Factorio flaws

Post all other topics which do not belong to any other category.
Frightning
Filter Inserter
Filter Inserter
Posts: 807
Joined: Fri Apr 29, 2016 5:27 pm
Contact:

Re: Factorio flaws

Post by Frightning »

Hedning1390 wrote:
Disclaimer
First of all I'd like to say I love this game and I have over 500 hours on steam, with a further couple of hundred hours not logged. That said I get a little triggered when everyone on these forums (and the steam forums) praise it as if it's the best game ever with no flaws, so I have compiled a list of some of the flaws. The list is somewhat ordered with the worst flaw first. Please don't take this list as a personal attack.
tl;dr version (please read the full text before getting angry):
1: There's only a handful of buildings, with few or no side-grades.
2: The beacons and requester chests pits creativity against efficiency.
3: The "win" condition is totally misplaced.
4: The biters are confused and almost a scam if considering the promos.
5: The lazy reskins are weird to me.
6: It's basically just a bunch of recipes and adding recipes is not very advanced.
Full version:
1: There's only a handful of buildings, with few or no side-grades. The 3 assemblers and 3 furnaces are straight upgrades, so can be considered as 1 each. The assembler, chemical plant, refinery and furnace while doing the exact same thing doesn't compete with each other because each recipe requires one and exactly one of them. This makes for a rather linear and boring experience. The only side grade in this game is your choice in power. The way solar panels, accumulators, boilers, nuclear reactors and transportable steam work and are balanced against each other is sublime. Imagine if the 3 assemblers and 3 furnaces could be changed in a similar way, from boring straight upgrades to a strategic choice, how much more interesting the game would be.

2: The beacons and requester chests pits creativity against efficiency. Even with the lackluster content you can still make some beautiful and creative designs. However the two mentioned things severely hamper this because the implicit goal of factorio is to be efficient:
-Requester chests kills belt logistics: Every recipe can be created with a requester-asember-provider chain. If you are not using this chain for everything but perfectly straight belt lines you are making your factory less efficient.
-Beacons force perfect grids: The most efficient way to produce things is to have 8 beacons per assembler and 8 assemblers per beacon. This can only be accomplished by placing them in parallel straight lines. Any attempt to use direct insertion breaks the pattern and thus cannot be used. Another design, efficient in another way, is to have the maximum 12 beacons per assembler. This can also only be accomplished one way which is already a boring square. On top of this each square has to be placed in a perfect grid else you pay for more beacons than you have to.

3: The "win" condition is totally misplaced. You win the game so fast that the people who only play to the "win" won't experience much of the game unless they go out of their way to do so. For the rest of us the "win" is totally meaningless, because it is just a science pack among others, not more significant than when you set up purple science. From my perspective it would be better to just remove the "win" prompt and treat it like any other infinite builder game.

The greater point is about pacing, and it ties up all the first 3 points here. While people who play to the "win" miss out on game content, so do those of us who play continuously. The beacon problem (above) is not a problem if you only play to the first rocket, because almost none of your assemblers will have beacons and modules. Neither is the first point a problem, because at least straight upgrades change what you can do in your designs, and the refinery introduces multiple outputs, which is sufficiently different to feel fresh the first time you unlock it.

If the intent is that we should only play to the "win" then I question even more the asking price of €25, because there are few random factors and even fewer branching and exclusionary choices that makes doing the first 6-10 hours over and over interesting. The most common argument I hear in defense of the price is how many hours you can spend, but even though your very first newbie play-through may take more than 10h you really can't spend that many hours in the pre rocket phase unless you are interested in speedrunning. Therefore I conclude that continuous games are essential for the value of the game, and therefore point 1 and 2 are problematic flaws.

4: The biters are confused and almost a scam if considering the promos. It seems to me that the developers have changed their view on what they wanted the biters to be from their original plans, but just left them in limbo. Almost every aspect of this enemy is simple and boring. There is lots of creativity when it comes to weapons, but because the enemy is so boring this is sadly almost wasted effort. I could explain why they are boring, but clearly the direction now is to just reduce their presence in the game, and it makes me wonder why we can't just get a "/c for all revealed chunks if entity=enemy do destroy" bomb.

5: The lazy reskins are weird to me. Compared to the other flaws this is a very minor one. I'm talking about many things only differing in color. The circuits, science packs, inserters, belts, chests, modules, 3 of the ores, etc. All several versions but with only a color differentiating them. The game is so incredibly polished in so many other aspects, so I don't understand why this is.

6: It's basically just a bunch of recipes and adding recipes is not very advanced. I could add hundreds of new recipes in the blink of an eye, and since the game is based almost solely around these recipes I can expand the game massively this way. The only effort is in balancing these recipes to make an interesting experience. This is mostly a problem in relation to the price, as being small and simple isn't a bad thing in and of itself. I'm not saying €25 is too much (although it may be considering the no sales policy), but some people suggest it is worth even more than this, and there is no objective ground for that.
I disagree with most of this honestly, but I want to give a constructive rebuttal:

1: Unlike the other buildings you mention, these are things you build within the first hour of gameplay (unless you're doing a lazy bastard run and have to mine up a stockpile of materials first) and continue to need, in quantity, for the rest of the game. It makes some sense to have some verticality to those buildings. They are also always more expensive to build the higher tier buildings, so the lower tier buildings may be worth continuing to use even when you have a higher tier available for at least some amount of time or in certain cases. Finally, they are not, strictly speaking, upgrades over the previous tier: the most significant example is Steel furnaces v. Electric furnaces regarding fuel usage for steam-based power setups; you have to use 2x Eff1 modules in the Electric furnaces in order to make them more Coal/Solid fuel efficient than Steel furnaces, but it's not the only example, Assembly machine 1s are actually more energy efficient than Assembly machine 2s (90kw for 0.5 crafting speed v 150kw for 0.75 crafting speed). Belts are also tiered for much the same reasons as assembly machines and furnaces are (and there, the fixed cost of upgrading is pretty massive).

