Friday Facts #304 - Small bugs; Big changes

[psihius]

Let's get back to what actually matters in Factorio - numbers, ratios and efficiency (all kinds)!

I get that you're being semi-facetious here, and I feel the same way but... not really ?

Taking all that into account, it is still performance and time-wise easier to deal with a basic recipe than setting up AOP and it seems it is actually more performance.

Aaand you've ruined it... (if the main point was hard proof about performance issues)

But there *is* a worthwhile question here : can the new Basic Oil Processing be at the same time :
- Too bad efficiency-wise for a (semi-)serious/experienced player to even consider for late game.
- Still good enough to not be too tedious for early game and first time players ?
(- And Advanced Oil Processing not being the mandatory technology to rush would be nice too, but perhaps this is asking too much...)

On the last paragraph - the beauty of new recipe is that when building at scale (less so for 1k spm, more so when you go to 3-5k spm) the hassle of placing it all and module requirements with AOP are so high that man, it just overrides my "I need to make it efficient" to "I'm gonna save 10 freaking hours".
Also, something I just realised - the productivity and especially speed module and beacon savings are ENORMOUS. It is going into 10-20k of modules saving territory at those scales easy at the lower end of megabase, also means less power draw.

As for performance - well, we can't really test it yet. But I and my colleagues did a lot of optimisation works (Clusterio 60k was an exercise in perfect optimisation in a nutshell to a degree that even experienced megabase builders have raised a few eyebrows), so we kind'a just can see from experience knowing how the game behaves that this results in a substantial performance gain just by virtue of scaling down amount of oil processing entities 2 times and scaling down amount of pipes by what, about 5-7 times? That's a massive amount of entities removed from processing.

[FuryoftheStars]

As for performance - well, we can't really test it yet.

There’s actually a mod on the portal right now that changes BOP to the suggested ratios in the FFF. And another for some of the heavy only with early cracking.

https://mods.factorio.com/tag/oil?version=0.17

[kreisa]

Edit - irrelevant

[RocketManChronicles]

This thread sure has grown a lot in a day.

I would propose to separate these two by both simplifying blue science -

I had to stop reading there. Science is the main means of progression and undermining that aspect only makes everything else easier as a result.

Part of the appeal for blue science being a bit of a hurdle is that you're going to be using less than ideal things for a while until you are ready to make the next jump into superior tech. This minor hurdle being removed makes many things obsolete very quickly and rushes you into the late game far sooner than was designed.

This entire thread has been explaining why simplification in such a niche game that was formerly more involved is already a bad thing, but that's one thing that would have even worse ramifications than the change to oil.

But most of you don't remember how hard it was to learn.

I do. It's the only reason I bought the game in the first place.

Learning how to master the train system after hours of trial and error was a thrill once things started to flow and I started to see how things seemed to function. It provided the tingly sensation seeing trains sail across the tracks without crashing. It felt very rewarding after spending the time to learn the ins and outs which also applied to oil and other systems. I could have used a wiki but that would have spoiled the fun for me.

If these systems were simplified at the time then that rewarding sensation would either be diminished or eliminated since it took no effort to do. The main appeal of puzzles is the challenge of figuring it out for yourself, but that's not possible if you're forced to take the easy route only. Easy games like that bore the hell out of me, which is why Factorio stood out among the crowd as something to challenge the mind. Now I turn to mods to sustain the stimulation that vanilla no longer provides.

I am among others here, the game is becoming too easy. The change to oil proposed just, ugh, makes oil refining too easy. The main driving factor that drew me into Factorio was the difficulty and the complexity to solve problems. I still remember when I first fired up the game and started figuring things out, the complexity kept me trying to solve the puzzles little by little, through trial and error. Trial and error are a beauty in this game, at least you can pick it all up and start over. Or even move to a new location and build a better layout that makes more sense. The proposed change to BOP just eliminates that puzzle, almost entirely. One input to one output is just.... stupid. Why have it go through a refinery? Just make it a chemical plant then, that is what Lubricant does. I understand the devs are going for this whole "New Player Experience," but in doing so, I think they are forgetting what formula actually makes that an experience in Factorio. Complexity is what drew me into this game, simplicity is pushing me out (well, to super heavily modded).

