Friday Facts #266 - Cleanup of mechanics

[bobucles]

As for ingredient limits being "yet another thing that had to be explained somehow," explain that the simpler machines are more limited in functionality, and can only craft simpler recipes.

I'm not sure there's an easy way to explain it. As it stands assemblers are "black boxes" where you put things in and new things come out. The nuances are not visible to the player and in truth they kinda never were. And before you say "yes they are" then tell me what this factory does. Or this one. Or this one.

factorio game.jpg (762.22 KiB) Viewed 6570 times

Notice a common theme? Without the all powerful information overlay active, you're almost completely blind. I think the information overlay has been a gigantic crutch for essentially all of development and no real game testing has been done without it. Even on the factorio guides one of the first steps for the player is "turn the UI on" because it's nearly impossible to play without it.

I think a big problem here feeds into the the important early game experience of "show don't tell". Players have a finite patience and it is more fun to watch things happen than to read about what should be happening. Some aspects of Factorio are handled extremely well in this regard. Mines do mining, belts do belting, and inserters swing around and it's a lot of fun to watch it all in action. Simple devices show simple feedback, but assemblers aren't very simple. They contain a lot of important information such as recipe, resources and activity and most of that information is critical to the gameplay at all times. So before telling me what the key essential difference between these three assemblers is:

assemblers.jpg (59.01 KiB) Viewed 6570 times

Maybe it'd be more constructive to see how the default feedback on their function can be shown in a better way?

With the video engine getting super optimized I think it's a great time to discuss how maybe things could fit better into the game's "natural" view. That way things can both look and play well without having to bury the experience in overlays. Maybe I'm talking crazy, but an overlay should help enhance the experience rather than being the core essentials.

[meganothing]

I agree everything because I think it will really clean things up. I do agree that the number of ingredients isn't the best mechanic for assembly machines. However, my initial thought is to replace this with the "technology level" of the items it can make (example below). It looks like a couple people had the same idea.

AM3: can make everything
AM2: can make everything unlocked by only science pack 1,2,3, military, and production
AM1: can make everything unlocked by only science packs 1 & 2

The problem with this idea is that it serves no game play purpose, it is just a limitation for limitation sakes.

Since the machines are plug-in replacements you don't need to adapt anything about your base design because of such a limitation, it just adds a small initial cost to setting up a factory. There are already reasons to use better assemblers and if someone wants to make an unusual design with lots of AM1's to achieve some metrics like cheapest factory for X, that is a good thing.

[quyxkh]

The problem with this idea is that it serves no game play purpose, it is just a limitation for limitation sakes.

Well, it's pretty clear that there's the crowd that regards design puzzles as impediments to play, and the crowd that regards them as the point of play.

[Guest_4544]

In my opinion to me it sounds like it the developers just haven't spend enough thought on it how those mechanics could be refined and/or exploited so that it is actually useful and in turn explain the new player how they are used in this context. Yes of course those mining hardness is pretty much useless because the only additional structure you can build that is a improvement over the coal powered drill is the electric drill and no additional mechanic that actually makes a use of the drill part itself like a fast and soft drill that is well suited for copper as those stone/metal types are relative soft and another drill for harder materials. So by introducing a new feature the drill bit it would introduce another gameplay depth that the player can exploit it to make his factory run even much better if he bothers learning this mechanic or he could chose to instead ignore it and use a generic drill bit for everything else if it confuse him too much. Obviously Factorio doesn't have this drill bit mechanic and in mod context they can only do so by adding a "drill module".

So in other words introduce the new player the core game mechanic what they essentially need to know everything in order to launch their very own rocket. The key is in explaining what the player needs to know, you could for example not explain those burner/fuel efficiency instead of just removing everything. Or maybe a simple/advanced mode so by default for the new player they are in simple mode so that they don't have to deal with much stuff and if the new player thinks he is ready to take on new challenges or complexity he could pick the advanced mode instead were less features is hidden or more is unlocked.