2: I generally agree with your criticism of how beacons are warping layout choices for module based setups (the modules themselves and how they interact are partly to blame for this, but so too are the beacons themselves). However, logistics robots are not straight up better than belts. (This coming from someone who much prefers robots to belts, I might add.) For one, belts have no energy costs, whereas robots can actually use a LOT of energy. The usual argument for why bots are better is that they allow for more compact layouts and work nicer with beaconized setups. But that massive energy cost means you will likely lose whatever space gains you made and then some due to needing more power infrastructure (this is especially true for solar users, who should take energy efficiency seriously). In my opinion, what really makes bots desirable is simplicity. They are a simple, yet very flexible solution to the problem of local transportation of materials (long distances are better left to trains).

3: Pretty sure most people aren't launching a rocket in less than 20 hours unless they're trying to speedrun the game. It gives a good 'pie in the sky' goal for new players that also serves to give them some direction in what they are ultimately building towards. Space science research is for people that want to go beyond the core game and build more, and bigger designs, and challenge themselves. So I fail to see how the 'win condition' is 'misplaced'.

4: Biters being easy is, mostly, a function of map settings. If you find them too easy, you can rectify that by changing your map settings so that they are more of a challenge (I've had plenty of fun with just changing the starting area size to very small so that I have to deal with them buggers from the very beginning, for example). Eventually you will reach a point where beating them in combat is trivially easy, because your technology is that advanced, but that is par for the course for everything in the game (things become easier to do as you progress down the tech tree). Otherwise there really wouldn't be much reason to research better tech (or the reason would be: just to keep up, which if you increase evolution rate enough, can become difficult to do).

5: The 'laziness' here is actually both efficient and useful, the recolored but otherwise identical icons tell you that they are clearly similarly purposed components and yet distinct (color making the latter obvious). It's convenient that this also takes not much effort from the devs to achieve this result (win-win).

6: This is a criticism that I have some agreement with. Most of the recipes in this game follow the pattern of combine items A through F (or fewer, typically no more than 4, actually), to get new item Z. They have added some recipes that use a catalyst which motivates a feedback loop design (Coal liquefaction and Kovarex process come to mind), but overall I agree that too much of the stuff we make in the game is 'combine components to make item'-style recipes. At this point I think Factorio itself is too far along to really 'resolve' this issue. I would rather see Wube finish Factorio (1.0+whatever they want to do with it in the near term post 1.0), and that their next project be a new game that is a 'spiritual sequel' to Factorio (I don't actually want it to be a Factorio 2 because that would mean they have expectations from Factorio 1 to meet, and I'd rather they not confine themselves like that). For that next, future title, I would LOVE to see them explore a similar style of game with much more recipes that make multiple products or use/need catalyst(s). That would create more varied and intricate design decisions for players to engage with, but it's something that, imo, is beyond the scope of Factorio itself at this stage of it's development (it's fair game for mods though, and mods like Bob's Mod have shown that there is worthwhile design space there to be explored).

Hedning1390
Fast Inserter
Fast Inserter
Posts: 231
Joined: Fri Jan 05, 2018 8:47 pm
Contact:

Re: Factorio flaws

Post by Hedning1390 »

Frightning wrote:...
1. Is your point here that they are actually side-grades because they require slightly more power? That's pretty weak. Why are you putting so much emphasis on power usage? It really isn't a big deal. Are you really keeping steel furnaces because of the power usage of electric? I think you are pretty alone doing that.

2. I didn't say logistics bots are straight up better than belts. I even gave an example where belts are better. It's the simplicity you mentioned which was the target of my complaint. You don't need to solve logistics problems when there is a "one fits all" solution called bots.

Btw again with the power arguments? The benefit of having compact designs is that logistics becomes easier and more efficient. The total space power included doesn't matter. You don't have to place your powerplants in the middle of your production. You can place your powerplants anywhere. Preferably somewhere out of the way. Map space is infinite. It is a non-issue.

3 If it takes 20h with an asking price of €25 it's terrible value for money. See the part where I talk about why continuous games are essential for the game's value.

4 Biters are like lawns. I can choose one that forces me to mow it every 2 days, or I can choose one that I only have to mow every 2 weeks, or a natural one that doesn't need much keeping at all. Neither is difficult, its just more or less of a chore.

5. If it's useful I guess you would prefer if the smg, shotgun, rocket launcher and flamethrower had the same icon, just with different colors too? I disagree strongly with this. If you can identify different weapons even if they look different you should be able to identify different storage containers even if they look different.
ASDFGerte wrote:..
I was talking about the balance between boilers, reactors and solar. Not how power works in general or the redundancy of efficiency modules (you may want to have a chat with above poster btw lol). Do you think the 3 ways of making power are poorly balanced?

1 dollar per hour is bad unless it is a very specific category of game. Sandbox games especially should expect a lot more playtime than that.
bobucles wrote:Proceeds to copy the exact same thing they've seen a million times before
Hello mr smart inserter. Long insterter here. Look at the little box to the right of all my posts. I haven't seen that much. I'm sure in a forum for a game that has existed for 6 (?) years most things have been said before.

Hannu
Filter Inserter
Filter Inserter
Posts: 850
Joined: Thu Apr 28, 2016 6:27 am
Contact:

Re: Factorio flaws

Post by Hannu »

Hedning1390 wrote: 1: There's only a handful of buildings, with few or no side-grades. The 3 assemblers and 3 furnaces are straight upgrades, so can be considered as 1 each. The assembler, chemical plant, refinery and furnace while doing the exact same thing doesn't compete with each other because each recipe requires one and exactly one of them. This makes for a rather linear and boring experience. The only side grade in this game is your choice in power. The way solar panels, accumulators, boilers, nuclear reactors and transportable steam work and are balanced against each other is sublime. Imagine if the 3 assemblers and 3 furnaces could be changed in a similar way, from boring straight upgrades to a strategic choice, how much more interesting the game would be.
I agree. Mods can give more interest through complex material flows, but after all, they are also too perfect and lead to simple solutions. In my opinion there should be variable environmental conditions in the game. It would make possible to make more interesting chains and strategic decisions. If everything is exact and predictable, game will be straightforward run to predefined goal. Optimal solutions are easy to find and only decision is that do you make optimal base or do you use some own rolegame restrictions.