Just like @Light says here, the sensation of finally figuring it out and having everything work after spending time on it; that is the true "New Player Experience." Dumbing down the recipes and other aspects of the game are simply just taking away from that feeling of accomplishment. Converting Crude Oil into only Petroleum Gas is nothing, there is no feeling here. "Oh wow, I just converted oil to gas, whoopty freaking do." Why not just have Crude Oil make plastic and solid fuel for Chemical Science then? The conversion from Crude Oil to only Petroleum Gas is just another step in a process, just one more conversion.

You want to improve the "New Player Experience?" Quit making the game too easy! The puzzles are the true aspect of the game, solving one problem then moving on to the next one. I remember playing vanilla for so long, solving one issue after another. Climbing that hill to automate "Science Pack 3" (yes, back when there were only 4 sciences, and the fourth was Alien Science) was a true feeling of accomplishment. Solving all those puzzles to unlock the most powerful tools in the game was a true sensation of watching it all work together. Nowadays, vanilla is too easy; way too easy. I fear that some of these moves to 'solve' the "New Player Experience" may actually not interest people. The puzzles will not be there, the feelings of solving complex problems will not be there. I played a 0.17 game in March, launched a rocket in 6 hours. The BOP was sufficient as is, as it made Chemical Science work. Petroleum Gas to make Plastic for the Advanced Circuits; Light Oil for the Solid Fuel; and Heavy Oil for the Lubricant that was making Construction Bots for me. It was a nice puzzle to solve as each output had a use and somewhere to go. Chemical Science is not the wall the devs think it is. The next true wall was both Production and High-Tech Sciences, as they involve volumes of material to fully automate.

A good "New Player Experience" needs to contain a rich balance between an understanding of mechanics and complexity. Handing the user the solution to the next puzzle doesn't teach the player to develop problem-solving skills for the game, they will just wait until the next solution presents itself, rather than trying on their own.

[mmmPI]

the beauty of new recipe overrides my "I need to make it efficient" to "I'm gonna save 10 freaking hours".

Also, something I just realised - the productivity and especially speed module and beacon savings are ENORMOUS.

I guess that's exactly the opposite of what the devs wanted to do, change the late-game.

It's not just about ratios to prevent the use of the receipe in late game even if you make BOP like 10 times less efficient, you also have evidences that lots of player don't really do all the math themselves, copy existing design, would still prefer spending 2 hours doubling oil patchs rather than 10 to design the whole AOP/cracking thing at scale.

This thread sure has grown a lot in a day.

I would propose to separate these two by both simplifying blue science -

I had to stop reading there. Science is the main means of progression and undermining that aspect only makes everything else easier as a result.

Part of the appeal for blue science being a bit of a hurdle is that you're going to be using less than ideal things for a while until you are ready to make the next jump into superior tech. This minor hurdle being removed makes many things obsolete very quickly and rushes you into the late game far sooner than was designed.

This entire thread has been explaining why simplification in such a niche game that was formerly more involved is already a bad thing, but that's one thing that would have even worse ramifications than the change to oil.

Maybe the ideas of rethinking blue science, or having construction robots earlier than logistics robots, would help gaining granularity and smoothen the learning curve , a good tutorial can be added on top of that, in exchange for less impact on the oil production chain.

As it was pointed out, there must be reasonnable means to appeal to people that would not play a game that is not stable, Factorio getting to version 1.0 also means that some new groups of people will consider it and didn't before that's seems obvious to me that there is a risk that the general satisfaction % is impacted if you don't welcome them, and if they were shy to play an early-access game, maybe they also want some more guidelines in the game, that is a stereotype ofc but i hope you get the idea.

[meganothing]

Some people in this thread posted some GUI improvement ideas (for example this list at the end of viewtopic.php?f=38&t=73443&start=340#p443944), a friend of mine had another one:

* The hover window has a status line that is very inconspicuous with all the other info lines there (about 20 for the refinery). If it shows a problem (like backlogging is) it should show its text in red color. Small change but it might have a BIG effect for beginners searching for the cause of the problem.

[Kenira]

Personally i really like the idea of having Basic Oil Processing only output heavy oil and make cracking available from the start. That and replacing red circuits in chemical science with something else so you don't need to produce plastic yet, or refine it into red circuits, which someone said was the biggest hurdle in the science pack anyway which i completely agree with.