I'd suggest that you developers at the very least keep the modding part of those mechanic untouched so that at least a modder can make use of this feature if they wish instead of introducing a rather unnecessary "API" incompatibility change, and perhaps just introduce the new player the essential part that they must know to figure out the majority of the new gameplay concept so that they can still understand the overall game but still can keep learning more if they desire or not if they don't like it. I do understand that sometimes a breaking changes has to be made even if it just for fixing a bug and thus the mod needs to be updated again to stay its compatibility but making such a large change were they even have to change their own "gameplay concept" for it? I am against it. For my own part I rarely use the arithmetic combinatorics and other such structures that makes use of this green/red cables yet for me it is another part that I could bother learning it to take advance to improve my factory or chose to ignore it so for now I decided to ignore it. If the majority of new players still haven't figured out this arithmetic combinatorics what would be your reaction to it be? Removing too for the sake of simplicity and in turn abandon even more veteran players that take advance of this new feature? A rhetorical question. Though I wonder if "you" made this FFF blogpost intentionally so provocative as you did in the past with the belts/robots debate some times ago or not I don't know, this one is a minor issue at least for me. I think one of the aspect that makes Factorio so great that is has a lot of mechanics that the player can explore if they wish to use such as trains based factory only or a bot based factory only and as well giving him the option to use both or only one aspect of this feature, so far none of those features in this FFF has been touched obviously but I am also afraid that the developers will orient their future changes much more over the new players instead of older/veteran players which understand more of this stuff more or less.

"We have 8 different damage types in the game now:"

In the context of modding it gives a "universal" agreed damage types that a modder could balance their ammunition type with. Acid as it name implies is acidic in nature and corrodes soft and hard materials depending on its acidity level, so far only the aliens spitter classes use this type because they are the only ones which have access to this acidic "ammunition". So far I know there is no vanilla weapon/ammo that uses the acidic and I don't mind it because there is at least a option to expand it modding wise to make it more useful, I haven't seen a mod that introduces only new damage types and nothing else that can be used as a library for other mods to base it off from.

TL;DR if you desire this simplicity so much then at least do all the necessary changes in your own structures and keep the Modding API as it is so that at least mods can take advance of those features instead of axing it completely.

[Rythe]

And now let me answer the questions of 'Why pickaxe?' and 'How explain ingredient limit?' more directly, because I think I glossed over the direct questions for reasons when it appears I need to answer them.

The reason you have a pickaxe is because you have a character who interacts with the world using their tools (see weapons/grenades, armor mods/bots, vehicles). Direct mining is very big in the early part of the game, so it makes sense that you'd have a tool for it that can be upgraded like with weapons vs biters and vehicles vs terrain. Mining implement vs resource seems obvious here. There's a foundational thematic and game mechanic consistency that you've managed to lose sight of.

And I'm not sure why explaining the ingredient limit on assembling machines is actually a problem? I mean, are you explaining the why of ammo limits on weapons? Item stack limits? Ingredient limits are an inherent limiting feature to assembling machines that mostly just makes sense, because again, progression gating is a core mechanic (tech tree) and production specialization is too (chem plants, centrifuge). Assembling machine ingredient limits is a soft intro to product specialization and an early concrete representation of tech tree progression gating inherent to some buildings.

The ''just makes sense' is why I skipped down to what seems to be the real problem of Assembling Machine 1 being kind of crap and not meeting the needs of its tech tier. And why bobucles skipped down to the representation problem (that could be fixed by making an assembly platform area on assembling machines and just plop down the item sprite onto it while mechanical doodads whirl about in mock assembling motions). Mostly, I think you should just trust that the player base for Factorio is smart enough to figure out ingredient limits and maybe even the black box nature of these buildings. And if it's really bothering you, have Compilatron go "Yay! Assembling Machine 2 can accept more ingredients which means it can make more things!"

Really, why is this hard?