But after all, Factorio is intended to be entertaining video game for everyone and not additional work for bored engineers. I understand very well why straightforward simplicity is best solution it you want huge sales from casual players.
2: The beacons and requester chests pits creativity against efficiency. Even with the lackluster content you can still make some beautiful and creative designs. However the two mentioned things severely hamper this because the implicit goal of factorio is to be efficient:
-Requester chests kills belt logistics: Every recipe can be created with a requester-asember-provider chain. If you are not using this chain for everything but perfectly straight belt lines you are making your factory less efficient.
-Beacons force perfect grids: The most efficient way to produce things is to have 8 beacons per assembler and 8 assemblers per beacon. This can only be accomplished by placing them in parallel straight lines. Any attempt to use direct insertion breaks the pattern and thus cannot be used. Another design, efficient in another way, is to have the maximum 12 beacons per assembler. This can also only be accomplished one way which is already a boring square. On top of this each square has to be placed in a perfect grid else you pay for more beacons than you have to.
In my opinion this is flaw number 1. You begin with interesting building game but after some point it is pointless copying of trivial structures, if you do not want to restrict yourself.
It seems that huge bot based megabase with simple mathematical perfect production ratios is some kind of default choice and every development decision must support it. I think that simple solution, collision detection of bots, would solve it with very interesting way. After that bots could be used to avoid too complex and ugly spaghetti in machine workshops etc. but they would not be usable for large volume transportation because crowding around endpoints. But of course it would ruin all megabases and get most players angry.
3: The "win" condition is totally misplaced. You win the game so fast that the people who only play to the "win" won't experience much of the game unless they go out of their way to do so. For the rest of us the "win" is totally meaningless, because it is just a science pack among others, not more significant than when you set up purple science. From my perspective it would be better to just remove the "win" prompt and treat it like any other infinite builder game.
I think win condition is quite OK, if it must exist at all. It is always very artificial choice in sandbox game but it seems that some players want to have strict predefined goal or they do not feel game entertaining. I think that for beginners it takes 20-40 hours to achieve win condition which is quite typical duration of playthrough in games. Then player can try more efficient, maybe bigger or more biters etc. and move to next game. I am sure that vase majority of million players will not play game though (up to rocket) many times. Most of them would never build hundreds of hours even if win condition would need it.

4: The biters are confused and almost a scam if considering the promos.
Yes, they could give much more. But I like them even in current state. They give interesting addition in midgame, but when you begin to build big they are just resource sink, which is OK. I like to have somewhat varying and unpredictable resource sinks. But well thought automated robot warfaring with programmable robots would fit so well in Factorio.
5: The lazy reskins are weird to me.
I do not care about that. In my opinion devs put too much work to get everything look good, but I understand that it increases sales more than complex game mechanics.
I'm not saying €25 is too much (although it may be considering the no sales policy), but some people suggest it is worth even more than this, and there is no objective ground for that.
How there could be an objective measure worth of what something is? Especially entertaining thing which is not needed at all. I have got order of magnitude more interesting playing from Factorio and Kerbal Space Program than from any AA game from big company. From that point of view I could have paid much more or I could also buy interesting content if these game companies would offer it. Unfortunately they probably give more eye candy and technically challenging content comes from mods (My conscience nags to me, I should also support some modders financially, I have enjoyed so much from their work).

gsezz
Long Handed Inserter
Long Handed Inserter
Posts: 60
Joined: Thu Jan 04, 2018 8:23 pm
Contact:

Re: Factorio flaws

Post by gsezz »

I've seen typical game times ranging from 10 hours to 60 or more hours for the first rocket.
I'm at least 100 hours into a map before I even think about the rocket, 150 until I launch it... :D

1: I'd like to see slow/fast versions for chemical plants too. But otherwise I'm happy.

2: Agree. Simple solution: I don't use beacons for production, only for smelting. But my reason is that I just don't want to figure out new ratios again, after calculating them without beacons. And I only use bots for quick hacks, or if I don't want to fiddle around with belts at the moment. That's my choice, and I'm glad the game allows such choices.

3: Whats the issue with the win message? Just ignore it, but leave it in the game for the speed runners. Maybe someday there could be an option to disable it, but there are plenty more important things to implement and fix.

4: Yep, the biters will need a complete rework at some point. A better AI, so that they adopt to defense lines and don't act as dumb as they do. Having huge groups of enemys getting wrecked by a hand full of landmines at the same spot, over and over again ist boring. There were strategy games with a better AI 20 years ago.

5: I totally disagree! I don't want to have different models for things that do the exact same thing only with a different speed, or a little twist. To differentiate them by color is in my opinion the perfect solution. I just wished the color scheme would be a more consistent throughout the game. Belts are yellow>red>blue, inserters are grey>yellow>blue, machines are grey>blue>green, ammo is yellow>red(>green). But I understand that it's probably to late to change this. People are already used to it.

6: Objective ground for the price: entertainment per cost. At this point I got over 900hrs of playtime for only 25€, that's a great value. The only other game that kept me entertained for longer was a subscription based MMO, that costed me several hundred € over the years. I own a lot of so called tripple A titles for 60€ and more, and not a single one has more than 100hrs of playtime, most of them barely get to 20hrs. The 25€ I spent on factorio were probably the best 25€ bucks I ever spent on anything regarding entertainment. It doesn't matter how simple a game is, or seams to be, the value is what you get from it.
I don't understand how you could even criticize the price, if you have played for several hundert hours yourself. I'm honestly interested in wich other game you played more for less money, because I'm constantly searching for good games. Free to play games with different monetization concepts are not comparable.

