Core game balancing

[nafira]

Hi,

I recently finished "There's no spoon" achievement, and it was a bit easy on certain point, and not hard but frustrating on others. The end was a rush because I made several mistakes which costs me around 2 hours at least, even for a modest player like me.

So I found some balancing that can be addressed since they clearly "break" the progression curve ! If you take a look at world's record (human or bot), they all came across the same walls as I did. Making it complicated : yes. Making it insanely difficult just with unlocking a science : no.


Phase 1 :

Some good things first :

• The start of the game is pretty straight-forward and clean
• organizing red soon is quite easy with some wizard-ish techniques ^^
• going full red+green science is quite well balanced

Here come the bad points :

• Since it's pretty easy to clean all red+green science, unlocking blue is trivial but making blue isn't, so get stuck at "waiting" for another useful science
• grey science don't make any sense even if not going for a run, it really cost a leg and it produces more pollution than you can defend yourself ; it's easier to stay with simple research at first, just add more turrets, which doesn't cost anything (compared to science) and don't create pollution
• Side note : Steel furnace are quite useless for their cost. No gain on any plan ... Electrical Furnace is the game changer
• Side note 2 : having rock to collect, especially huge rocks, make the a bit too easy at the beginning, or on the other hand, the game is too slow if you don't have rocks

Phase 2 :

At this point, you have two choices :

1. Rush to petroleum to get bots and progress fast
2. Secure your space, reduce pollution and only go for modules at first ; even go solar

First one is ballsy but rewarding, the other one is how I usually play the game. Also the first one is a pain in the ass since you need to build a nightmarish number of green and red circuit assemblers : better get prepared and know your ratios and space management.
Really, this part became very complicated with sulfur, even if it doesn't look so. Instead of making everything with petroleum, you have to make intermediates.

In the end, you need more copper than Iron (what an Iron-y )


At this stage, even if setting things in motion for blue science made you cry, you didn't prepared yourself for the worst thing in Factorio : Steel. Freaking Steel.
Even if you reach bots, motors are very light on the recipe so it's negligible.


Phase 3 :

Nearly no recipes needs it as a "commonly used material", so as a normal player, you don't worry about it. The thing is, you need an incredible amount of steel to continue the game. A clear wall stands before you. If you hadn't prepare, you'll get stuck like 2h at least just building the start (speaking for beginner time, faster for us). And you're not even really using it.
When purple science comes, you just wish you were dead.
You want to go yellow first ? Don't worry, you'll need steel too, even if it's small, because you'll need like 4 assemblers dedicate to each yellow science assembler. And for all the assemblers needed for greens and red circuits, you'll need more steel than you think.

And I forgot Blue chips ! What a pleasure to discover if you start that you can't even put 6 assemblers without emptying a full lane of greens !! And a full lane of greens, it's around 40 assemblers.

Phase 4 :

The Rocket !

At this point, you easily made : electrical motors, concrete, battery, modules, and one the ingredient is clearly too easy, Rocket Fuel ! It's insanely easy to produce those in huge quantity, compared to the 20 copper + 1 steal (so 10 Iron) to create a low density module. And don't mention control unit which uses processing units, like yellow science, and speed module which is very costly considering you need green and red circuits everywhere, especially green ones.

So to sum up :

• Silo is easy, even trivial
• Getting to research low density modules and control unit is far too costly compared to rocket fuel
• Rocket fuel is trivial, I did it with 8 assemblers, just storing it
• low density module is easy enough but is used everywhere and need a big throughput of copper : if you don't have copper, it's just no fun
• making control unit is really too expensive due to yellow science usage in processing unit

As I see it, the "early game" (Phase 1 + 2) is pretty enjoyable, for any kind of players at any speed.
Next phases need either preparation or just waiting, and doing "useless" things to kill time. While I was the achievement, sometimes, I had no other choice than to just watch. If I pulled some resources from the chain, I would break it and slow it down, without any rapid gain.
Clearly it doesn't feels right. If you automate too much, you can't correct the curve of progression easily : either you made a mistake or not. It's punitive, and not rewarding to fix your production


I wonder how many people just "watch" their game, even as casual player, because they are short on every resources and trying to increase production will not help it since you can't pollute as you wish ?


Sorry for the long post, which some will find useless, but I had to share what I felt while doing this. I learn other aspects (like the Steel Wall) because I never tried to reach to rocket that fast, and the game shows you many more things you didn't saw before in your 30-40h run.

[SoShootMe]

Interesting (if long) post! I agree with a lot of what you wrote, but a few points I wanted to comment on... which have ended up making this rather a long post too .

Phase 1 :
...
grey science don't make any sense even if not going for a run, it really cost a leg and it produces more pollution than you can defend yourself ; it's easier to stay with simple research at first, just add more turrets, which doesn't cost anything (compared to science) and don't create pollution

Military and Chemical science packs have comparable total resource requirements, and both are a big step up compared to the Automation and Logistic science packs. Of course, unlike Military science, it is impossible to avoid Chemical science.

I've noticed Iron ore indirectly pumped into biters being an appreciable fraction of total use. I'm not sure how long at least Physical projectile damage 3 (the first to need Military science packs, which themselves need lots of iron) would take to be a net gain; perhaps too long for "No spoon" but I expect they are worth it in a more leisurely paced game.

Side note : Steel furnace are quite useless for their cost. No gain on any plan ... Electrical Furnace is the game changer

A good chunk of the extra raw resource cost of N steel furnaces over 2N stone furnaces (to have the same output) is made up by savings in belt and inserters. For the compact 4x12-furnace ore smelting layout I use, one set of Steel furnaces is ~600 resources more than two sets of stone furnaces. That's only 20s of the full 1.8k/m input the setups would consume.

