too fast "early stage", balancing for midgame

[tecxx]

hy,
tldr; early stage is passing too fast and needs more love. trains and basic train networks should be useable much earlier, especially and clearly before robotics


the long post:
i want to give personal opinion / feedback regarding the early and mid stage of a typical factorio game.

early stage:
99% of the games i played start like this:
- plug one metal extractor and a smelter directly in front of it. put wood/coal manually inside. smelt iron plates to build a second extractor.
- craft second extractor, place it on coal.
- craft third extractor, place it on copper, smelter in front, run in a circle and refuel all with coal.
- after about 10 minutes there is enough metal, copper and coal to build the pump, electric extractor, generator, 3 boilers and some belts.

and at this point, "stone age" is over and not a single gun clip, turret, armor piece, or red science bottle was produced.
so what use do the burner powered items have except passing the first couple of minutes?
this first stage of the game just seems way to short. especially when compared to what happens next - a typical game runs for over 40hours until the current end goal is reached. if the game time would be 2-3hours on average, the 10-15 startup minutes seem ok. but like this, its just a nuisance.

i want to relate to another discussion point going on in the forum "remove the pickaxe tool". when i first read the suggestion, i wondered what people are complaining about, this is part of the progression - you start with simple tools, and advance to automated machinery. now, after playing countless weekends till late in the night, i understand the request. the pickaxe item is utterly useless. and so seem the burner based items.

anyway my suggestion is not to get rid of burner based items - i'd like to see them play a more important role in the first to mid stage of the game. advancement to electricity should take a while, be a bit more costly, and not be immediately necessary to get the game rolling. implementing a change in this regard would be hard, though.... the early game shouldn't become lengthy and boring of course.
some ideas:
- burner based simple assembly stations, e.g. to automate belt manufactoring
- requiring big machinery (e.g. generator) to be built in assembly stations / factories
- additional tech levels to research, e.g. "electricity", to allow pumps and generator crafting, or "steam machines"
- steam network. use heaters and pipes to power facilities - this could serve as a early game "electricity-like" power system


mid stage:
trains take too much advanced research to really help in early midgame. we usually reached fully working train networks roughly at the same time as when we built the first roboport. what this means is that after the automation of level 2 research, engines and steel, the next logical step would be to expand out to far away resource patches and "bring stuff in", so research for level 3 can continue. and for this, rail systems should primarily be used.
but instead, once the first train rolls out to do its work, we are just a tiny step away to place the first roboport, so unloading at the main station is already done with robot sorting mechanisms, and the "endgame" phase has been reached.

I dislike this because the "progress" seems wrong. rails mainly rely on stone, steel, engines and coal. with this, you can craft rails and engines. why require level 3 research to build and run trains? level 3 should be mainly for robotics, or move robotics to level 4 research (which, as i already posted somewhere else, is way too easy to craft). but this are just some not well tought-out comments, not sure what a possible solution could be.
this issue is also related to resource placement on the map - there's just too much to really fuel the incentive to expand. for example, oil field placement - we tried playing with a reduction of available oil, and this clearly put the game on a more enjoyable level. seeing this in context - oil could be required in basic form for trains, but in heavily processed forms for higher tier chips, a related change to this could be a possible improvement. but i am confident the resource placement issues will be tackled in the future.


thoughts?

[FishSandwich]

The great thing is, if you're not happy about a part of the game it's usually moddable, and has been!

Check out the Marathon Mod!

[tecxx]

edit: sorry i misinterpreted your post.
looks like a great mod, will check it out. seems to be exactly what i was writing about.
hope some of this make it into the main game!
cheers

[Geonosis]

I tlike these suggestions a lot.

Regarding the early train issue: for me, the biggest hinderence to building a train is the engine unit, which I always build way to late and therefore build trains only for fun and not really because I need them necessarily. Making trains earlier and easier available would -for me- improve the experience early on.

Also, I think the ideas for the early game play are very valuable: Making more possible without relying right away on electricity. I would love to continue playing in early stage, using coal inserters and buring drills with complicated coal delivery networks, but since you need electricity straight away for research (to get belts etc.), it really makes no sense. Therefore, the idea of a simple alternative to electricity, such as steam power delivered by pipes for example, really appeals to me.

If the research "electricity" is early on available, but without it, one could proceed in the game beyond just producing iron and copper plates, players would have the choice if they go for the next stage right away of first work a bit with the coal / steam powered system. Imagine the new challenges that would arise from early technologies like this. I would love to find some complicated solutions to power (long-armed) burner insertes with coal while at the same time they insert iron wheels into a assembling machine that is powered by steam, which is provided by a large pipe network.