Hannu
Filter Inserter
Filter Inserter
Posts: 850
Joined: Thu Apr 28, 2016 6:27 am
Contact:

Re: Factorio flaws

Post by Hannu »

Frightning wrote: 1: Unlike the other buildings you mention, these are things you build within the first hour of gameplay (unless you're doing a lazy bastard run and have to mine up a stockpile of materials first) and continue to need, in quantity, for the rest of the game. It makes some sense to have some verticality to those buildings.
In perfectionist or OCD point of view. But it would be technically more interesting if you had to make choices between different assembling machines. Now you can begin with lower tiers and replace them later with higher tier so that old machines are fully used. It would be more interesting to decide if I build fast cheap production chain, which I can not convert to better later without scrapping machines, or use more time and resources to build from better (and different) components so that it would reward in longer period. Factorio lacks almost completely such mechanics, which (in my opinion) would fit very well in this kind of engineering game. Because those are typically the most important decisions in real engineering. Optimizing last percents of performance, which is very popular in Factorio players, is extremely rare and special case in real world engineering. Maybe, if you develop some high tech sporting equipment or spacecrafts, but not in vast majority of industry.
They are also always more expensive to build the higher tier buildings, so the lower tier buildings may be worth continuing to use even when you have a higher tier available for at least some amount of time or in certain cases. Finally, they are not, strictly speaking, upgrades over the previous tier: the most significant example is Steel furnaces v. Electric furnaces regarding fuel usage for steam-based power setups; you have to use 2x Eff1 modules in the Electric furnaces in order to make them more Coal/Solid fuel efficient than Steel furnaces, but it's not the only example, Assembly machine 1s are actually more energy efficient than Assembly machine 2s (90kw for 0.5 crafting speed v 150kw for 0.75 crafting speed). Belts are also tiered for much the same reasons as assembly machines and furnaces are (and there, the fixed cost of upgrading is pretty massive).
I disagree these things. In theory, you are right with furnaces, but in practice energy is free in Factorio and there are more than enough resources if you do not especially select resource poor or deathworld settings. I have always cleaned large ore patches with commands to get my base look nicer and reason to build more realistic large rail network.

That high cost of belts is not true. You can see production stats and calculate how much resources belts have taken. I had once relatively large belt and train based factory, which used about 1 M iron plater per hour. Total cost of blue belts after about 200 hours and all expansions was about 1 percent of total production. Other infrastructure (except maybe tier 3 modules) costed even less, even I did not use beacons and needed hundreds of assemblers. Belts, assemblers etc. are practically free in Factorio. Their cost is negligible compared to research or rocket launching costs. Also, if you play until first rocket launch, cost of base is almost nothing compared to cost of research and rocket manufacturing.
However, logistics robots are not straight up better than belts.
Yes they are. All biggest bases use bots and even best computers can not handle belt based megabases nearly as large. It is very clear thing technically. Factiorio is optimized for bot transport, period. Only way to argue belts better are aesthetic or rolegame like.
For one, belts have no energy costs, whereas robots can actually use a LOT of energy. The usual argument for why bots are better is that they allow for more compact layouts and work nicer with beaconized setups. But that massive energy cost means you will likely lose whatever space gains you made and then some due to needing more power infrastructure (this is especially true for solar users, who should take energy efficiency seriously).
No, in largest bases processor performance is the only restrictive resource. Everything in Factorio world are free and infinite, if you do not set limits by yourself. It just needs some playtime to set up solar fields, but you can probably find enough ores to fill the whole world with solar panels.
In my opinion extreme megabases should not be the most important developing target, but it seems that there are so many youtube players and their fans that it is practically true. Many interesting suggestions have blocked because "it needs too much CPU cycles and megabases would suffer".
If you find them too easy, you can rectify that by changing your map settings so that they are more of a challenge (I've had plenty of fun with just changing the starting area size to very small so that I have to deal with them buggers from the very beginning, for example).
I know I can adjust number of biters to insanity and over that, but I would like to have some intelligent challenges instead of just running around base to fill turrets and fix walls. Plan defense and attack strategies, allocate production resources, program robot armies to do dirty work against enemies who feels to be intelligent and natural creatures instead of numberless hordes of lazily programmed robots who just run forward and kill everything they encounter without plan or objective until killed.

But my overall opinion is still positive. They are well balanced and annoying enough but not too much. I would not change them, except by making trains kill all enemies.
5: The 'laziness' here is actually both efficient and useful, the recolored but otherwise identical icons tell you that they are clearly similarly purposed components and yet distinct (color making the latter obvious). It's convenient that this also takes not much effort from the devs to achieve this result (win-win).
I agree. Some mods have dozens of beautiful production entities, which looks like real production plants, but their small icons are very difficult to distinguish.

Aeternus
Filter Inserter
Filter Inserter
Posts: 835
Joined: Wed Mar 29, 2017 2:10 am
Contact:

Re: Factorio flaws

Post by Aeternus »

ASDFGerte wrote:
Hedning1390 wrote:What is your issue with power?
For normal maps (there are some specially designed scenarios where this doesn't apply), power is usually just plopping down some steam engines whenever you need more. Then when you reach higher tech, you do the same with reactors. Then in very late game, you either use deactivated nuclear or plop down a million friggin solar panels to save ups. Apart from said specially designed scenarios, i've never had to actually plan fuel use ahead of time or do anything to reduce power usage, let alone using efficiency modules (who the §$%§$ uses efficiency modules, except for building power armor mk2?).
I do. Efficiency mines (and even smelters) on an isolated power grid eliminates the need to run power to remote mines. A local solar power grid, some accus and blammo, selfsufficient outpost.

One thing I will agree with is that smelters are indeed boring. There is no "bulk smelter" in vanilla, and in megaplants, smelters quickly become the bottleneck and end up being -huge- clusters. A 3x 24*24 block (24*60 for steel) beaconed to get them all working at 5ish speed (with production mods in there) makes up the bulk of my factory. And yes. That gets boring fast, not to mention tedious to set up.

Zavian
Smart Inserter
Smart Inserter
Posts: 1641
Joined: Thu Mar 02, 2017 2:57 am
Contact:

Re: Factorio flaws

Post by Zavian »

Frightning wrote:I generally agree with your criticism of how beacons are warping layout choices for module based setups (the modules themselves and how they interact are partly to blame for this, but so too are the beacons themselves).
A suggestion that came up during the bots versus belts discussion is to rework beacons so that each additional beacon gives a diminishing return. eg the first beacon could give 100% bonus, the second 50%, the third 33% etc. (or 100%, 50%, 25% etc) . By moving to diminishing returns you can hopefully get people to start building more diverse and interesting setups with less beacons leaving more room to route belts etc, rather than the currents 8x8 bot based builds.