In terms of pollution, the extra raw resources and need for steel/brick smelting mean extra for construction, and both furnaces produce the same pollution for a given output. But steel furnaces halve the coal required for a given output, so also halve the pollution due to the coal miners. For the layout above, I'm saving overall after ~20 minutes of full output, or ~40k ore.

Then of course there is the space saved, which can be a benefit by reducing the area you must defend. This is also true comparing steel furnaces to electric, which are the furnaces I find useless (until later). I guess at least some of this is a consequence of style of play.

Phase 3 :
...
Steel. Freaking Steel.

Production science feels like the biggest jump in the game to me and, in resource terms, the biggest part of that jump is due to the steel you need. But as well as needing ~3x the iron of Automation, Logistic and Chemical science packs combined, Production science packs also need ~2x the copper and wire/circuit assemblers.

As I see it, the "early game" (Phase 1 + 2) is pretty enjoyable, for any kind of players at any speed.
Next phases need either preparation or just waiting, and doing "useless" things to kill time. While I was the achievement, sometimes, I had no other choice than to just watch. If I pulled some resources from the chain, I would break it and slow it down, without any rapid gain.
Clearly it doesn't feels right. If you automate too much, you can't correct the curve of progression easily : either you made a mistake or not. It's punitive, and not rewarding to fix your production

Although I mentioned the Production science resources above, for me there was a more significant underlying jump than resources, similar to what you describe: I largely muddled through up to Production science, but to progress here I had to learn how to expand and how to plan for scale. This took some time and I think it is a point in the game that may divide players, with some deciding the game is no longer interesting, and others finding it becomes interesting in a different way.

As you also seemed to suggest, I find progressing beyond Production science (Utility science and rocket) feels easier, perhaps too easy; a case of more of what's been done already, rather than new challenges. It might feel more balanced if this was the point where you needed more focus on the enemy (defence and/or offence); to some extent it will depend on the map, but I think I've always had to deal with most of that already, usually as a distraction from getting Chemical science packs up and running.

[Hannu]

As I see it, the "early game" (Phase 1 + 2) is pretty enjoyable, for any kind of players at any speed.
Next phases need either preparation or just waiting, and doing "useless" things to kill time. While I was the achievement, sometimes, I had no other choice than to just watch.

Achievement should need preparations and skills. If there were easier speed achievement for beginners and casual players, more skilled would like to have another "pro" -achievement which is harder.

I wonder how many people just "watch" their game, even as casual player, because they are short on every resources and trying to increase production will not help it since you can't pollute as you wish ?

At default settings there is no reason to delay expansions to avoid pollution. Just build few turrets and walls where needed and make occasional clearing runs. I make usually one turret creep expansion at beginning, another when I get tank and when I have power armor I begin to conquer large areas for rail base. Now spidertron is very good for that purpose.

[nafira]

Interesting (if long) post! I agree with a lot of what you wrote, but a few points I wanted to comment on... which have ended up making this rather a long post too .

Phase 1 :
...
grey science don't make any sense even if not going for a run, it really cost a leg and it produces more pollution than you can defend yourself ; it's easier to stay with simple research at first, just add more turrets, which doesn't cost anything (compared to science) and don't create pollution

Military and Chemical science packs have comparable total resource requirements, and both are a big step up compared to the Automation and Logistic science packs. Of course, unlike Military science, it is impossible to avoid Chemical science.

I've noticed Iron ore indirectly pumped into biters being an appreciable fraction of total use. I'm not sure how long at least Physical projectile damage 3 (the first to need Military science packs, which themselves need lots of iron) would take to be a net gain; perhaps too long for "No spoon" but I expect they are worth it in a more leisurely paced game.

Side note : Steel furnace are quite useless for their cost. No gain on any plan ... Electrical Furnace is the game changer

A good chunk of the extra raw resource cost of N steel furnaces over 2N stone furnaces (to have the same output) is made up by savings in belt and inserters. For the compact 4x12-furnace ore smelting layout I use, one set of Steel furnaces is ~600 resources more than two sets of stone furnaces. That's only 20s of the full 1.8k/m input the setups would consume.

In terms of pollution, the extra raw resources and need for steel/brick smelting mean extra for construction, and both furnaces produce the same pollution for a given output. But steel furnaces halve the coal required for a given output, so also halve the pollution due to the coal miners. For the layout above, I'm saving overall after ~20 minutes of full output, or ~40k ore.

Then of course there is the space saved, which can be a benefit by reducing the area you must defend. This is also true comparing steel furnaces to electric, which are the furnaces I find useless (until later). I guess at least some of this is a consequence of style of play.

Phase 3 :
...
Steel. Freaking Steel.

Production science feels like the biggest jump in the game to me and, in resource terms, the biggest part of that jump is due to the steel you need. But as well as needing ~3x the iron of Automation, Logistic and Chemical science packs combined, Production science packs also need ~2x the copper and wire/circuit assemblers.

As I see it, the "early game" (Phase 1 + 2) is pretty enjoyable, for any kind of players at any speed.
Next phases need either preparation or just waiting, and doing "useless" things to kill time. While I was the achievement, sometimes, I had no other choice than to just watch. If I pulled some resources from the chain, I would break it and slow it down, without any rapid gain.
Clearly it doesn't feels right. If you automate too much, you can't correct the curve of progression easily : either you made a mistake or not. It's punitive, and not rewarding to fix your production

Although I mentioned the Production science resources above, for me there was a more significant underlying jump than resources, similar to what you describe: I largely muddled through up to Production science, but to progress here I had to learn how to expand and how to plan for scale. This took some time and I think it is a point in the game that may divide players, with some deciding the game is no longer interesting, and others finding it becomes interesting in a different way.

As you also seemed to suggest, I find progressing beyond Production science (Utility science and rocket) feels easier, perhaps too easy; a case of more of what's been done already, rather than new challenges. It might feel more balanced if this was the point where you needed more focus on the enemy (defence and/or offence); to some extent it will depend on the map, but I think I've always had to deal with most of that already, usually as a distraction from getting Chemical science packs up and running.