[Ohjay83]

Yet more quality of life. The game is going to be polished so much, its gonna have a hole in it

Lazy bastard is going to be a breeze now. I would consider lowering the craft count to reflect these changes.

Have you considered wiping the achievement stats on 1.0 launch?

That would devestate me. I did all the achievements.. and I am waiting and longing for many more interesting achievements that suggest you play a game/map with a whole other focus.

[glee8e]

Amidst this chaos of assembler tiering, I want to amend something about kovarex's word on mining category.

Also, with the hardness, the argument why to remove it completely was, that we still have the mechanism of "mining category".
So if a mod wants to lock mining of certain resource to a certain miner, he can just use a mining category that the basic drills not handle.

Assembler recipes also have such sort of category. It's called "recipe-category". Do you know internally chem reactors, refineries, centrifuges are also assemble machines? They just use different recipe-category. Recipe tiering can still be acheieved with this even without the slot limit, albeit a little tedious.

----

As a side note, I do believe that vanilla recipes need tiering too, instead of just laying that task off to modders, otherwise the vanilla progression line would be too greatly affected (more specifically, you can automate all science packs production without having to get to certain automation techs). About the puzzled new player? Well...

• Add a property "unable_categories = {{category="assembler-tier-2", reason="insufficient-tier-2"}}" to the assembly machine prototype that shows what can't be crafted due to tier limit.
• Color the tier 1 recipe background in grey, tier 2 in blue, tier 3 in gold, etc.

There are actually quite some ideas. However these wouldn't come to you dev's mind if all you want is to deliberately lower the barrier of entry, by all means.

[leighzer]

I agree that it is a good idea to streamline the base game. This will improve new player first impressions and help retain new players. Once someone is hooked and vanilla isn't complex enough, they can always download mods to introduce more complexity.

I was neutral about the assembler recipe limit removal at first. Yet, as long as the devs document the changes well and provide modders with reliable ways to create, incorporate, and maintain mods I see a nice compromise.

Seasoned players that are missing mechanics they grew to love can always mod them in, all the while devs can take vanilla where they wish. I do feel for modders as this equates to more time you have to invest to keep your mechanics. As long as devs invest in supporting the modding community/increase the ease of modding, I see this as the best balance. Without mods, it would be a choice between the dev and players vision of the game. Fortunately this isn't the case.

Keep up the great work! I say streamline confidently!(with modders in mind (I'm sure they are))

[qwerty_ca]

Generally speaking, I like the new changes. Simplification is always a good thing. I especially like the removal of the axe idea. I hope you guys can do the same for other things like power armor, personal roboport capacity, personal nuclear reactor capacity etc.

[bobucles]

In the context of modding it gives a "universal" agreed damage types that a modder could balance their ammunition type with. Acid as it name implies is acidic in nature and corrodes soft and hard materials depending on its acidity level, so far only the aliens spitter classes use this type because they are the only ones which have access to this acidic "ammunition".

The point of damage types for damage types sake is silly. There's no reason to make more flavors of damage just because they can exist. The point of a damage type is to implement some kind of desired game mechanic. The rock/paper/scissors math under the hood is calibrated to FIT the game design, not the other way around. So:
- I want this rocket launcher to be really good against giant armored things
- I want this flame thrower to be really good against wet squishy things
- I want small rapid fire weapons to be really bad against heavy armor
are examples of stipulating what the desired game balance is supposed to be. Stipulate it first, then you can look at various armor systems to match it. For example, flat damage reduction makes rapid fire weapons REALLY weak. That makes it a good choice for one goal. A damage multiplier helps a weapon do exceptional damage against certain tagged targets. That's really helpful if the tag is easy to understand such as doing a big damage bonus against "big things". There are countless dozens of ways to generate AoE damage, DoT damage, front loaded damage, back loaded damage, aggressive biased mechanics, defensive biased mechanics, so on and so forth. Blizzard games end up having incredibly complex relationships due to them using a sprinkle of every kind of mechanic imaginable.