Here is a layout to make 20 blue belts worth of green circuits. I was never really happy with it because it only fits 6 beacons around each assembler, when a simpler bot based build gets 8 beacons per assembler. If the 7th and 8th beacons had diminishing returns I might have used it more. (Though it was also a pain from a belt layout pov. The 20 belt version needs copper and iron from both the east and the west, and outputs green circuits to both the north and the south. And when it was designed in 0.14 belt performance really wasn't up to a belt based factory on this scale).
20BeltGreenCircuits.jpg
20BeltGreenCircuits.jpg (455.77 KiB) Viewed 5493 times

bman212121
Fast Inserter
Fast Inserter
Posts: 141
Joined: Mon Apr 18, 2016 8:08 pm
Contact:

Re: Factorio flaws

Post by bman212121 »

Hedning1390 wrote:
bman212121 wrote:You are forcing yourself into a "boring" play style because you can't envision doing it any other way. If you can't get past having to use beacons because they are there, then download a mod to just turn them off. That will force you to actually come up with new and challenging ways to tackle the same problem.
I have already replied to that:
Hedning1390 wrote:If you are not using beacons you will have to build so much more factory for the same production. And if you don't use productivity modules to keep the speed reasonable now your entire factory will pretty much consist of furnaces, which would both look boring and repetitive, and be boring and repetitive to build too.
If you think creativity is not mutually exclusive with efficiency then what is your definition of efficiency, because it is clearly not the same as mine.
The whole point of the game is to build things, if you find that boring and repetitive then you probably need to take a break and move to something else. That said, even if you have 10,000 furnaces on the map, it does not need to be "boring and repetitive". If anything the blueprints are what make it boring, because it gives a player a lazy way out so they don't have to think anymore. But if you can use your imagination, you can challenge yourself and stop making the game boring. Rather than just making one big setup and blueprinting it, how about trying to come up with a different way to do the same thing each time you build it? Rather than just building a row of beacons, and then slapping factories next to those, try making different patterns.

Efficiency in Factorio can mean many things, you need to look past resource efficiency. You can be efficient with: Power, space, pollution, resources, belts, inserters, etc etc. Using a boring beacon layout is doubtful that it's going to be efficient for pollution or power, so maybe see what you can do to make a base that uses the least amount of power possible. Or make a base that has almost no pollution. Or maybe challenge yourself to see if you can build a factory that uses the least amount of inserters possible. That would require you to try to chain as many items in a row as you can, and definitely wouldn't have room to waste on beacons. I also bet it won't look boring and repetitive, because it will require you to un-align things from the grid in order to use less inserters. There is no achievement or thing to check off in game to provide you with alternate scenario, but using your creativity you can probably come up with new ways to be efficient.

bman212121
Fast Inserter
Fast Inserter
Posts: 141
Joined: Mon Apr 18, 2016 8:08 pm
Contact:

Re: Factorio flaws

Post by bman212121 »

Bauer wrote: Last week, I worked for about 6 hours to create a building-block that produces 600 green circuits/s with Cu and Fe plates as input. I was going to multiply this by at least 3, maybe 5-6.
Being bored of the assembler, beacon rows, I tried to create something new. I came up with a belt-based solution, not perfectly efficient, but a design that would make the assemblers work continuously with 6 beacons to allow for belts. My UPS dropped so badly (from 58 to 11 after building 2 of such building blocks, unplayable!!) that I decided to tear everything down. It took me 20 minutes to design and build a bot-based 8/8-type building block. The impact on UPS is negligible. That's a sad story.

This is why I think that I am forced to NOT diversify.

I have a working design, it does what I want it to do, I like watching it, I'd still prefer to watch a belt based setup. But, hey!, I know I cannot have everything.
It would be great to read a FFF about this to see the developers view on things.
Well in the case that you're making, I do think there is not really going to be a way around that. The game itself can only be so efficient, and the scale at which you're trying to make things is basically unprecedented. I get why you'd maybe want to build something that big for infinite research, but at the same time I think your issue is a top 1% of the top 1%. 600/s works out to be ~2.1M / hr, which is 7x higher than the highest achievement in the game. Then you wanted to take that and go up 5x, so like 35x or making 10.8M green circuits / hr. I'm sure there is probably a small number of users in the forums who are trying to push the envelope like that, but under normal play through you don't even need to break 300k / hr to win the game. My current base which is on a scale bigger than I've ever designed before can only support around 500 - 600K / hr, so even just one of your setups is already 3 - 4x the biggest thing I've ever created. (Which is already bigger than any setup I've seen created by anyone else I play with)

I do wish the actual numbers of the game were scaled down a bit more, because the sheer volume of items does becoming overwhelming. In order to make those 600 / s, you also need to be able to make 1800 / s copper cables, 900 / s copper plates, 600 / s iron plates, and 600 / s copper plates. That means the game is handling 4,500 items / s without any type of buffer. But on belts that would mean you might have 10x that amount in play, so 1 setup might have 50K items in motion, so two setups would require updating 100K items every second. That's simply a ton of stuff to update, so I can see where it's just not going to work. The devs have stated the biggest limitation to the game right now is memory bandwidth, they can only change around 400MB per tick, otherwise the UPS will go down. When the numbers start going through the roof I don't know if there is much you can really do about it, as you're hitting a limit of what your machine can handle, and not really a limit of the game. If you could take a couple 0's out of the research, ie infinite starts at 10 pots instead of 1,000 pots, it would give more headroom before you hit the wall. But at the same token I'm sure someone will try to push the numbers well past that, and hit the same limitations again.