Thanks for taking the time to share Always enjoyable to read !

For military :
What I meant was more : you need to setup a big set of pollutive assembler/miners, to create and use grey science. Because when you use a grey science, which is a lot of iron, copper and steel, you also need red and green science. This is not to be forgotten. Grey science don't pollute that much, but the whole package does (IMHO). And what it gives you is not rewarding at all in term of real defense.

It doesn't beat adding more turrets or just upgrading to piercing rounds. With that you defend yourself against big swarms with a very low cost.
And don't forget that you have to add blue science after only 2 upgrades, which kinda kill the thing in terms of costs.

Indeed in a pure leisure/discover game, there's no reason to avoid grey science. But if you're not casual, there's no use (again IMHO) due to pollution (total re+green+grey+labs needing energy) and adding a few turrets is more help than 300 red+green+grey science pack.


For steel furnace :
Indeed you half your use on coal... on paper !

Long explanation :

Because to produce 1 steel furnace you to

• smelt 30 iron plates into 6 steel
• which where 30 iron ore smelted before
• 10 bricks which is smelting 20 stone
• Total : 6 smelts at 16s + 30 smelts at 3.2sec(iron plates) + 10 smelts at 3.2 sec

So to create 1 steel furnace, which will save 50% of coal, which costs not much to compensate (in terms of energy), you have to smelt 46 times object which is 224sec at 90kW = around 20MJ = 5 Coal.

You also need extra energy to drill materials (which again cannot be used to create more useful items) :

• 30 iron ore = 5.4MJ
• 20 stone = 3.2MJ
• roughly 2 coal coming from iron and stone ore (8.6MJ)
• I'll consider the 5 coal negligible (technically 670kJ for a grand total of 9.2Mj)

So 224 sec of smelting + 114 sec of mining and a 29.2MJ energy bill(7.5 coal), just to build 1 steel furnace. And the pollution it go with.

Is it really worth it ? giving that you need to take out serious amounts of iron ore from a belt.
Coal ? To save 1 coal, you need to smelt 27 times a simple one (a coal last 13.35 uses, so 43sec approx.), but you need to compensate 7.5 coal at least, so 200 smelting before starting really saving, at a rate of 1 coal every 27 smelting. But to drill them, you need ...54s, let's say 1min. During 1 minute, you need to power the drill 5.5MJ which is a 1.5 coal, which reduces the gain made again.

Very shortly : you save 1 coal every 2 minutes per smelter ; 30 coal per hour which costs 60sec (over an hour) to mine and 1 coal per hour in term energy. A serious player will have 300 smelters by then, reducing to 150 to replace : 4500 iron to spend on this, to save .... 4500 coal per hour.
Before empty a starting 150k coal patch ... you'll make it last 33h instead of 16 hours.

But giving that you use 13 ore before using 1 coal, you have to smelt .... 2M times. At this point you are already on the next patch of Iron or Copper because your first one are empty.

Very long calculation for a short story : given a 150k starting coal patch, you'll need to smelt 2M plates before being in need. It's not worth it to save your coal for this purpose. And they are no use after that.
Instead you can use a 5 stone furnace, which only double the footprint of smelting area.


And a quote :
"but to progress here I had to learn how to expand and how to plan for scale".

It's precisely my point !
You either fail to plan and lose time, how to put it correctly, without any way to compensate by upgrading another side of the factory. You just failed to rise properly.
Seen from the other side of the mirror : Factorio say to you "the only way is the planned way, otherwise you will be slow as a snail".
What if I just want to play "as is", developing linearly during my playing time ? Factorio says "No, play well or wait on parts where you failed".

It really doesn't push you to explore new ways of going from the start to the launch.


Next thoughts :
Giving that "go planned until the end or wait" thing, I'm wondering about some very debatable things. First Petroleum.
It feels good to worry only of petroleum when you just want to make blue science work. It removes a step, clearly, and for a big time. Also, it makes it rewarding to upgrade to Advanced Petroleum, and more than double the pure petroleum output if entirely cracked.
But on the other side, it push bots (and motors for short) very far. It also pushed Tier 3 belt further.
So on one hand we created a bigger step to avoid huge steps, but it resulted into a smaller one but with much things to do with it : light + heavy + lub + cracking + starting only this point of electrical motors and Tier 3 belt.

It's like a tsunami : "Hey the sea is going back guys ! (10 seconds later) ---Oh shit, a wave that carries 100 times the small waves we had before, just strike us".
Not a step, just underwhelming ! And not just a bit.

As I see it, the "early game" (Phase 1 + 2) is pretty enjoyable, for any kind of players at any speed.
Next phases need either preparation or just waiting, and doing "useless" things to kill time. While I was the achievement, sometimes, I had no other choice than to just watch.

Achievement should need preparations and skills. If there were easier speed achievement for beginners and casual players, more skilled would like to have another "pro" -achievement which is harder.

I wonder how many people just "watch" their game, even as casual player, because they are short on every resources and trying to increase production will not help it since you can't pollute as you wish ?

At default settings there is no reason to delay expansions to avoid pollution. Just build few turrets and walls where needed and make occasional clearing runs. I make usually one turret creep expansion at beginning, another when I get tank and when I have power armor I begin to conquer large areas for rail base. Now spidertron is very good for that purpose.

I totally agree for the achievement, it's preparation and skills, and it's what I did. I just failed on Steel (forgot to buffer), and I was not used to go further in purple and yellow science, due to the "end-game" type of ratio and placement.


And so you take time to conquer, to avoid retaliation, because if you expand too fast, you'll get swarmed.
So in other terms, you need to spend time on a useless task (bitters comes back), to let you expand. It's part of the gameplay, hopefully for us, but it's taking, IMHO, a mandatory way : you have to clean a territory (preferably using water ) to build some other factories (not just for pure expansion). We all do that, sadly.
Slightly different as I see it. It's more a chore, even for a 5h player who already died twice due to bitters

Difficult to address that as balancing though.