I've seen mods that basically make a rainbow of every damage type possible. Factorio isn't a game where you match the colors of your bullets to the colors of the enemies to win. So a huge range of damage types doesn't make sense. What does make sense is setting up a set of general guidelines, and finding the correct damage types to fulfill those guidelines.

[wlfbck]

Reduce the number of damage types to something like: generic, impact, heat, and acid.

Gotta disagree here. I wouldn't know where to put laser or electric under there. Why not just merge generic and impact into physical and add energy/energetic for laser and electric?

(heat being fire and explosion is obvious, although i wouldn't have named it heat, but that doesn't matter. Same goes for poison and acid)

Found a nice article. Basically they're using easier to understand terms for mechanical (=potential+kinetic), thermal and chemical. Laser is clearly heat, i remember melting down enemy tanks in Earth 2150 ^^. If they're removing moddability for that there definetly needs to be direct (un-resistable) damage, but i'll just hope it stays moddable.
energy2.png

See i would have never made the connection between heat and laser, also take a look at the reddit thread for this FFF. There is quite a lot of people who aren't sure what these new types are exactly. I'm still not sure what impact is supposed to be, i guess bullets? or is it when you get hit by trains?

[eradicator]

I think a big problem here feeds into the the important early game experience of "show don't tell". Players have a finite patience and it is more fun to watch things happen than to read about what should be happening. Some aspects of Factorio are handled extremely well in this regard. Mines do mining, belts do belting, and inserters swing around and it's a lot of fun to watch it all in action. Simple devices show simple feedback, but assemblers aren't very simple. They contain a lot of important information such as recipe, resources and activity and most of that information is critical to the gameplay at all times. So before telling me what the key essential difference between these three assemblers is:
assemblers.jpg
Maybe it'd be more constructive to see how the default feedback on their function can be shown in a better way?

With the video engine getting super optimized I think it's a great time to discuss how maybe things could fit better into the game's "natural" view. That way things can both look and play well without having to bury the experience in overlays. Maybe I'm talking crazy, but an overlay should help enhance the experience rather than being the core essentials.

A very - very - good point. I sometimes switch off the overlays to have a look at the factory. But then i instantly notice that i can't see anything besides inserters and belts anymore. "Classic" building games like the Anno Series or Settlers have a seperate building for each "recipe", all of them with a cute little animations showing the farmers farm, the miller mill, or the smith smith. It's a pleasure watching the motion, and i deeply miss it for assemblers. (German even has a word for it: "Wuselfaktor", which literally translated would be something like "bustling-factor", because these types of games are really popular in Germany). Ofc a different building per recipe doesn't fit factorio, but i wouldn't say no if we got some specialized high-end buildings like a "circuit board assembler", or some new animations for the existing ones. Chemplants already have a different animation color for each recipe, but many people probably don't even notice because it's such a small difference. Just showing a small stack of the item being made lieing somewhere around the machine would be an awesome improvement (think wood next to a lumberjack's hut).

[ownlyme]

[...]Ofc a different building per recipe doesn't fit factorio[...]

It would be possible to make the machine's appearance change depending on the recipe chosen.
Of course that would be a ton of work for the graphical artists and make the job for modders more difficult because they wouldn't want to add generic machines into a factory that is so lovely detailed.
I think a viable solution would be changing the appearance depending on the kind of recipe what's chosen.
For example all chip-assemblers would look the same, all assemblers making iron gears, iron sticks, copper cables... would look the same etc.

[HoGo80]

No, please dont do it... Dont start this thing. Making the game more simple, easier for the dumbs. It was a reason I stopped to play WoW.
There are steps you have to follow, there are rules you have to keep in head and you will get a bigger feeling of success / achievemt later. Not cheap wins right from the beginning.
I know, you will get more players, more money, yet you just will make it an another easy game for easy-lazy people. More people will buy, but more people will leave it right after few days.
It was A game for me, but you are turning to wrong direction.

