why people may quit playing. Considering the pace of the game up until oil, it could be this is the point where they decide this game just isn't for them. Maybe they started installing mods, which disable achievements (and yes, I actually fall into that category... I only have 7 achievements. Oil is one of them... launching a rocket is not). Point is, you really can't use this as a metric for determining a point in the game where things may be getting too hard.

“Advanced circuits are too advanced to require them this early in a science pack.”
I find this a fairly valid point, but almost all of the things that chemical science pack unlocks do require advanced circuits, and the future science packs build on top of that complexity (and of course the amount of advanced circuits you need) further.

Going along this route and that I also see general complaints on the complexity (not just advanced circuits), I believe the current requirements for chem science is something like 3 advanced circuits, 2 engines, and 1 solid fuel. Have you considered replacing advanced circuits with plastic, and maybe reduce engines to 1? Of course, question... if the chem science is supposed to be about mastering oils, what do engines have to do with it?

“Basic oil processing should only output heavy oil and unlock both cracking recipes.”
This would mean that if the player is not going for early robots, they would just have to do 2 more cracking steps for all of the petroleum gas. That is 2 extra steps adding to the tedium and vastly decreasing the puzzle of advanced oil processing - a typical new player is really not going to set up a circuit network, and the AOP solution would not be different in anything, just more efficient per crude oil.

I agree that it should not be HO or LO output only and then introduce cracking early. But I also dislike PG only (in part because I feel like it simplifies things too much, in part because of all the things that must be moved back behind blue science now, and in part because it just straight up feels wrong). Have you considered only pushing LO to advanced oil? This would cut the complexity down (reduce the number of fluids being handled from 3 -> 2), but still allow the player to make everything that is needed now. HO should be the most efficient use for solid fuel, but then make rocket fuel just take straight liquid LO so that LO still has a primary use.

“Put sulfuric acid, or sulfur and water in chemical science pack recipe.”
I really dislike any science pack to have a fluid input, just looking at the crafting menu and seeing red background is alarming.

Not that I'm advocating for a change to the chem science to this (though that wouldn't be a bad idea to replace engines with it, imo), there really isn't anything wrong with having one show red. Would be a change of pace to have a science require a fluid. But this one isn't an important topic to me.

“They are trying to overengineer the game, until it is no longer the game I fell in love with.”
This is incredibly painful to read, and yes we do consider this at every corner. I believe the changes we do overall are going in a good general direction, and if some version as a whole would be majorly superior, there would be more “0.12 recipes” stule mods and they would be much more popular, assuming we aren’t doing something awful in the non-moddable parts of the game, then people would just stay on older versions.

I do feel this way, but the feeling comes from a complex origin. It's not only because of this change, and it's not just you guys. I've had many games that I've played over the years that I bought and played specifically because of certain unique features that would later be nerfed/removed, quite often to the point where I lost interest in the game because what I loved most was gone and they became indistinguishable from every other game of their type out there. Why keep playing if what you like isn't there anymore? After a while, it starts to get old and frustrating. I'm starting to have the feeling inside "welp, here's another one". And please don't take this as an angry or ranting post. Emotion is lost in text. I'm just a little depressed.

“Petroleum gas has too many uses.”
Anything that is not from petroleum gas could introduce a need for reverse-cracking or forcing to make for example solid fuel from petroleum gas in order to save light oil for something else and keep the refineries running. I think it’s ok.

PG really does have too much pulling from it. Prior to chem science after unlocking oil, there really is a very high demand on PG, yet all 3 fluids are produced at near equal rates. This really is what causes the backup. To go with my suggestion above on having BOP do HO & PG, moving LO to advanced, you should move sulfur to HO. Plastic will still have a higher demand than sulfur, so there shouldn't be any situation where you're low on HO but surplus on PG once everything is actually set up. You can tweak the production numbers a bit on advanced oil to help, but once the player unlocks everything, they can easily mix and match the processes available to avoid "needing" reversed cracking. Though, to be honest, having reverse cracking would not be a bad thing, and actually exists (alkylation).

“The wall is the ever growing complexity of science packs.”
“The wall is the ever growing resource demands of science packs.”
I can’t really imagine how would the progression be interesting if the science pack steps would be about the same price and/or complexity increase steps. In the beginning it would just be insane, and in the late stages just a chore. Should there be smaller steps? That’s a tough question, but I think 0.17 has improved on the distribution of these steps quite significantly.

When folks say these things, I think what they are more trying to say is that the jump from red/green to blue is significantly greater in time & complexity compared to the previous/than what it should probably be.

“The new recipe does not make sense as it is not realistic.”
It is always great if we can stay believable or compatible with real processes, but gameplay logic should always come first in my opinion, unless it’s converting raw fish into rocket parts.

As I stated earlier, part of what attracted me was how many things were striving for the concept. While we're not dealing with exact realistic ingredient needs and results, the concepts are still there. We're not mining a single "metal" resource that then is somehow usable for everything. Putting crude in and getting only PG out almost feels like that.

Especially if you read this far, thank you very much. Hopefully you have found some answers, and even more hopefully you understand our aim is not to ruin the whole game, and that making a change does not mean it can never be changed, altered or reverted.

Thank you very much for all of your replies. All of them.

V

While I still don't understand, thank you for replying.

