Friday Facts #376 - Research and Technology

Posted by kovarex, V453000, Klonan on 2023-09-15

Hello,
This week we will talk about some technology and research related changes coming in 2.0 and the Space Age expansion.


Quality clarificationskovarex

It was rather interesting to read the reactions to the last FFF-375 about quality.
It sparked a lot of debate and we noticed a lot of misconceptions about it. So we would like to address some of the biggest ones:

  • When you don't put quality modules somewhere, everything works as before, so you don't have to sort everything everywhere as some people thought.
  • The random aspect of the mechanic seems to be very upsetting to a lot of people.
    I have to agree, that there are a lot of situations where I hate the randomness in games, but the context really matters greatly.
    For example, I decided to never ever play Settlers of Catan again, as the random dice can decide the winner of the game way too often for my liking.
    But the Factorio case is very different. The reason is, that you rarely depend on the random result, because you (hopefully) produce at a large scale and the law of large numbers will just transform probabilities into ratios. The more you produce the more it evens out.
  • The core of what makes Factorio good is that the horrible huge grind you would need to get all the resources can be mitigated by automation.
    It is quite similar with quality, as the annoyance of the grind and unpredictability related to randomness in other games can be just pummeled by the sheer output of your Factory.
  • The most simple approach is to produce all the ingredients in normal quality.
    So most of the Factory (typically) produces non-quality stuff and mainly provides for research (non-quality) and the huge amount of items to be processed by the small quality part.
    In our playthroughs, the parts of the factory which were dealing with the quality were less than 5% of the whole Factory, so don't worry about being forced to solve all the mixed quality items everywhere.
  • The names are the easiest thing to change. Maybe I don't take myself and the game too seriously and found it amusing, but if we had a very good counter-proposal which feels good and is clear when it comes to tiers, we can still change it.
  • The fact that there isn't just "the one" blueprint for circuits, etc., because it depends on the stage of the game and your dedication to quality is something we see as an improvement. Either you have more blueprints for different stages of the game, or you have to improvise more. And mainly, if you actually shoot for the best setups, you can still have the same approach as with beacons, that you build for a legendary setup and just accept that it is not perfect now and upgrade it as you get more of the good items, very similar with beacon setups before full tier-3 modules are available.
  • I used the "it's optional" argument, but I didn't mean it as an excuse for "It's shitty but it's ok, because you don't have to use it".
    I understand that if the mechanics provide an upgrade, you can't say it is optional, as our gamer-optimizer brains are just programmed to progress to better stuff. And it is painful to decide to ignore some better stuff just because I don't like the mechanics to get there. This leads to the biggest proverb we always repeat: "The best strategy needs to be fun".
    The main part of the optionality was that you decide optionally where and when you want to use it.
    There are still players who prefer to play without beacons and modules, so some small value in the optionality for these exist as well.
  • The last very frequent opinion was, that it would be better use this opportunity to introduce a lot of interesting custom recipes to get all the qualities of different items.
    But avoiding this was the cornerstone of the design. I can only design the game around what I enjoy, and I just don't enjoy having to keep track of a huge number of unique recipes, which all work technically basically the same.
    The fact, that the most simple approach (mostly) means using similar blueprint to produce different kind of quality items is generally a goal, as it helps to downscale the time investment needed to progress, since this is just one of the expansion mechanics, and we want to keep the game time reasonable.
    I would compare it to repeating the design of rail intersections, no one is complaining, that they can repeat their blueprint in many places, because we have blueprints and construction robots to avoid the grind related to building repeating patterns.
    In other words, there is a relatively straightforward way to go through it and get the best items, but the design still allows for some very non-trivial custom setups if you want to optimize more and enjoy the theory-crafting. Easy to learn, hard to master.

Overall, the design was driven by our own enjoyment. We play the game and we don't want to bloat it with grindy clutter.
I can't expect everyone to enjoy games the same way I do, it is okay and I respect that. But I believe there will be enough players who will enjoy the ride with us.

This is a simple build to create Assembling machine 3 in quality.
If there's more than 100 of any quality, it recycles them hoping to get the next tier.
The build only receives the basic ingredients from "the normal factory", and all of the quality complications remain here.
In earlier stages you'll likely have just a few of these kinds of builds for items you prioritize the most, however with time you will likely add more of these.
...with which some more upcoming features might help you. ;)


Discovering new materialsV453000

When it comes to how planets and especially their unique resources are discovered, we have realized problems with how Factorio has been doing this for a while now.

If a planet is unlocked by a technology on some science tier, then the starting resources of the planet would also be gated behind these same science packs - because what else, you don't have anything else available yet. This would mean that you could (and most typically would) already be researching ways to process the new unique resource that you have not even seen yet.

Another most fundamental conflict was that we wanted the progress on each planet to result in a single new science pack, resulting in a chicken-or-egg situation. The problem would be, this science pack should need most of the new unique processing steps on that planet, therefore you need to somehow get to discovering the processing first.