[indjev99]

I agree I think it should take longer. I mean imagine some big server with a lot of players there could be a coal age and not every new player joins, plays for 10 minutes, and he has electricity.

[SHiRKiT]

If it is FUN, then it goes. Stone age currently is just BORING (except for new players =] )

[Malawi]

A research-hut at the beginning of the game might be a cool idea. One would have to bring specific materials and wait. You would get the science lab/science packs only after researching it in the hut.

[tobsimon]

and at this point, "stone age" is over

I do agree, that the early game feels rushed, it's items too soon obsolete and all in all uneventfull, when you have some practice.

Under the impression of your post, I first wanted to make the suggestion to add water mills as early game energy source, but…

The character is an astronaut. There is no stone age. There is no development of technology, everything is basically known to the character. This is not civilization or age of empires. What happens is, that this character bootstraps a factory without preexisting infrastructure. So he can firstly build simple stuff and only later in the game more complicated stuff. Electricity is the first thing anyone would make in this situation, or not?

In many ways this concept is flawed. First, the science tree, which totally thwarts what I just said. Then the electric motors, which many items oviously contain, but are not acknowlished in the recipes and only produced much much later. Many more examples could be made. And one question remains. Is it actually possible to bootstrap a factory without preexisting infrastructure on a timescale less than a few thousand years of technological advancement? Nobody ever did that.

The real question is, how to model the bootstrapping process in a sensible way?

• It's done with a science tree now, which is absurd and shouldn't be there in the first place.
  (I think somwhere even the devs called it a bad compromise.)
• I think some mod introduced tech-download stations instead, which are radio-connected to your mothership.
• Disallowing hand-crafting for many more items can help (also a mod), but also makes the game (too) complicated.
• Item production could be made crazily interlinked and complex, but that would probably make the game unfun.

I have no answer (yet). The whole thing definately needs a rework, but that may be just my perfectionism speaking.

BTW, the early energy deliverance system should be a steam engine powered line shaft. Later you plug only a generator to the shaft but firstly everything else including a scoop wheel to water the boilers everywhere with wooden flumes. Inserters and the like don't have burners themselves, but sit on a line shaft. That really seems low tech and believable also .

But actually, steam engines need lubricants, you know…

[Drury]

I'm glad laser turrets don't require 50 different types of lenses and mirrors, 1500 different parts that would hold them in place, 600 various gears and 14 servos for swiveling mechanism (and yeah lubricant) as well as various transformers, capacitors, resistors, wires of various width, length and material and general electronics, reinforced steel plating of various shapes and widths, held together with thousands of bolts and screws. I like building at least a single turret in my lifetime, even before having access to lubricant and other lategame tech.

I very much approve of the mechanical energy harnessing in earlygame though. That's my idea as well.

[tobsimon]

, yeah exactly.
We understand each other. Mechanical energy transfer is the logical step before electric power.
But still no better solution on how to pace progression, than the tech tree and/or product drain.

[Tinyboss]

It's done with a science tree now, which is absurd and shouldn't be there in the first place.
(I think somwhere even the devs called it a bad compromise.)

I don't think it's too absurd. Supposing the character understands that all these technologies exist, it still doesn't mean he knows every detail. And even if he does (say, he has Wikipedia on his spacephone), he'll still have to spend time tweaking his designs to make sure everything is just right. That's what I imagine the science research is: perfecting the implementation, not discovering the principles.

[Drury]

Ultimately research is a progression system that has it's place in the game for gameplay purposes. It is in fact the very backbone of the game and de facto the goal throughout the game. All your efforts are concentrated towards science pack production and increasing the effectivity of science pack production.

Not only does it provide a goal, it also supplies the player with new tools at a reasonable rate. This prevents the player from getting overwhelmed. You start off with only very few options, and increase them by researching new stuff. This is very important for keeping new players interested and in control. Imagine you, as new player, started the game off with the entire tech tree unlocked. You would have no idea what to do and what all of that stuff in the crafting menu does and means and then you'd just quit thinking the game's not your thing, simple as that.

[tobsimon]

I fully agree on the need for a progression system, still I don't think research is a good way to do it.

Apart from the problem with story consitency, which may be explained some way, here are some more issues, I have with it.

It doesn't pace very good. The red phase works all right for me, but starting with the green phase, I frequently am in the situation, that science moves well and fast, and I research stuff I currently have no time or desire to use right away. In short, my factory isn't keeping up with my research progression.

The popup is annoying. I simply don't want to have to care about, what tech to research next. There is no real choice in the research tree anyway. I would bet that in most plays, it is more or less the same succession.