[Serenity]

There should be sinks for light and heavy oil before cracking so players are not frustrated by having product that can't use. You could make solid fuel, but you don't use that much solid fuel until late game.

They did add solid fuel to blue science partly for that reason. You didn't need much, but a full assembly line of science could use the output of a single chemical plant.

Many people do inefficient, weird things, yes.
Your point?

Over your head

[BlueTemplar]

[...]
I even started my own YT channel because of this game and am currently nearing my 200th episode of my third map.
And every time something like this changes, everything gets f*cked up.
It gets a little frustrating, to say the least, at times.
And I'm sure I'm not the only one who is building ginormous setups, with hundreds of hours invested, only to have it become broken/useless because of something that didn't need an update.

But I guess I can already see the arguments and the, often so eloquently made (*sarcasm), objections now:

1: It's beta you are not allowed to have expectations. The game can change at any time.
2: You don't like it, don't play it.
3: Just don't play experimental.
etc etc bla bla bla bla.

I started to already write answers to the arguments but then I stopped, because I feel people who bring this up, don't get it and don't want to get it anyway.
[...]

The game is not even in beta (= all major features have been done, only bugfixing and balancing is left).
It's in alpha.
Nobody forces you to update. Even if you're using Steam, AFAIK, you had to specifically pick the 0.17.x "beta" version, which is the only one of the "betas" that will force auto-updates.
Also, you can always roll back to a previous version if you feel that the new changes break too much. (AFAIK, Steam makes it even easier than the directly available website links ?)

You want more eloquence ?
Well, how about this example :
The just released 0.17.59 changed the balance of green circuits in expensive mode.
(And what felt to be out of the blue to me, though I might have missed some balance discussion ?)
And, unlucky me, I had just redone my whole circuit production !
Have you seen me complain about that change ?
(Of course, people using expensive recipes are in the minority... but so are those using experimental... just a larger minority.)

[...]
Additionally, [this change] breaks existing setups. Any proposed change that breaks backwards compatibility with previous versions of the game needs to have an extremely good rationale for implementation, especially as part of a minor version change. If this was for a 1.0 or 0.18 change, or fixed some critical bug or exploit, then fine. Not acceptable just for a balance tweak, especially in the middle of the 0.17 release cycle.

As I already said in the 0.17.59 thread, this is pretty much to be expected just(?) before the first 0.17 stable. It's after that it would have been (more) problematic.

However, Silfir does have a good point :

Maybe a bit of a provocative question:

The players you're worried about - the ones that might get frustrated trying to solve the multi-input problem - aren't they playing on stable right now? The vast majority of them? It takes at least some savvy or independent research for players to opt into the experimental branch, I would imagine.

That means you haven't even exposed those players to the chemical science pack - which I imagine was already an attempt at solving the problem you're now trying to solve again.

Isn't it worth first introducing the chemical science pack to Stable - the default install - and checking how new players are handling it? You're about to introduce a slew of drastic changes to how oil processing works, after all - isn't that more suited for the version 0.18.0 of the new experimental branch?
[...]

Especially considering that the GUI rewrite isn't finished yet, and that the missing tutorials are planned for 0.18 !
This would also fit with the "Final game balancing" planned for 0.18 !
(Maybe should even be postponed to the second 0.18 stable, to see how players reacted to the new tutorials ?)

[jodokus31]

“Pushing robots back is not fun.”
This was pretty much my first response when I first heard about the whole idea, but now I’d say it’s really not that huge difference between late logistic and early blue science, especially with the basic refinery being quicker to set up. I will be observing this very closely what exact effects will it have. We were considering to add burner powered construction robots a few years ago, but I can’t currently see how and when would those appear.

It is a huge difference. Why? With your proposed change to oil you pretty much force players to completely rebuild their oil setup. And in addition you take away the possibility to automate that process by moving construction robots behind that burden.

I don't think, that you have to rebuild oil, if you account for AOP before.
However, I think it's a real problem (the rebuilding part) for newcomers, while con-bots might not be the highest prio for them.

“The new recipe does not make sense as it is not realistic.”
It is always great if we can stay believable or compatible with real processes, but gameplay logic should always come first in my opinion, unless it’s converting raw fish into rocket parts.

@Wube: You should not underestimate the impact on many people, if realism suffers in favour for gameplay (esp. if it's not accepted as improvement for whatever reason). I'm not into chemistry very much, but the oil changes seems weird for me from this perspective. The current BOP feels much more believable...

[Orum]

I'll echo the sentiment that while this is better than the previously proposed changes, it still addresses the "problem" in the wrong way. Namely, if the concern is complexity, this just delays the inevitable, and will encourage new players (who will likely not set up their piping correctly for Adv. Oil Processing) to stay on basic when they can't select AOP because of the fluid mixing potential. Instead a tutorial for oil, like there is for trains, would be a better solution.

If the concern is there aren't enough things to use heavy/light oil in before AOP, bring cracking forward! Keep BOP as it is now, with three outputs, but give us cracking with the initial Oil Processing research. This would at least allow players to grok how balancing fluids is supposed to work, instead of just dumping everything into temporary solutions that will eventually back up.

But if you're sticking to your guns on the "one input, one output", at least flag the other inputs as what they would be for AOP. That is, label the water, light, and heavy oil I/O, but put something over them to denote they're not used, like a red 'X'. Prevent players from connecting other fluids to them, so at least they can switch to AOP in the future without too much trouble.