[Luaan]

If the motivation for changes is to make entry level simpler, I get it. Just look at the Steam achievement percentages to see the low proportion of people who stick with the game.

Do you really think that 55% people getting all the way to oil processing is a bad number? What games are you comparing to? Also, keep in mind that you don't get steam achievements when playing with mods. I've spread this game directly to about ten people so far, and most of them didn't play more than a few hours of vanilla - but they're still playing with mods to this day. And I wonder how the number of people finishing the game dropped with the massive complexity of the research changes.

Clean up the GUI? Sure. Improve some ratios? Sure, why not. But while "remove all the unnecessary complexity" is a good design approach, the "unnecessary" part is quite critical here. I've always supported pretty much all the changes Wube made to the game, and I very much like their approach at eliminating things that are superfluous. Boiler efficiency is an important game mechanic - if anything, I'd love to see burners expanded; right now, most people just have everything hooked up to one power grid. Pickax and armour durability? Pretty much useless. Things that you only make once (unless you lose them)? Well, the game has a great hand-crafting system - making a one-off thing is just a click of a button; if you want to make that more straightforward, reduce the crafting times on things that you don't really need to automate.

Ingredient limit doesn't make sense? Sure, it doesn't. But why not make the assemblers distinct in a different way? Organise products by complexity. Remove the ingredient limit, but not the progression - why not just say e.g. "engine can only be built in L2+, electric engine in L3+...". Sure, it's going to need some balancing, but so are your changes. Heck, why even have three tiers of assemblers if they're just recoloured variants with higher speed and module count? Why not have assembler research improve speed and module slots? Do assembler tiers really add that much to the game? Half the people use the cheapest version for everything they can, the other half always uses the latest they have available. Why clutter the inventories and add extra production steps for no reason? Consider how assemblers vs. chemical plants work - they're both "just" assemblers, but a lot more distinct (and visual). Bob and Angel further improve this with things like electronics assemblers - it makes the game more interesting without really making it any more complicated (I mean, the manufacturing facilities; Angel's mods are of course fiendishly complicated ). Of course, it's a lot easier to just cut, rather than trying to figure out more interesting things to do with things that are kind of marginal, and I understand that you're trying to get to 1.0 with as few new features as possible - Factorio has been in development for a very long time, and getting it to 1.0 would surely bring a great sense of finally finishing a game. But software is never finished - and few games have paid so much service to that idea while also being awesome as Factorio.

Improving the new player experience is a great thing. But for all of the changes you're making here, somebody already suggested a great solution that fixes the problem without just cutting the feature out, and making it more interesting. And we're not talking about some great development challenges here. If it's useless, cut it out. If it's interesting but underused, think for at least 5 minutes about how you could rework it to be better. No such nonsense should ever block a new player, of course - but having it as extra depth to be explored as you're getting comfortable with the game is something completely different.

[Venatos]

these are just beautifull simplifications! thank you so much!

""Reduce the number of damage types to something like: generic, impact, heat, and acid.""
generic and impact can savely be merged into "kinetic", people know and trust it.

[bobucles]

Half the people use the cheapest version for everything they can, the other half always uses the latest they have available.

That has never been my online experience. The vast majority use almost pure Tier 2 assemblers for the majority of pre-rocket because It Just Works. The main time that T1s exist is in the super early burner stages or if someone simply forgot to upgrade an early game build. No one uses the "cheapest option possible" because
A) that's not how to play Factorio. the goal is always bigger, faster, moar, and
B) After realizing T1 assemblers can't build a frickin' entry game inserter you realize they just suck.

kinetic

Also known as ordinary "physical" damage for most.

[dood]

Good.
You might also want to remove armor durability, it doesn't really serve any purpose since this isn't a don't starve type game in which your breaking equipment forces you to constantly go out and scavange for resources in order to fill declining bars.
The final modular armor you'll be sticking with has so much durability that it is hardly a game mechanic anyway and I don't see how lowering durability so it becomes a problem leads to fun.