Alternatively, my second favourite idea is only having the Advanced Oil Processing from the start, and the oil processing technology will unlock cracking. (Plus not requiring red circuits for chemical science to again make getting into oil easier).

[Dupl3xxx]

As I've been saying for over a year, 'first oil' has a very simple solution that is 'deadlock free' and it's available here: https://mods.factorio.com/mod/Zr_simpleoil

The reason for this suggestion is that 'blockage from full output' is not a concept that's tough well, if at all, to new players. Furnaces will keep burning fuel to plates into a full stack of 100. This is a players first encounter with anything that makes output, and it shows that it will fill a lot before blocking. 100 seems like a 'natural' stopping point, since that's the stack size. This teaches a player that a machine will output until it has one stack in its output slot. I've seen many new players confused as to why the machine they placed stopped building more than 4 inserters and then stopped. They run around checking every step, because blocking by full output isn't clearly communicated. Why is an assembler blocking at 4 inserters that have a stack-size of 50, while plates block at 100, their stack size?

Using the chem-plant as the first entry point into oil means you *have* to rebuild or build new to take advantage of the more advanced oil processing. it clearly shows it as optional, but better. And since you force the player to use cracking, you've already thought them what to do if they get to much of any one fluid.

Yes, that means you get another W+A -> W+B -> W+C -> D production line. Not the most exciting, but this is about showing the player new and exiting stuff without 'setting them up to fail' like current BOP is.

Additionally, you might want to change the blue science pack to require heavy oil, instead of gas, to reduce the needed complexity of the first cracking.

TL;DR: Game doesn't teach that 'full output' blocks further production, and in some cases shows that output will keep filling (furnaces!). Solutions is either to teach full output blocks, or to make first oil non-blocking. Compared to 'New BOP', cracking would give a better sense of progress, avoid blocks and teach more new concepts without overwhelming the player.

[Preserteo]

The problem of the designers is that they must not have realized that many, but many many players do it precisely because of their difficulty and the time that must be invested to solve the problems that the game proposes.

Simplifying the game to make it easier for other people to play it seems like a joke, if what we want is even more difficult, the introduction of more liquids (gasoline, diesel), or more minerals, or more recipes, or more ...

[mmmPI]

Simplifying the game to make it easier for other people to play it seems like a joke, if what we want is even more difficult, the introduction of more liquids (gasoline, diesel), or more minerals, or more recipes, or more ...

If many new players are attracted to the game because of version 1. The idea is not simplifying the game for them. But help them cope with the actual difficulty. To "teach" them better, the same lesson.

Maybe by making more steps but smaller steps, the idea is the same height at the end and even higher in the future with ever-expanding modlist.


When i started there was way less different things , there where added with time, for someone starting a game that has been developped during 7 years, you may want to help them "catch up". Again not simplify the game as is the late game should be different so that even brain-dead can reach it and don't give bad ratings.

As a parallel starting a full angel+bob pynadon map today doesn't feel as complex for me as for other players because i have already discover the content little by little when it was added, it was many small steps here and there, that now makes me feel confident about solving huge puzzle. But i don't think starting with the 3 of them in one go as your first modded experience is good. The level of complexity shouldn't be lowered ( to my opinion too). But there is still room for making the curve less steep at the beginning.

[mcdjfp]

Except that the whole point of cracking is that the ratios between what you get and what you need are off, so you're rewarded for setting up the cracking instead of just dumping the little excess in tanks, and "voiding" it by removing those tanks once in a while...

Personally, my favorite solution is to add a oil tutorial or two and realize that the game is not for everyone. (there are plenty of games out there that don't interest me and I don't want their developers to go chasing after me). I have seen too many restaurants, TV stations, computer games, and others ruined when the owners started chasing a new target audience (whatever it might be) and ended up driving away more of their current supporters than they gained in new supporters. This just seems like the best of the ideas I have seen so far if the developers insist on changing things. Heavy oil only at a chemical plant is 2nd on my list, but both are far below add tutorials.

As for the cracking, there is no way to create a perfect ratio that will work for everyone through their end game base. I would hope by having an improved ratio (designed for the midgame) it would allow the issue to hit more gradually. Cracking would still be almost essential, just not as rushed. (Of course even now enough chests can allow someone to brute force their way through the problem) Whatever the solution delaying the lesson too much does not seem to be a good idea.