[SoShootMe]

A good chunk of the extra raw resource cost of N steel furnaces over 2N stone furnaces (to have the same output) is made up by savings in belt and inserters.

Indeed you half your use on coal... on paper !

What I wrote could have been more clear. My point was that reduced coal consumption means reduced pollution from mining coal, and my argument was that even though making steel furnaces creates more pollution, this is likely to be made up for by the reduction while in use.

A confession: I used incorrect raw resource values for the steel furnace (5 steel instead of 6 due to faulty memory, 10 stone instead of 20 due to stupidity) when working out the figures I gave. For the layouts I compared (1.8k/m plates), this raises the raw resource difference to ~1.3k, and pollution break-even point to ~35 minutes of full output, or ~70k ore.

This makes both my quoted statement and first paragraph weaker than I thought, but I think they are still valid. In particular, you looked only at the cost of the furnaces (higher for steel furnaces), not the cost of belt and inserters (higher for stone furnaces).

It seems to me that in the end, there is primarily a trade-off between less space (steel furnaces) and fewer resources for construction (stone furnaces). The benefit of the former is difficult to quantify but I think you are overestimating the latter. I'm still unconvinced steel furnaces are "quite useless" as you described them, even going for "There is no spoon".

[jodokus31]

Steel furnaces are IMO nice and fast to double output of the smelter-array in-place. It's more expensive than to just double, but the same logic applies to higher tier belts. And footprint matters, esp. with biters on IMO.

Electric Furnaces are very expensive and need a redesign of the smelter-array and are mainly nice to use with modules. Not really an alternative for spoon runs.

[nafira]

A good chunk of the extra raw resource cost of N steel furnaces over 2N stone furnaces (to have the same output) is made up by savings in belt and inserters.

Indeed you half your use on coal... on paper !

What I wrote could have been more clear. My point was that reduced coal consumption means reduced pollution from mining coal, and my argument was that even though making steel furnaces creates more pollution, this is likely to be made up for by the reduction while in use.

A confession: I used incorrect raw resource values for the steel furnace (5 steel instead of 6 due to faulty memory, 10 stone instead of 20 due to stupidity) when working out the figures I gave. For the layouts I compared (1.8k/m plates), this raises the raw resource difference to ~1.3k, and pollution break-even point to ~35 minutes of full output, or ~70k ore.

This makes both my quoted statement and first paragraph weaker than I thought, but I think they are still valid. In particular, you looked only at the cost of the furnaces (higher for steel furnaces), not the cost of belt and inserters (higher for stone furnaces).

It seems to me that in the end, there is primarily a trade-off between less space (steel furnaces) and fewer resources for construction (stone furnaces). The benefit of the former is difficult to quantify but I think you are overestimating the latter. I'm still unconvinced steel furnaces are "quite useless" as you described them, even going for "There is no spoon".

Indeed, there's a gain in long term pollution, but you have to invest quite a long IMO to make a use of it. The only thing that uses steel furnace is to craft steel.
The main problem of the starting game is that you pretty much cover all the ore patch very fast, and replacing stone furnace with steel one doesn't produce more. You're already output everything you can from the ore patch, and you don't need a 100 steel furnace for this.
They "feel" pretty useless, even they aren't, as every part of the game. They should also have like a 10% gain in productivity, because it's an advanced furnace.

And this is the main problem why we don't use steel furnaces for common smelting : it's because the size of ore patch is fixed, and so the maximum output is caped. No need to have a smelting capacity oversized.

Steel furnaces are IMO nice and fast to double output of the smelter-array in-place. It's more expensive than to just double, but the same logic applies to higher tier belts. And footprint matters, esp. with biters on IMO.

Electric Furnaces are very expensive and need a redesign of the smelter-array and are mainly nice to use with modules. Not really an alternative for spoon runs.

My post is not about how it fits with the spoon. It just made me question some things about the core gameplay.
You can "double" your smelting speed, if you have 1k/min ore to smelt, and a smelting speed of 2k/min thanks to steel furnaces, it won't give you anything. You just set in motion to much horsepower for nothing.

And Electric furnace are so costly to make that you don't even consider to use them. They are for production science and that's all.
In the main game there's many many uses :

• no more coal
• less pollution, like a lot
• synergy with trains
• modules

If you think they are only uses because they "are mainly nice to use with modules", you totally didn't get the purpose of electrical furnaces ...

[DanGio]

I think the furnace tiers are well balanced, as the other tiers in the game. Generally, higher tiers aren't just *plain* better, and that's good, because otherwise Factorio would be all about grinding upgrades like in a clicker game. Look at red belts : 2* faster, 8* more expensive. Assembling machine 2 : 1.5* faster, 2* times more expensive and less energy efficient.

[jodokus31]

Steel furnaces are IMO nice and fast to double output of the smelter-array in-place. It's more expensive than to just double, but the same logic applies to higher tier belts. And footprint matters, esp. with biters on IMO.

Electric Furnaces are very expensive and need a redesign of the smelter-array and are mainly nice to use with modules. Not really an alternative for spoon runs.

My post is not about how it fits with the spoon. It just made me question some things about the core gameplay.
You can "double" your smelting speed, if you have 1k/min ore to smelt, and a smelting speed of 2k/min thanks to steel furnaces, it won't give you anything. You just set in motion to much horsepower for nothing.

And Electric furnace are so costly to make that you don't even consider to use them. They are for production science and that's all.
In the main game there's many many uses :

• no more coal
• less pollution, like a lot
• synergy with trains
• modules

If you think they are only uses because they "are mainly nice to use with modules", you totally didn't get the purpose of electrical furnaces ...