[Jap2.0]

Good.
You might also want to remove armor durability, it doesn't really serve any purpose since this isn't a don't starve type game in which your breaking equipment forces you to constantly go out and scavange for resources in order to fill declining bars.
The final modular armor you'll be sticking with has so much durability that it is hardly a game mechanic anyway and I don't see how lowering durability so it becomes a problem leads to fun.

They are removing it.

[planetfall]

My two cents ashtray full of spare change:

-Removal of things from the game engine itself, such that mods can no longer do those things (with a rationale of effectively, "you shouldn't want to do those things"), is almost inherently a bad decision.

Deleting mining hardness outright seems... harmful? Instead you could just have it equal across the board, and have some kind of global flag for whether to display mining hardness/power stats. It defaults to false but can be changed by mods. The only way I can see this specific change being justifiable as an engine-level removal is if the extra computation was making miners use excessive processor power, which I wouldn't assume to be the case but I'm not sure where to find that sort of data.

-I like the idea, posted somewhat upthread, of making the base mining speed be equivalent to the current iron axe and have the new iron axe be an upgraded version. I think it's somewhat silly to remove axes entirely. Perhaps it would be better to have mining tools have infinite durability but give them a much higher resource cost, bringing them in line with the model used for armour.
One of the things that really bothers me is how fast you reach the top tier of mining tool (literally within the first tier of research), and wouldn't be opposed to 1-2 more tiers (laser cutter, AI multitool?) later on that require more complex components. Again, making them like armour, which I personally draw a parallel with because of the relative position of their UI slots.

Another point is that despite defence being a fairly negligible resource drain once you set automation up, there is no voice calling to remove ammunition from the game. Maybe the solution is to make construction bots require an axe in the recipe, so that there is a reason to automate production. (I do still prefer the infinite durability model, though.)

-People have already beat to death the fact that assembler input limits are a crucial element of progression and the real problem here is how to communicate rules to the player. Despite that, I do think that for the base game at least, it would be better to base recipe progression on crafting categories rather than input limit. Though obviously leaving the functionality available to modders would be ideal.

It seems exceedingly silly, to my mortal mind, that a T1 assembler can make the purportedly complex brain of a stack filter inserter, or high tech items like logistic bots, level 1 modules, and even ROCKET CONTROL UNITS, but not the simple inserter that your character already knew how to make before inventing the assembling machine.

A problem related to this is that assembler 2 is unlocked too soon after assembler 1. Not only that but the recipe for assembler 2 is just an assembler 1 + the same intermediates again. So there is no point, ever, to simply automating assembler 1, because by the time you have the ability to do so you automatically have the right materials on-site to make assembler 2. And after that astonishingly short gap, assembler 3 is unlocked fairly late.

There's also the problem of chemical plants and oil refineries being significantly better than assembler 2 when unlocked (in terms of crafting speed and module slots) and then never getting an upgrade ever, which makes machine-progression-as-a-through-line feel a bit patchy, but that's at best tangential to the topic at hand.

Some changes in this department that I think would make the progression better include:
-In addition to most things that take 2 ingredients, assembler 1 can make basic and long-handed inserters, base level splitters, and (for the sake of Lazy Bastard) labs, assembler 1 and burner miners.
-Assembler 1 CANNOT make any recipe that uses steel, red circuits, processing units, or either type of engine units. Also, no combinators, advanced inserters, or red belts/undergrounds, probably?
-Move assembler 2 to green science; change its recipe to require different components (maybe assembler 1 + steel + circuits + filter inserter? Spitballing here.)
-Perhaps make assembler 1 more energy efficient than 2 (to represent the additional electronics required for greater complexity) such that when to upgrade is more of a strategic decision.
-Assembler 3 is needed to automate rocket related stuffs, level 3 modules, blue belts, uranium bullets/shells and nukes. Or whatever else is deemed "late game enough."
-Have the "made in" thimmajig show on all recipes by default