Another thing that might help is a button in the building GUI that explains why the building isn't running. Some messages could be simple such as, "not enough iron plates". Others could add a tip to the message, "Light Oil output full. Consider creating solid fuel for boilers and smelters or researching Advanced Oil Processing to open new options." Once Advanced Oil processing was researched it could change to suggest cracking to Petroleum gas as an option instead of the research.

[Pi-C]

I'm doing a rail world, and manually cutting down trees to build the track is the worst

There's also :
https://mods.factorio.com/mod/Bluebuild
https://mods.factorio.com/mod/Nanobots

And of course the time-tested vanilla method to get rid of trees :
https://www.youtube.com/watch?v=JvxdPIeUBrs&t=42s

Regarding trees, in my current game I always carry a couple of wood harvesters with me. Just plop them down in the middle of a forest, add chest and inserter, connect to power -- and the area will be cleared while you can do something more interesting at another place!

(I get a lot of wood out of it, but I can use it quite well as I also play with Bio Industries + Wood Gasification and generated my map with a bare minimum of coal resources. Getting petroleum from wood gasification quite early may feel like a cheat. However, I can only use it to make solid fuel and Bio Industries' fertilizer at the moment as advancing to Blue science requires coal to make plastic bars, and the technology to make coal from wood is locked behind Blue science. So I'm forced to go out and find a bit of coal in order to prime my own coal production -- and I can't even rely on fancy weapons for that because military science also depends on coal!)

[Engimage]

I see one small but effective UI addition that would help.

In machine UI window there is a progress bar where you can see that machine is working.

Change the background of this progress bar to red (and possibly make it gradually blink) and add a text message on top of it saying “output storage full” or “not enough resources”

This will clearly show to the player that machine is not operating and will show exactly why.

[BlueTemplar]

[...]
The blue science is really behind a really big barrier, with not too many rewards given for the intermediate steps you take when setting up the automated production. The solutions by Yandersen sound quite good.
[...]

Well, (bot-less) Modular Armor, (Efficiency1) Modules and Solid Fuel are all pretty great... but mostly in games with a military/biter challenge.

the beauty of new recipe overrides my "I need to make it efficient" to "I'm gonna save 10 freaking hours".

Also, something I just realised - the productivity and especially speed module and beacon savings are ENORMOUS.

I guess that's exactly the opposite of what the devs wanted to do, change the late-game.

It's not just about ratios to prevent the use of the receipe in late game even if you make BOP like 10 times less efficient, you also have evidences that lots of player don't really do all the math themselves, copy existing design, would still prefer spending 2 hours doubling oil patchs rather than 10 to design the whole AOP/cracking thing at scale.

His example though is not "late game", it's "so far past the end game that you need to ride a rocket fleet to see it behind the horison" ! (And very few players do...)

Also, it's the hypothetical people that you are talking about would never bother with Nuclear Reactors either. But Nuclear Reactors are still here, for those that like the complexity. And so Advanced Oil Processing will still be here. Can't everyone just get along ?

EDIT : Following the "mandatory complexity" logic, one would want to require used up uranium fuel cells for nuclear fuel, and nuclear fuel for rocket !
(A manually-fed kovarex-less non-properly-ratioed nuclear reactor is arguably less complicated than an advanced oil processing setup with cracking and solid fuel output.)
Would making building and powering a nuclear reactor mandatory to win make the game actually better, though ?

[Preserteo]

Simplifying the game to make it easier for other people to play it seems like a joke, if what we want is even more difficult, the introduction of more liquids (gasoline, diesel), or more minerals, or more recipes, or more ...

If many new players are attracted to the game because of version 1. The idea is not simplifying the game for them. But help them cope with the actual difficulty. To "teach" them better, the same lesson.

Maybe by making more steps but smaller steps, the idea is the same height at the end and even higher in the future with ever-expanding modlist.


When i started there was way less different things , there where added with time, for someone starting a game that has been developped during 7 years, you may want to help them "catch up". Again not simplify the game as is the late game should be different so that even brain-dead can reach it and don't give bad ratings.