You seem to assume, that one builds all the miners and smelters immediately with stone furnaces. f.e.: 2 yellow iron belts (48 stone furnaces and 30 miners each).
But if I only build 1, unlock red belts & steel furnaces, I can upgrade it for the second yellow belt, without building another smelter, which takes time. And each next smelter, that I have to build manually is twice as fast. Or I just build 24 steel furnaces for 1 yellow belt.

The most useful aspect of electric furnaces is, that you can insert prod-modules and beacons around, for extra plates in later game. Another aspect is to insert efficiency modules to reduce pollution. Or just without modules to reduce pollution, but this ones seems a bit weak for the cost and the bigger size (9 tiles vs. 4 tiles). The missing coal requirement makes them for me also less interesting from gameplay aspect.
I tend to don't build electric furnaces before rocket, maybe, if I severely need to reduce pollution in a deathworld setting.

[Serenity]

And this is the main problem why we don't use steel furnaces for common smelting : it's because the size of ore patch is fixed, and so the maximum output is caped. No need to have a smelting capacity oversized.

Who is this "we"? You need to expand to other ore patches eventually. That's not an option. Sure, you can get by with stone furnaces for some time, but once you really get into red and blue circuits and production science the iron and copper demands increase greatly. At that point it's a great help to just double your smelter output in place by upgrading the furnaces and belts and then simply hook up a train. Why should I build another stone furnace smelter instead? Why should I make my smelting area twice as large as it already is?

And the production cost of machines is completely irrelevant unless you consume them of science. Hunting for some achievement is a special case and not a basis for any game balance.

[nafira]

And this is the main problem why we don't use steel furnaces for common smelting : it's because the size of ore patch is fixed, and so the maximum output is caped. No need to have a smelting capacity oversized.

Who is this "we"? You need to expand to other ore patches eventually. That's not an option. Sure, you can get by with stone furnaces for some time, but once you really get into red and blue circuits and production science the iron and copper demands increase greatly. At that point it's a great help to just double your smelter output in place by upgrading the furnaces and belts and then simply hook up a train. Why should I build another stone furnace smelter instead? Why should I make my smelting area twice as large as it already is?

And the production cost of machines is completely irrelevant unless you consume them of science. Hunting for some achievement is a special case and not a basis for any game balance.

Calm down and act with the name you have here.

"I" don't use very often those because I either ran out of space in my patch meaning I'm developing in an other area, and so I've done the jump to Electrical Furnaces before using them for science, which is clearly an indicator that I'm not running for a time.
And I quote you "the production cost of machines is completely irrelevant". Same applies for electrical furnaces.


I'm not trying to make a judgment over Steel Furnaces, even if it seems so.
But when I expand to train, it's because I need to go bigger, and to do that, you (or "I" should I say) need to be able to smelt it either on the spot or at your base.
And if your base is going bigger, you're going to need coal for plastic and most of all, green modules, which decrease the amount of coal needed by 75% (on mining drills) and Assembler.
And it reduces pollution also !

I use Steel Furnace, because it's still cheap enough, even in for the spoon, but for the spoon, you have other problems than space. In a normal play, as a beginner or as a veteran, I don't stay very long with them. Sure, you have them fast (it's basically the first blue science I pick). Why ? Because steel is hard to produce.
But I never used them when expanding, because you need to "spaghetti more" with your base and trains. And you still use coal.


To avoid further misunderstanding, this is how I see Steel Furnace :

• Steel furnaces need to be a greater science to search : 300 red+green+blue
• Just dividing time by 2 isn't very ... interesting. Sure, it has it's purpose, but something is missing : you're building hell of an expensive furnace, which is costly in term of science(not with 75, but with 300 yes)
• Instead of simply reducing time, it should produce 5% extra products to say "This furnace is made of steel and help us create a plate with less resources. It's an enhancement, not just a faster one"
• The steel furnace help us reduce the energy needed because it doesn't dissipate energy like with stone furnaces ; let's reduce the power use to 60kW (instead of 90)

I don't have the translation for this in English but there's a saying "avoir le cul entre deux chaises" in french which is a bit familiar, which literally translate by "having your ass between two chairs", meaning that you have a purpose, but the context of your existence makes you irrelevant regarding the other two possibilities.

Stone furnace : fast to build and cheap, do the job
Eletrical furnace : expensive, faster, cleaner and doesn't need coal

In the middle : Steel furnace, expensive(same), faster(same), cleaner(a little bit) but it does need coal. And that's why it's actually not very interesting to use them (unless you play more vanilla than the vanilla game, and always use coal to smelt things)

[jodokus31]

[*]The steel furnace help us reduce the energy needed because it doesn't dissipate energy like with stone furnaces ; let's reduce the power use to 60kW (instead of 90)

The stone furnace uses 90 kw for crafting speed 1.
The steel furnace uses 90 kw for crafting speed 2. That's half the energy for the same product.
The electric furnace uses 180 kw + 6 kw idle for crafting speed 2, thats more than double the energy than the steel furnace. With modules, it gets of course better.

(That's a relict, where boilers had efficiency 50%, but left in the game for whatever reason)

[nafira]

[*]The steel furnace help us reduce the energy needed because it doesn't dissipate energy like with stone furnaces ; let's reduce the power use to 60kW (instead of 90)

The stone furnace uses 90 kw for crafting speed 1.
The steel furnace uses 90 kw for crafting speed 2. That's half the energy for the same product.
The electric furnace uses 180 kw + 6 kw idle for crafting speed 2, thats more than double the energy than the steel furnace. With modules, it gets of course better.

(That's a relict, where boilers had efficiency 50%, but left in the game for whatever reason)

As calculated before, the main goal of Steel furnace is to reduce the footprint of the smelting area. The energy saving is nearly negligible.
Half of "not much" is still "not much" (remember it's less than 30 coal saved per hour per Steel furnace).

The big interest of Electric furnace is that's it's cleaner (if not powered by Steam) and doesn't need coal, which is a game changer is the gameplay.