EDIT: I am curious though, so what type of system did you end up hitting the limit on? The main factor is going to be memory speed. If you're hitting the limit on say DDR4-2400, you might be able to pick up some DDR4-4000 and if I were guessing that should be about 2x the memory bandwidth. Probably not enough to save what you're doing, but would still give a lot of headroom for the times where you're just starting to hit the limit. I'm only using DDR3-1600 and I haven't seen UPS drops in vanilla. But The thing to keep in mind is that Factorio is synchronous so the server controls the pace. IE my server might only be running DDR3-1333, so even though the memory in my pc is faster, it doesn't matter. If you have clients that are slower than the server, they will just get dropped or not even be able to join. (I have an older system with DDR2-800, and in late game it's not possible for that machine to ever catch up or even join into the game) I'd need both the server and my client to be high end DDR4 if I wanted to play on a map that couldn't run on just DDR3, or you'd have to play it in single player so then only your pc's memory bandwidth matters.

Hedning1390
Fast Inserter
Fast Inserter
Posts: 231
Joined: Fri Jan 05, 2018 8:47 pm
Contact:

Re: Factorio flaws

Post by Hedning1390 »

Zavian's picture perfectly showcases the flaw of both belts and beacons. Thank you Zavian for posting it

To all you who don't see why:
A fully beaconed such setup draws 10 iron, 11.8 copper and produces 15.4 circuits (/s/assembler pair). In his setup he is only producing a maximum of 10 circuits, which is a massive decrease, and limited by belts, so his assemblers will stall from time to time. Also using 1:1 ratio when the ideal ratio is 14:15 those circuit assemblers will stall even if he had the belt capacity.

Any way you look at it it is less efficient: His belts needs to loop around the entire structure to get entry from all sides. What we don't see is the massive belt spaghetti between this and the train stations. He is using a lot more resources in modules and assemblers, because he needs more of them to produce the same thing, and the power usage is greater for the same amount of product (I know some of you are obsessed with power usage).

He isn't even diverting from the row design that much and he still loses a massive 35% efficiency.

This isn't just because of beacons neither. He is saturating the belts fully. If the assemblers were to work at full capacity (fully beaconed) he could only go 1 row deep with belts, meaning this would shrink from 8 rows to 2, and the rows themselves would decrease from 10 to 6 (could possibly maintain 10 with braiding), which creates inefficiency in beacon (and speed module) use since there's a lot more edge beacons. This would not be the case if he switched to bots. Switching to bots would also remove the belt spaghetti between the blueprint and the train stations and would allow for the ideal ratio of 14:15.

And to those that thinks this is "unprecedented" "top 1% of the top 1%". 20 belts is what you need for 1k science per minute, which is like the first goal in a continuous game. If you have ever heard the phrase "megabase" 1k and up is what that phrase refers to. You are seriously missing out on the game if you don't ever play to this stage. There are so many new challenges and you get so much insight into the game once you reach this stage.

Here's the picture for those that don't want to scroll up: download/file.php?id=36875

Bauer
Filter Inserter
Filter Inserter
Posts: 346
Joined: Fri May 05, 2017 12:48 pm
Contact:

Re: Factorio flaws

Post by Bauer »

Hedning1390 wrote: He isn't even diverting from the row design that much and he still loses a massive 35% efficiency.
And let me add: This is the best you can do.
A 8/8-design requires that you put assembing machines for Coils and Green Circuits in different rows.
Since one stack inserter belt to machine has not sufficient throughput, you need two (tunneling the row of beacons with underground belts).
2 inserters for Coil + 1 for Fe-plates + 1 for output is too much if you also want to sqeeze in a power pole.
Of course, you could use half of a belt for Fe-plates and for output. However, then you can use 2 assemblers that will also be understaturated. I wouldn't want to call "2" a row.

And yes, the spagetti is a nightmare, especially if you want to get the ratios right.

Bauer
Filter Inserter
Filter Inserter
Posts: 346
Joined: Fri May 05, 2017 12:48 pm
Contact:

Re: Factorio flaws

Post by Bauer »

bman212121 wrote:
Bauer wrote: Last week, I worked for about 6 hours to create a building-block that produces 600 green circuits/s with Cu and Fe plates as input. (...)
I'm sure there is probably a small number of users in the forums who are trying to push the envelope like that, but under normal play through you don't even need to break 300k / hr to win the game. My current base which is on a scale bigger than I've ever designed before can only support around 500 - 600K / hr, so even just one of your setups is already 3 - 4x the biggest thing I've ever created. (Which is already bigger than any setup I've seen created by anyone else I play with)
Hedning1390 answered already: With 46451 green circuits/min you can make (fully equipped with tier 3 prod. mod.) 1k science/min... which is where a megabase starts.
When you think about it... appropriately, it should be call kilobase. ;-)

Let me add: There is no such thing like too many green circuits.

User avatar
5thHorseman
Smart Inserter
Smart Inserter
Posts: 1193
Joined: Fri Jun 10, 2016 11:21 pm
Contact:

Re: Factorio flaws

Post by 5thHorseman »

Weird. Not one mention of what is actually a flaw in Factorio, which is that blueprints are this weird pseudo-item that clutter your inventory and are hard to manage, in spite of and sometimes because of the management tools given to us.

But yeah, not having 6 more levels of assembler is bad too. No wait. It's not.

Hedning1390
Fast Inserter
Fast Inserter
Posts: 231
Joined: Fri Jan 05, 2018 8:47 pm
Contact:

Re: Factorio flaws

Post by Hedning1390 »

Bauer wrote:And let me add: This is the best you can do.
A 8/8-design requires that you put assembing machines for Coils and Green Circuits in different rows.
Not unless you want to use the 14:15 ratio. This runs at the speed of the wire assembler: https://i.imgur.com/3FSfRn3.png
5thHorseman wrote:But yeah, not having 6 more levels of assembler is bad too. No wait. It's not.
I think you totally missed the point. I wasn't talking about making more assemblers, my complaint was that they were too similar already.
As for blueprints they are working on an UI update. It makes no sense for me to point out flaws in things they are still working on. Also while I agree blueprints can be improved I don't find them hard to manage nor cluttering my inventory. Try blueprint books and using the library. Keep a disposable blueprint on your hotbar for temporary stuff. You can clear a blueprint with shift right click.