As a parallel starting a full angel+bob pynadon map today doesn't feel as complex for me as for other players because i have already discover the content little by little when it was added, it was many small steps here and there, that now makes me feel confident about solving huge puzzle. But i don't think starting with the 3 of them in one go as your first modded experience is good. The level of complexity shouldn't be lowered ( to my opinion too). But there is still room for making the curve less steep at the beginning.

I do not agree. What has led me to continue playing Factorio is because of the problems it creates, I don't want to play a Tycoon, there are many, too many, and everyone has something in common, a linear learning. Factorio no.

That's why I play it, that's why I don't want to play with simplified goals, it takes away my freedom of thought, and it takes away my desire to play.

And as happens to many people.

[Reika]

Would making building and powering a nuclear reactor mandatory to win make the game actually better, though ?

I strongly suspect a lot of us would vocally say "Yes", and that exact idea is why my FTweaks mod makes the satellite require one unit of nuclear fuel (to both require nuclear progress and simulate an RTG).

[dzaima]

Having read about half of the posts here, I still don't quite understand why do so many people think the oil change simplifies/dumbs down the game in any way. Sure, it makes the start of oil simpler, but you've still got to figure out advanced oil processing & cracking at some point sooner or later (assuming there's a good incentive to do so, e.g. lack of petroleum, some recipes, or just being a good factorio player and wanting efficiency). It doesn't remove any difficulty, it just smooths it out into two parts of learning about how oil works. As long as the research for advanced oil processing clearly mentions that it's a lot more efficient, I'd expect that most of the players will go for it - I'm guessing most would be way more happy, and get more satisfaction out of upgrading their current production line than finding another oil patch and join it to their base. Those who really don't want to go trough the hassle (or care about UPS) also have the ability to.. not. What's so bad in that?

(this is not to say I completely agree with the change (well, I mostly do), and I may be biased because I've never used coal liquefaction and don't really care about getting solid fuel until it's strictly required, and always struggled with my base dying when light & heavy oil backed up since it has exactly 0 use to me before blue science)

[BlueTemplar]

Would making building and powering a nuclear reactor mandatory to win make the game actually better, though ?

I strongly suspect a lot of us would vocally say "Yes", and that exact idea is why my FTweaks mod makes the satellite require one unit of nuclear fuel (to both require nuclear progress and simulate an RTG).

Hmm, maybe you're right ? (And push uranium even farther than oil on the map? In any case, this could be a nice game mode, with a different kind of difficulty change than DeathWorld / Marathon / RailWorld !)

As a nitpick that you might be aware of (and that I almost forgot about - thought about posting about a satellite-RTG too),
winning the game doesn't require a satellite in 0.17 anymore.

[mmmPI]

Would making building and powering a nuclear reactor mandatory to win make the game actually better, though ?

I strongly suspect a lot of us would vocally say "Yes", and that exact idea is why my FTweaks mod makes the satellite require one unit of nuclear fuel (to both require nuclear progress and simulate an RTG).

This doesn't make it harder for newbs, or overwhelming as something like "i don't have time in my life to learn all that" since you wouldn't be concerned while learning other stuff, it means you could still enjoy factorio almost the same way. It would make "beating the game" a greater achievement. If you take the staircase comparaison, it means the general level is higher, because you add and extra step almost at the end. It doesn't prevent people from trying to climb so it's fine. It would encourage people to set hard goal, at the cost of sometimes not achieving them and having to deal with that failure, which is different lesson as teaching people how to suceed by being consistant ant doing things little by little towards the end not rushing


EDIT:

about a different mode // marathon railword

Sharing this view that this is the other side of the granularity spectrum, making the hard part harder for those who wants it. I like it

[Omnifarious]

(this is not to say I completely agree with the change (well, I mostly do), and I may be biased because I've never used coal liquefaction and don't really care about getting solid fuel until it's strictly required, and always struggled with my base dying when light & heavy oil backed up since it has exactly 0 use to me before blue science)

As a random aside, the only time I've used coal liquifaction is when I had a coal patch near a uranium patch and I decided I didn't want to ship sulfuric acid to the uranium patch. So I set up a tiny oil chain that just produce sulfuric acid and nothing else (except, of course, for siphoning off just enough heavy oil to keep the process moving).