Pollution :
stone furnace : 2/min
steel furnace : 4/min (so the only gain is from the mined coal not needed)
electrical furnace : 1/min and if you translate that into steam pollution due to consumption, it goes back to 2/min. But you still have the coal removed from the equation ! Pretty nice
And as you mentioned it, you can enhance it with a simple module !

[jodokus31]

electrical furnace : 1/min and if you translate that into steam pollution due to consumption, it goes back to 2/min. But you still have the coal removed from the

Wait, how do you get 2 pollution per minute with steam power?
186kw is 10.33 % of a boiler, which pollutes 30/min, so ~ 3.1/min only for boiler power.
Correct me, if i'm wrong.

Solar or Nuclear are of course much better in terms of pollution.

Btw I'm only arguing against "Steel furnaces are useless" and don't say that electric furnaces are bad

[nafira]

electrical furnace : 1/min and if you translate that into steam pollution due to consumption, it goes back to 2/min. But you still have the coal removed from the

Wait, how do you get 2 pollution per minute with steam power?
186kw is 10.33 % of a boiler, which pollutes 30/min, so ~ 3.1/min only for boiler power.
Correct me, if i'm wrong.

Solar or Nuclear are of course much better in terms of pollution.

Btw I'm only arguing against "Steel furnaces are useless" and don't say that electric furnaces are bad

You're right, I just took the fact that putting 2 module inside to reduce consumption by 60% is nearly mandatory and really rewarding considering production costs.
It's like for drills, I rarely saw someone not taking at least a part of it's resource to reduce drills consumption
But, straight out, indeed, it stays at 4 considering raw needs. It's more than 4 with full calculation I think

[Aru]

Steel furnaces are not much more expensive than stone, considering inserters and belts, and they save time and space to put down. They are simple and cheap to use in place of stone furnaces. The advantage may be slight, but so is the cost, and so I think they are worthwhile, at least for new smelting lines if not replacing stone ones, though ultimately not very impactful or important. Electric furnaces use the same power and produce the same total pollution (via steam). They are a lot more important, but they are also a vastly larger investment that takes much longer to reach. Steel furnaces can easily excuse delaying electric furnaces until efficiency modules and construction robotics are done, which makes replacing the smelting lines easier, reduces power and pollution, and removes much of the disincentive from committing to a lot of early furnace lines (which is, having to deconstruct them later, very slow without robots).

Skipping or committing to steel furnaces, are similarly practical, I think. I wouldn't mind steel furnaces being made very slightly cheaper, or if stone furnace was part of the recipe like before, which makes the upgrade even more direct.

As for the jump to blue science and using more copper than iron, those were both far more extreme before the 0.17 changes. Now you use ~ 35% more iron than copper for science.

And if the game is at any point too slow... well, it isn't, the game goes as fast as the player. If you are impatient with waiting for your base, honestly, I think it's time to turn off the game and take a break, and come back when you can play better. Skillful practice makes you faster at everything. Or just sit and patiently wait if that's your thing. If you don't think you have enough materials, smelt more materials.

As for managing pollution, I don't worry too much about it. I try to scout biter bases in advance with a car, and kill them with turrets before the pollution reaches them. Best defense is a strong offense. The way I see it, I will want to kill those bases eventually, for one reason or another. When I do, it will use the same time and ammunition, and add the same evolution, whether I do it sooner or later. So I might as well do it sooner, before they attack. Base defense is easy when you don't have to defend because you already killed everything. Plus this forces more bases to spawn with the current score rather than a later one (because you travel farther out), which means smaller worms, and you will have a buffer zone against base expansions. The faster you can progress, the further ahead of the evolution score you will be, generally.

Don't worry much about evolution, scout the edge of your pollution cloud, kill everything, build more smelting and mining, build more smelting and mining, you'll have plenty to do and plenty of plates. Some materials are about volume, others are about design complexity. Put efficiency 1 modules in your drills so they don't get attacked. A nice thing to know, a furnace smelting steel consumes the same iron that another furnace produces. So, a line of steel-making furnaces uses up to an entire line of iron-making furnaces. A set of science packs every 2 seconds (my personal standard until infinite science), not including space science, uses 8 yellow belts of iron and 6 belts of copper.

I mean no offense, but whenever anyone says factorio is slow, or that science is too hard, I think they are approaching it with too much of a single-player guided campaign type of mindset. (I'm guilty of this too, to a degree, everyone is.) You have to throw that completely away and approach it as an open ended, sandbox, creative engineering game. You have to set your own pace and objectives to enjoy it, which is a different kind of mindset, but it adds a different layer of enjoyment, too. Then, you are free to to progress with the full force of your mind, instead of waiting for a path that isn't coming.

[nafira]

Skipping or committing to steel furnaces, are similarly practical, I think. I wouldn't mind steel furnaces being made very slightly cheaper, or if stone furnace was part of the recipe like before, which makes the upgrade even more direct.

Agreed, and this is a the thing I'm speaking about : either it's irrelevant or you go all the way down. No middle way, like enhancing current. It's just actually a bit less polluting and a bit of space, but far from noticeable (indeed for new line, it is).

As for the jump to blue science and using more copper than iron, those were both far more extreme before the 0.17 changes. Now you use ~ 35% more iron than copper for science.

The jump in Copper is not in blue, the big jump is for yellow science. I just doubled it just for yellow and spare low density.
And considering you develop purple and yellow science at the same time, the jump in copper need is quite incredible.

And if the game is at any point too slow... well, it isn't, the game goes as fast as the player. If you are impatient with waiting for your base, honestly, I think it's time to turn off the game and take a break, and come back when you can play better. Skillful practice makes you faster at everything. Or just sit and patiently wait if that's your thing. If you don't think you have enough materials, smelt more materials.

...