Frightning
Filter Inserter
Filter Inserter
Posts: 807
Joined: Fri Apr 29, 2016 5:27 pm
Contact:

Re: Factorio flaws

Post by Frightning »

Hedning1390 wrote:
Frightning wrote:...
1. Is your point here that they are actually side-grades because they require slightly more power? That's pretty weak. Why are you putting so much emphasis on power usage? It really isn't a big deal. Are you really keeping steel furnaces because of the power usage of electric? I think you are pretty alone doing that.

2. I didn't say logistics bots are straight up better than belts. I even gave an example where belts are better. It's the simplicity you mentioned which was the target of my complaint. You don't need to solve logistics problems when there is a "one fits all" solution called bots.

Btw again with the power arguments? The benefit of having compact designs is that logistics becomes easier and more efficient. The total space power included doesn't matter. You don't have to place your powerplants in the middle of your production. You can place your powerplants anywhere. Preferably somewhere out of the way. Map space is infinite. It is a non-issue.

3 If it takes 20h with an asking price of €25 it's terrible value for money. See the part where I talk about why continuous games are essential for the game's value.

4 Biters are like lawns. I can choose one that forces me to mow it every 2 days, or I can choose one that I only have to mow every 2 weeks, or a natural one that doesn't need much keeping at all. Neither is difficult, its just more or less of a chore.

5. If it's useful I guess you would prefer if the smg, shotgun, rocket launcher and flamethrower had the same icon, just with different colors too? I disagree strongly with this. If you can identify different weapons even if they look different you should be able to identify different storage containers even if they look different.
ASDFGerte wrote:..
I was talking about the balance between boilers, reactors and solar. Not how power works in general or the redundancy of efficiency modules (you may want to have a chat with above poster btw lol). Do you think the 3 ways of making power are poorly balanced?

1 dollar per hour is bad unless it is a very specific category of game. Sandbox games especially should expect a lot more playtime than that.
bobucles wrote:Proceeds to copy the exact same thing they've seen a million times before
Hello mr smart inserter. Long insterter here. Look at the little box to the right of all my posts. I haven't seen that much. I'm sure in a forum for a game that has existed for 6 (?) years most things have been said before.
1: I used power usage as an example because it's an aspect of the game that a lot people don't appreciate because they choose mapgen settings that make power demands largely a non-factor. If you actually play a map where resources are not abundant, (or space free of biters for that matter), you will then start appreciate the importance of those things. It's like playing a game on the easiest settings and then complaining that the game is easy....it's easy because you chose to make it that way. Factorio does something which very few modern games do (that used to be FAR more common in the past) and that is giving the player a LOT of agency over specifics regarding how the game will play (game setup options, on average, have decreased in games dramatically since the early 2000s). There are people on this forum who've stuck w/ Steel furnaces even for endgame. To my knowledge, their reasons for doing so were partly power related, but also due to the 2x2 vrs. 3x3 footprint. I still have some Assem1s and many Assem2s even at endgame because if I don't need the throughput of an Assem3, there's no real reason for me to place one, same thing with belts, I usually only use Red/Blue where the throughput is actually needed (saves a ton of Iron, btw).

2: A 'one-size-fits-all' solution that is at the END of the tech tree, almost as if it's a reward for progressing that deep into the tree (you can spare yourself the 'headaches' of dealing with belt routing and related issues, if you so desire, at the price of logistics bots infrastructure and energy demands...so a generally higher fixed cost, and actual running energy costs). Until you actually get all the way down the tree, gotta deal with belts (and trains once that's researched).

3: Most single player AAA games have maybe 30-50h of content and are typically priced at $60 at release. Factorio already looks like a good deal next to those things, even before acknowledging that Factorio has vastly greater replay value than those kinds of games ever will. Let alone Factorio also has multiplayer, uses randomization to make replaying more varied and interesting, and is extremely mod friendly as well. Factorio is honestly more comparable in terms of replay value to a game like Skyrim than it is to the kinds of $60 titles I am referring to here, and it's goes for half the price Skyrim went for on it's release. (This doesn't even get into the whole DLC debacle and the fact that Factorio has none and probably never will have any; it just doesn't make sense to have DLC when Factorio is so moddable that people could make a free mod that provides precisely the content of said DLC).

4: Have you ever tried maxing out biter settings plus jacking up their evolution rates and expansion rate? Good luck surviving that, much less managing to expand for resources before what you can get from your starting area runs out (especially if you choose very small starting area). Again, map settings....learn how to use them to get the experience you want from the game. (Feel free to ask for suggestions from other vets on here, as well as experimenting on your own time). Eventually, the biters will become easy to deal with, just like every other problem in the game. Then you have a whole new set of challenges to consider (problems of scale are something only people that have actually built megabases/gigafactories have the experience to truly appreciate; even I haven't quite gotten to that point yet, though I will...eventually). Player time is one resource that absolutely fixed, and the longer you go in the game, the valuable that resource becomes and more important it is to use it efficiently. Every other resources scales, player time does not.

5: No, because shotguns don't look like smgs and neither of them look like rocket launchers. Circuits....circuits look like....well...circuits. So using a color-coded circuit-looking image for the 3 types of circuits in the game is entirely logical and on top of that, efficient.

Bauer
Filter Inserter
Filter Inserter
Posts: 346
Joined: Fri May 05, 2017 12:48 pm
Contact:

Re: Factorio flaws

Post by Bauer »

Hedning1390 wrote:
Bauer wrote:And let me add: This is the best you can do.
A 8/8-design requires that you put assembing machines for Coils and Green Circuits in different rows.
Not unless you want to use the 14:15 ratio. This runs at the speed of the wire assembler: https://i.imgur.com/3FSfRn3.png
I like the desing. However, it allows only 2 rows, does it?