It takes the focus away from the main goal. Like you said, research becomes the de facto goal to work towards. In the early game (pre research), this is different. Expensive early game buildings, like the steam engine, cause you to establish a robust raw material production (not fully automated probably). It is very satisfying to be finally able to build the next big thing, you worked for. I feel no satisfaction whatsoever in the unlocking of a new technology, it's more like a "ah, finally, that took long"-relief. So, science is a goal, but I feel like it's the wrong goal to work for, when on the other hand you should get a good production for rails, motors or robot frames going, which would really bring you forward, but which I regularly neglect, because I need to get blue science done (first, because I have nearly exhausted red+green research).

It is the only abstract thing in an otherwise very hands-on gaming experience.


Seasoned gamers are very much used to science trees. So much, that they feel natural. It doesn't mean, that they are the right thing to do.


I'm not sure if feasible, but I'd love to see a natural progression system, where technology is unlocked when you are able to use it. Maybe it is enough to simply count the amount of raw material accumulated or to fullfill other goals like "build and run this production line" (1). These prerequisite should be visible, when hovering the faded-out symbols in the build menu (e). So the science tree is fused into the build menu. Buffs like the stack bonus need to be done differently (e.g. module), but then again, I think it is highly questionable whether they should exist at all. Technology where a prequisite is not yet available may be faded-out even more or be compleatly invisible.

1) Fast belts and inserters and splitters may use a prequisite in the form "exceed a transport volume of 500 items per minute on a belt segment or inserter", to detect the moment, where you will need them. The Splitter and underground belts may need a certain overall belt transport volume to be unlocked. New defensive weapons need a threat level to be reached, like number of biters in the sourroundings or damage done by attacks.


If we do away with the huge product sink, that science packs are, we probably need a substitute for that. Either more expensive buildings, or a bigger continous need, like ammunition, energy, maintenence, food and housing. While food and housing is probably a bad idea, maintenence might not be so bad, if done right.

[bobucles]

I'll admit the green science research gets burned through pretty quickly. I'm fairly novice at the building aspect, so I don't have a good factory rolling by the time blue science makes its demands known.

[Blu3wolf]

I like the tech tree honestly. How many astronauts do you know who are familiar with every step of industrial production, and know how to build absolutely everything featured in factorio?

Im guessing its a short list.

Someone mentioned earlier that there is a mod which redoes science labs to be download stations - gameplay wise they work the same, but thematically you are downloading schematics and storing it on increasingly complex circuits (the research packs get repurposed as the storage medium). Im not sure which mod did that, but its one of the ones that NARMod plagiarised from.

I quite liked how NARMod played out, but it still had the same problem: that green research goes very slowly, then very quickly. It was pretty cool how it disabled handcrafting though, and decreased the amount of inventory space you get. The component parts that it breaks things down into makes the game make a lot more sense in terms of technology order - for the most part. Starting out with the ability to make very small electric motors and slowly being able to make bigger ones is very nonsensical.

I dunno how to fix the science problem, though. If you increase the research requirements to fit on an exponential scale, then you would correct the timing, but effectively remove the choice of what to research. Some mechanic needs to slow down the progression of research I guess.

One idea I guess could be to require science labs to be spread out - say unless they are at least 100 blocks away from one another, they cannot function. This would slow down the expansion considerably - although it would also bog down the factory design quite a bit too. Doesnt really fit thematically either. Any better ideas?

[bobucles]

Some mechanic needs to slow down the progression of research I guess.

The closest thing to this is with infrastructure. Each new tier has a minimum requirement of the processing steps the player needs to begin research. I think one of the issues has to do with the pacing between the tiers. The level of infrastructure is simple to get the first red, then okayish to get green, but the learning curve spikes up on blue and drops down on purple. In detail:

Prelude: Set up some miners, conveyors, and a power system.

Red : Easy. two smelters and 2 assemblers. Done.

Green: Moderate, 7 production steps, 5 of which are new and some of which you probably made anyway. Smelters, gears, wire, basic circuit, conveyor, assembly arm, potion. Green can piggyback off of red stuff for the most part so it feels fairly simple.

Blue: Hard, at least 14 production steps including all the oil, 10 of which are new. Smelters, Steel, wire, basic circuit, assembler, fast assembler, smart assembler, oil, refining, cracking, plastics, advanced circuit, sulfur, batteries.

Purple: Easy but grindy. Artifacts. Only artifacts.