I mean no offense, but whenever anyone says factorio is slow, or that science is too hard, I think they are approaching it with too much of a single-player guided campaign type of mindset. (I'm guilty of this too, to a degree, everyone is.) You have to throw that completely away and approach it as an open ended, sandbox, creative engineering game. You have to set your own pace and objectives to enjoy it, which is a different kind of mindset, but it adds a different layer of enjoyment, too. Then, you are free to to progress with the full force of your mind, instead of waiting for a path that isn't coming.

None taken

I don't consider Factorio slow or slower than it should be ; I'm used to and like slow games. Pretty much the opposite : I don't like competitive games, because it's too stressful and I've the impression to be hassled.
So I often need 40h to finish a game, because I want to play nicely, expand properly, and such. I'm never in a rush. And as far as my experience was going, if you wanted to go faster, you could, the game was scaling up. Until I tried the spoon.



The "problem" I try to address here, is that, finishing the spoon made me realize there's a huge gap between the "early game" (blue science researches included) and the rest; it's far more linear than the rest of the game at start and the *pwoof". Any "somehow" organized player can reach blue science in 3h top. My science rate is 2 SP/sec up to blue.
The slower you go, smaller the wall will be.
And this is "the wall" I'm speaking about, it's because in a matter of an hour for example(which might be pretty accurate somehow at 120 SPM), you finished all blue researches (not counting military ones), so you have 1 hour to make purple science available basically.... If following the logic.
Meanwhile at a rate of 30 SP/minute, you have 4 hours to build a fourth of that.

So the faster you finish blue science, smaller will be your time available before purple science "must" come up.
And the slower you go, the less you will feel that wall, until there's no wall, like in most of my or our games. I think you only see it if you go under 10 hours to setup all science (not especially finishing research).

I wrote a huge text, which I will replace with this graph :

spoon.jpg (129.53 KiB) Viewed 8198 times

This is the save from my run, 50 min after launch :
- iron is scaling linearly, which after 4h falls down because I didn't buffer steel, and was a bit slow making purple (my mistake)
- 1h before the end, copper demand, which was linear, jump out of nowhere, and double it's need : and still I had a stash of 12k green circuits, helping a lot.

THIS is not normal scale. I surely lost 2h with steel and inexperience, but the wall is clearly visible here, don't you agree ?
Purple science was up and running 2h30 before launch, slow but running. 30 min later, at full speed, it took loads of iron, and then copper came out, taking all space I had on all my patches.

Sure we are putting the game to the edge, but imagine a fast player : he's basically stopped unless he find some huge copper patches (in footprint not richness).
I think it's showing how copper is not well balanced in the course of the game while iron is.

[4xel]

Hi,

I recently finished "There's no spoon" achievement,
[...]
So I found some balancing that can be addressed since they clearly "break" the progression curve !

The game is not supposed to be balanced around speerunning though. The progressiong curve is broken because you skipped a lot of what hte game has to offer.

If you take a look at world's record (human or bot), they all came across the same walls as I did.

I don't have that impression. To me it looks more like they were prepared (as need be for a speed achievement) and rammed right through that wall.

• Since it's pretty easy to clean all red+green science, unlocking blue is trivial but making blue isn't, so get stuck at "waiting" for another useful science

As a beginner, the fix was easy for me: research slower, take time to appreciate what you unlocked and don't worry about lab idle time. Aiming at 45 or lower SPM immensely help getting through later researches as well.

Now that I speedrun, rushing to bots is essential, and having good researching speed is useful for late techs such as the Silo, and there is a gap in between, and I don't see it as an issue.

• grey science don't make any sense even if not going for a run, it really cost a leg and it produces more pollution than you can defend yourself ; it's easier to stay with simple research at first, just add more turrets, which doesn't cost anything (compared to science) and don't create pollution

How many assemblers do you have on grey science? You can go very far with only one assembler on walls, one on grenades, one on red ammo and two on military science. The impact should be barely noticeable. You can even handfeed them to make things simpler. You probably need grenades and walls anyway. It's a common rooky mistake to try to do everything in perfect ratio and to aim at a given SPM. It does not make sense in general, but it makes much less sens with military. Assess your needs and go from there. If you need 500 military science for laser turrets over the course of 4 hours for landmines and some upgrades, building 10 assemblers for perfect ratio, or 20 to keep up with other science, is completely absurd. 500 military science over 4 hours is something around 1 assembler.

If you feel like you don't need military science at all don't go for it. Again the game is not balanced solely around speerunning. Biter evolution is affected by time, pollution and nest killed. A casual player fooliing around and killing biters for fun will have more use for military science than a speedrunner who focus all his effort and resources into launching a rocket.

• Side note : Steel furnace are quite useless for their cost. No gain on any plan ... Electrical Furnace is the game changer

I am not aware of a single top speedrunner ever using electric furnaces. The capital cost is just not worth it. Pollution is a non factor even on deathworld, there are much better places to use modules.

Of course, if you don't make any military science, you don't have landmines. And if you don't have landmines, pollution is an issue. But even then, efficiency modules in miners and pumpjack and solar panels are better use of your resources than electric furnace (and they're considered very bad for speedruns).

Steel furnace, on the other hand, is a common sight. As already explained, the cost isn't that high once you factor in everything you save on belts and inserters, and you also half the construction time, which is not a small gain!

Also, the interpretation of your math is wrong.
"saving 1 coal every 2 minutes per smelter" sounds right, but we mostly don't care about whether it makes patches last longer. Accounting for logistic, the difference in cost between one steel furnace and two stone smelter is in the ballpark of 50 ores, even accounting smelting. "saving 1 coal every 2 minutes per smelter" means a steel furnace breaks even in 1hour40. That's shorter than momst speedrun, and that's only one of the perk of the steel furnace (the other being space and construction time, both halved).