Aeternus
Filter Inserter
Filter Inserter
Posts: 835
Joined: Wed Mar 29, 2017 2:10 am
Contact:

Re: Factorio flaws

Post by Aeternus »

5thHorseman wrote:Weird. Not one mention of what is actually a flaw in Factorio, which is that blueprints are this weird pseudo-item that clutter your inventory and are hard to manage, in spite of and sometimes because of the management tools given to us.
But yeah, not having 6 more levels of assembler is bad too. No wait. It's not.
Blueprint books anyone? I usually carry 4 of those: Rail, Base, Power and Combat/Defense. Each of those generally has less then 20 basic blueprints. When I want to piece together something larger I plop down one blueprint in a 4x4 grid, drag a normal blueprint around it and expand to 16x16, then repeat the pattern as neccesary to however big it needs to be. No clutter.

If you're finding the basic humdrum pattern boring, then experiment. Instead of doing metal + copper -> EC, then EC+Plastic+Copper -> AC, then AC+EC+Acid -> ProcUn, do a direct Metal + Copper + Plastic + Acid -> ProcUn + SpeedMods -> RCS, then repeat -that- brick until you have the amount of stuff you need. It'll also allow you to interlace smelters a little... a metal smelter here, a copper smelter there... If you make that belt based, it'll be slower then a full bot based brick, but it might look a mighty lot more impressive.

User avatar
5thHorseman
Smart Inserter
Smart Inserter
Posts: 1193
Joined: Fri Jun 10, 2016 11:21 pm
Contact:

Re: Factorio flaws

Post by 5thHorseman »

Hedning1390 wrote:As for blueprints they are working on an UI update. It makes no sense for me to point out flaws in things they are still working on.
I've seen no indication that they even think blueprints are unfinished, little alone that they plan to finish them.

But anyway I've derailed enough. I don't agree with your proposed changes, I think they'll just clutter the game, which is at peak clutter already.

Hedning1390
Fast Inserter
Fast Inserter
Posts: 231
Joined: Fri Jan 05, 2018 8:47 pm
Contact:

Re: Factorio flaws

Post by Hedning1390 »

Frightning wrote:...
I have played on "hard" settings. I've answered this already, my answer is up there somewhere if you want to read it.

On my "there is no spoon" run I used mainly stone furnaces simply because I didn't have time to change them. That early it doesn't matter what furnace you have. What you and I consider "endgame" is wildly different though and once you go beyond the first rocket you cannot have stone or steel furnaces. This is also partly what I mean by people who only play to the "win" missing out, because that early into the game there are lots of things in the tech tree that you might either just skip over, or don't need to even look at.

You might not be concerned about what happens post first rocket, but I am and I have explained why several times. This entire post is about continuous games. You don't need to convince me that pre 1st rocket play is good. What you need to do is either show how the flaws I pointed out in continuous games are actually not flaws in continuous games, or make an argument for why the game is worth €25 it even without continuous games.

Circuits may look like circuits, but why does processing units look like circuits? A processing unit could look more like a server. Why do logistics chests look like boxes? That makes no sense. How exactly are the bots interacting with them? Beaming them inside with star trek like teleporters? Why does iron ore look like stone ore and copper ore but uranium ore gets its own look? Why do all science look like flasks? Are you telling me we are liquidizing furnaces and engines and the mix is a purple fluid? Rocket fuel used to be a shade of solid fuel. It's not any more, but now nuclear fuel is a shade of rocket fuel. Would you prefer all of them being the same cube, just with different colors?
Bauer wrote:
Hedning1390 wrote:
Bauer wrote:And let me add: This is the best you can do.
A 8/8-design requires that you put assembing machines for Coils and Green Circuits in different rows.
Not unless you want to use the 14:15 ratio. This runs at the speed of the wire assembler: https://i.imgur.com/3FSfRn3.png
I like the desing. However, it allows only 2 rows, does it?
Yes. Even if there were some magic way of getting everything in a standard line the throughput required of 14 per assembler would limit it anyway, and a row of length 2 isn't really a row.

Frightning
Filter Inserter
Filter Inserter
Posts: 807
Joined: Fri Apr 29, 2016 5:27 pm
Contact:

Re: Factorio flaws

Post by Frightning »

Hedning1390 wrote:
Frightning wrote:...
I have played on "hard" settings. I've answered this already, my answer is up there somewhere if you want to read it.

On my "there is no spoon" run I used mainly stone furnaces simply because I didn't have time to change them. That early it doesn't matter what furnace you have. What you and I consider "endgame" is wildly different though and once you go beyond the first rocket you cannot have stone or steel furnaces. This is also partly what I mean by people who only play to the "win" missing out, because that early into the game there are lots of things in the tech tree that you might either just skip over, or don't need to even look at.

You might not be concerned about what happens post first rocket, but I am and I have explained why several times. This entire post is about continuous games. You don't need to convince me that pre 1st rocket play is good. What you need to do is either show how the flaws I pointed out in continuous games are actually not flaws in continuous games, or make an argument for why the game is worth €25 it even without continuous games.

Circuits may look like circuits, but why does processing units look like circuits? A processing unit could look more like a server. Why do logistics chests look like boxes? That makes no sense. How exactly are the bots interacting with them? Beaming them inside with star trek like teleporters? Why does iron ore look like stone ore and copper ore but uranium ore gets its own look? Why do all science look like flasks? Are you telling me we are liquidizing furnaces and engines and the mix is a purple fluid? Rocket fuel used to be a shade of solid fuel. It's not any more, but now nuclear fuel is a shade of rocket fuel. Would you prefer all of them being the same cube, just with different colors?
~snip~
I know of people on these boards that built at least 1 RPM megabases (pre 0.15 days) using Steel furnaces instead of Electric furnaces, it's viable, even if suboptimal by some measures of efficiency.

The whole issue with UPS will NEVER go away, at some point, your computer's performance will become the bottleneck because the simulation is that computationally intensive. The fact that bots are better for UPS (specifically when done in a modular fashion, as tends to naturally occur in train+bot based bases), is unsurprising if you understand a bit about what goes into simulating the bots versus belts. This is why the devs have worked hard for 0.16 on optimizing belt simulation code, completely redesigning the data structure to improve performance (by a quite respectable amount, I might add). Frankly, even most megabase builders don't actually min-max the UPS efficiency of their designs (have a look at the UPS wars thread to see what that actually can entail....things like clocked inserters and such).

Post Reply

Return to “General discussion”