The alternative would have been that the science pack is made available immediately on arrival, but then it couldn't really be too interesting as it wouldn't include any of the processing you have discovered past that point.
The science pack is the primary export for each planet, so in order to prevent planets being just simple mining outposts with a rocket silo, we wanted to make sure most of the processing contributes to produce such science pack.

We have tried to design and implement several approaches in various forms of temporary science packs and even special entities that would act as temporary laboratories, but all of them were just adding mess that wasn't useful afterwards, as it was just a stepping stone to get back to producing a proper science pack that you would eventually ship back to laboratories on your home planet.

Trigger technologies

The solution that ticks all the boxes is in the end incredibly simple - we added a new method of researching a technology. Instead of processing science packs in a laboratory, technologies can now be completed with a trigger.
Currently we have a few technology triggers that unlock research:

  • Mining an entity
  • Crafting an item/fluid
  • Launching a rocket with a certain item

This means that each time you land on a planet, you discover and exploit its new resources in a way that feels quite natural, as everything happens while you're actually there.

This will make a lot more sense with specific examples once we show you the actual planets and not just this technical background, but there's more...

We can apply this elsewhere

As we're sure you remember, there are times even in the base game where you do discover new resources - specifically oil and uranium processing. Just like researching processing for a new resource on a different planet, it's similarly strange for the same reasons that you can research deep into oil refining tech tree while you haven't even seen a crude oil patch yet, or unlocking nuclear reactors without having met uranium.

Now in order to really get into oil processing, you will need to have touched some crude oil. This helps prevent the situation where especially in the "green science" tier, a lot of players would research very far ahead from what their factory can currently process, and then later feel discouraged seeing all the piled up recipes to work through.

Early game

There's more coming from this - when you crash land on the very first planet! Why does the player immediately know how to craft steam engines, inserters, transport belts? It always feels much better in games when your progression starts as low as possible and you can earn all of the things in the process, which makes it all feel much more deserved in the end.

With the triggers, we could create new technologies for even basic things like pipes, early power generation, labs and the Automation science pack.

Some unlocks, like electric mining drill, radar or repairs packs don't need to be available at the very beginning, so in turn we could add them into their own new "red science" technologies.


Early game technology graph. You can also see for example Steel axe technology now requires you to craft actual steel plates, so you can't just get the steel axe upgrade without having made steel, which has always been weird since steel axe is no longer a real item.

As a result, the following will be all of your starting recipes. Getting back to having unlocked all the recipes you are used to at the start doesn't really take any extra time than previously - the early trigger technologies only require a minimal amount of items to be crafted. But it still contributes to making the progression feel better.

As will become apparent with more things in future FFFs, during the development of the expansion we have arrived to various topics just like this one which stretched beyond the acceptable weirdness threshold, and had to address it in Space Age. The improvements also impacted the base game in a good way - and we are certain that mods will make good use of this as well.


Research queue always onKlonan

Research queue was a feature that was long requested, and after initially adding it, some playtesting of the game with the feature enabled led us to feel that it had some significant drawbacks (FFF-254). After some more playtesting and community feedback, we settled on a compromise whereby the research queue can be enabled with a map setting when starting the game (FFF-255).

This was okay, but it came with its own problems. A major source of frustration was when you would start the game, play for a while, only to realize after a few hours that you (or your friend who started the server) forgot to enable the research queue. There was no way to enable the queue after starting without using console commands, which would disable achievements.

During 1.1 development, we wanted to address the problem of new recipe discoverability a little bit better, so we added the recipe notifications (FFF-363). The recipe notifications meant that one of the initial negatives of the research queue (for new players and discovering items) was no longer present.

Over the years there have also been a fair share of bugs related to the research queue that we have had to fix, and having 2 code paths and GUI layouts in all the places (Queue vs No Queue) was becoming annoying. In the end we decided to just completely embrace the research queue, and remove all the code related to not having it enabled. So now it is always on and you cannot turn it off, there is no off. There is no longer the strange 'After you launch the rocket' condition for the queue, and no longer will you ever forgot to enable it before starting the game.


Productivity researchesKlonan

We wanted to add some more infinite technologies, and also reduce the resource pressure as the game goes along. The mining productivity is great, and it works very well, but it doesn't give the player much choice in the end-game. The other infinite technologies are more specialized, lots of combat ones, but other than maybe worker robot speed, not much for your production.

So we added a new type of productivity research, the recipe productivity research. Each level will increase the 'built-in' productivity of certain recipes, such as steel, processing units, rocket control units, etc.

We also wanted to let you unlock something super powerful for the end-game, which led us to the idea of the research productivity. Each level will increase the built-in productivity of your labs, and as you might know, that is the most effective place for your productivity modules already. It will require the end-game resources, of which details will come in some future FFFs.


As always, let us know what you think at the usual places.