Green science is fairly simple to get and introduces a lot. It maybe introduces too much. Some of the things which take over 200 red and green science should probably be boosted to blue tier, and instead take 20-50 of all 3 (or some blue and 2x-3x red/green). Blue science probably needs too much to get started, and purple science takes practically nothing after that. This gives a bit of a downhill slide for endgame.

I think blue science should be easier to deal with. The full advanced circuit setup alone makes green science look easy. There is a big hurdle where you need the ENTIRE oil network at once, which is a problem because it throws the entire pipe learning curve at once. The high petrol demand of plastic AND sulfur first thing means you NEED to rush directly to advanced oil tech. Advanced oil is blue research! I'd say drop the battery from the potion requirements, and instead upgrade the steel bar to an engine unit. I'd say a recipe of basic engine, advanced circuit and smart inserter would be good. It's one factory only thing (plus steel), one oil thing, and one production chain thing. It also makes blue potions completely sulfur free.

If all the sulfur tech got pushed firmly into blue tier that would probably be best. Sulfur production demands so much gas, it very clearly needs advanced oil to work. Defenses won't suffer because gun turrets are beefy now, and the player can get most of their weapons and modules without sulfur. Laser turrets would come later, which isn't crippling as long as other weapons are good. Battery modules would come way later, rockets would come later, and both items could use serious buffs anyway. They can't be appropriately powerful AND stay in the green tier at the same time. Late sulfur also pushes accumulators more into late game. This somewhat encourages a solid fuel stage of energy production, and provides noob safety by not letting all the oil be used up on accumulator production. Sulfur empties out oil wells pretty fast.

Unfortunately blue sulfur production places logistic bots deep into blue tier. This is due to the flying frame needing batteries. I don't see any way of keeping blue sulfur and green logistic bots unless the flying frame recipe drops batteries, or a third bot type is made to use basic engines and solid fuel. A solid fuel bot would be kind of redundant but it would fill the gap.

Purple science is a joke. While you reasonably need high level tech to deal with endgame biters, players can still harvest several hundred artifacts with middling tech. That's enough to get the full powered stuff for endgame. The purple potion needs its material cost increased a LOT. This is a good place for a recipe that needs lubricant (processing unit or electric engine) and something really high tech. A logistic bot would take a little bit of everything, actually. For a themed recipe add a slowdown capsule to the recipe, so you need an alien artifact, a slowdown capsule, and a logistic bot. Ooh, I wonder what spooky things could be going on that requires all three things?

[Drury]

Requiring oil industry to research advanced oil processing throws me off too, but I remember liking the challenge of blue science. It was this monumental task that I never thought I could do and then just went and did it. Best feel ever.

If anything, I'd bring advanced oil down to green. It's a pretty crucial upgrade to just go without on your blue science voyage, even if for a short time.

Purple science is weird because not everybody agrees it should be a thing. Some people just hate biters and want them to be an optional part of the game. I think the ease of it comes from exactly that, so you wouldn't have to smash a ton of nests to get the benefits, right there, and right then.

It's not an excellent compromise and I think the devs realize. Guess we just have to wait and see what they come up with.

[bobucles]

If anything, I'd bring advanced oil down to green. It's a pretty crucial upgrade to just go without on your blue science voyage, even if for a short time.

I'll have to disagree. Just about everything important is already available at green science. Vehicles, trains, oil, processing circuits for SOME reason when they really should be blue... you pretty much have the full game by the end of the green research. Blue has the high tier edition of items you already had at green.

That's the reason I think sulfur should be moved entirely to blue. Sulfur is used to make a good number of items, most of them high tech. Sulfur demands so much oil production that it doesn't work without advanced oil tech. Why force the player to need advanced oil at green, or to need blue research just to have viable blue research? That's kind of a hiccup, green doesn't need even more stuff. IMO it's better to simplify the oil requirements for blue potions, put the big oil consumers at blue tech, and demand lots of oil for purple research.

Using ingredients that can't be hand crafted to produce the blue potion(basic engine is the simplest) is already good enough to make sure players NEED a production system to get things going.

Purple science is weird because not everybody agrees it should be a thing.

Purple being crafted from ONLY artifacts is pretty meh. Players should need more factory to develop purple science, just like anything else. IMO the solution is to add high tier components to the recipe. Purple potions are a good place to demand sulfur items, which means getting the advanced oil blue tech. I chose logistic bots as an ingredient because they're a deeply involved recipe and players want to build them anyway. Slowdown capsules are a good place to practice using logistic bots because they're annoying to craft by conveyor. Keep in mind that purple science comes in batches of 10 so demanding a complex, expensive item is more about building the big production chain, rather than being a nightmarish resource sink.