• Side note 2 : having rock to collect, especially huge rocks, make the a bit too easy at the beginning, or on the other hand, the game is too slow if you don't have rocks

The early game isn't supposed to be tedious, rocks are a good thing. If you like tedious early game, try out the lazybastard achievement (or play like it if you already have it)

• Nearly no recipes needs it as a "commonly used material"

Rails, tanks, piercing rounds, engines (and vehicles), trains...
If you're good enough none of these common items are necessary for a speedruns, you should be good enough to know in advance you will still need steel later on and prepare.

One of the first thing beginners are told when they interact with veteran is don't buffer. Because yes, your 500 copper cables won't last long once you automate a factory, and yes a 15000 gear wheels buffer is stupid when you need green circuits to progress and your starter iron patch is exahausted.

One of the first thing upcomoing speedrunners learn, however, is do buffer. If you know what you need in advance, starting early is recommended for what you need the most.

The thing is, you need an incredible amount of steel to continue the game. A clear wall stands before you.

Not really. Again, not balanced around speerunning and speedrunners should be prepared.

A normal player is not supposed to zoom linearly and easily through the techs and rush a rocket. Later tech unlock more possibilities, which are supposed to be explored before moving on.

It's not red, green, blue, purple, yellow with a big walls between blue and purple.
It's more like electricity, red,
factory lines, green,
cars, trains (makes most upcoming walls much smaller), modules, explosive, flamables, blue, grey
Electric furnaces, purple
robots (also a gamechanger), modular armor, power armor, yellow
Artillery, nuclear power, beacons, rocket

If you follow those kinds of steps, your progress is much more incremental.

• Silo is easy, even trivial

I fail to see how it is an issue. Your first assembling machine cost less than what it took to unlock, and very quickely process several times its own cost of materials. Rocket Silo are much simpler to build than rockets IRL as well.

• Getting to research low density modules and control unit is far too costly compared to rocket fuel

It is again a progression thing. You're supposed to play with rocket fuel in your car and train long before you touch yellow science.

• low density module is easy enough but is used everywhere and need a big throughput of copper : if you don't have copper, it's just no fun
• making control unit is really too expensive due to yellow science usage in processing unit

So only RCU is hard? Sounds fair to me, it's the new thing and it unlocks white science.

Next phases need either preparation or just waiting, and doing "useless" things to kill time.

Yes, speed running and achievement hunting require preparation.

While I was the achievement, sometimes, I had no other choice than to just watch. If I pulled some resources from the chain, I would break it and slow it down, without any rapid gain.
Clearly it doesn't feels right. If you automate too much, you can't correct the curve of progression easily : either you made a mistake or not. It's punitive, and not rewarding to fix your production

Not sure what you're trying to say. Mistakes can make you lose some time, but bots makes it really convenient to fix them. That being said, no amount of factory fiddling can fix not having enough resources. Go outpost for that.

I wonder how many people just "watch" their game, even as casual player, because they are short on every resources and trying to increase production will not help it since you can't pollute as you wish ?

Many players just watch their factory for many reasons, but if your mindset is to speed run, then yes you can pollute as you wish and no, if you need more resource, you go fetch them, you do not wait for more. I did the math a big while ago, and typically, the worse polluters require, marginally, an aditional 5 to 10% of their energy consumption dedicated to defense upkeep in bullet or laser rays. So not a negligible cost if biters are nearby, but it's still much preferable to expand than not.

[blazespinnaker]

For there is no spoon and getting on track planning, there is a simple formula for rough estimation planning I have shared before that works very well.

It's the sum of sequences plus a few variables - EntitiesPlacedPerPeriod*PeriodInMinutes*n*(n+1)/2. Using this, you can determine your build order and priorities and time it will take.

For example, to get to rail tech for getting on track achievement, you need 178 lab minutes of research.

If you can place and fully fill labs slightly more than once per 10 minutes, than you can get to rail tech in about 50 minutes. Eg: 1.18*10*5*6 = 177.

If you can place labs and production faster, you'll get there much sooner. I believe the record is around 22 minutes last I checked.

This can work for launching a rocket silo as well, which requires 1000 lds/rcu/jetfuel, ignoring prod modules. Assuming you've reached rocket tech, you now want to solve for E*P*n*(n+1)/2=1000 and increase your production incrementally.

eg: Let's say you want to target 2 hours this time, but still use 10 minute periods, so solving for E*10*12*13/2 = 1000, you get E = 1000/780, or about adding 1.3 worth of the trio - LDS/RCU/Jetfuel production per minute every 10 minutes.

And it check outs, 1.3*10 + 2.6*10 + 3.9*10 + .. + 12*1.3*10 = 1014

https://www.wolframalpha.com/input/?i=1 ... +12*1.3*10+

However, with rocket rush, you have all techs and so you have bots at that point. It would be more efficient if your bot count is increasing at an accelerated rate.

In that case, you want to be summing exponentially increasing production (say 50% per minute). Eg, in minute one, you are producing 1.5 trio per minute, and in minute 2 you are producing 2.25 per minute, 3.375 per minute in minute 3.

This would be the sum(p^N) = p(p^n-1)/p-1, where p = 1.5. Easier to wolfram here, https://www.wolframalpha.com/input/?i=s ... k+%3D+1000

You get about 14.5 minutes of rapid exponential production increases to produce 1K rocket trio. An interesting challenge for sure, but doable I believe with appropriately planned imported blueprints. Note speed modules are required to handle significant back end consumption of rocket parts.

You could just build all your production at once rather than incrementally, but you'll either end up having to build too much or sit around twiddling your thumbs while waiting for it to end. Ideally, you want to be adding production and speed modules up until the very last minute.

The formulas of course is obviously not exact, though very useful as a sanity check to get a rough sense of what is required.

If you want exact, than a full step by step spreadsheet which enumerates each phase and production resulting from that step. That can be an extensive effort however.

[Koub]

@blazespinnaker : are you sure you're replying to the right thread ? I can hardly see the link between your analysis and the OP.