After some discussion about my previous post, Factorio's Recipes - Design Values, I decided to gather a few more thoughts about Factorio's recipes. I mostly want to zoom in on the tier system I described through the lens of the science packs.
Screencaps from https://wiki.factorio.com/
The most obvious tier structure for Factorio recipes is the gating provided by the various types of science packs. Each of the packs requires the player to prove that they've mastered some aspect(s) of the game in order to progress. As with any aspect of the game, the mastery can often take the form of handcrafting a small quantity of initial packs to get the first few technologies. And that's fine for the purposes of gating access to that tier of technology. Initial technologies often help with solving some of the challenges with scaling up production, so allowing them to be accessed more easily helps players appreciate the new power they're gaining.
Red science requires the player to demonstrate three key skills. The first is producing iron and copper plates. As the foundation of almost every recipe in the game, it's fitting that this is the first challenge. Second is converting those iron plates into gears. Third, which isn't listed in the recipe, is the production of electricity. This one doesn't count as part of the recipe, but it is worth calling out. While the first packs require that the gears and the packs themselves be handcrafted, the player will soon be able to automate this production with access to assembly machines. And they really need to do so, as the quantity of packs required for future technologies ramps up dramatically. Thus the player is forcibly exposed to the ability to handcraft science, but then quickly shown the folly of continuing down that path. The most basic automated setup for red science involves an assembly machine producing gears, direct inserting them into the machine producing the packs. This is pretty wasteful of the gear machine's time (0.5 seconds active for 5 seconds of pack production), but a beginning player would be unlikely to notice/care.
Green science's biggest difference from red science is the introduction of multiple layers of intermediate products. Both inserters and conveyors use gears, which the player has already shown mastery of. But they also need to begin producing green chips, which in turn require copper wire. Thus, at its deepest point, the green science pack is 5 post-ore layers deep (copper plate -> copper wire -> green chip -> inserter -> green pack) to red science's 3. The player could continue producing these products via direct insertion, but there's 2 factors that are likely to drive them away from that idea, and towards belted intermediates. First is that gears are now needed for 3 different elements of the science production process.
Second, excitingly, the intermediate products of green science are useful in their own right. If the player produces a belt and/or box of conveyor belts and inserters, they can swing by periodically to top off their own supply for base-building purposes. By forcing players to mass produce useful tools for general purpose factory production, the green science recipe makes sure players are well stocked to develop into the future. Side note: This is one of my strongest critiques of the otherwise fantastic Seablock set of mods - most or all of the science requires bespoke intermediates that have no other use in the game.
Two things stand out to me about military science. The first is that it's completely optional. No part of the tech tree all the way to launching the rocket requires them. The second is that its recipe requires you to demonstrate mastery of three different core military technologies. It doesn't feature turrets. But I would argue that turrets, walls, ammo, and grenades make up a very strong core for early to mid game defense, and this requires 3 of the 4.
Blue science tests, to my eye, three key skills. The first is producing a high quantity of green circuits. With each 2-pack requiring 6 green circuits, it's definitely time to start seriously considering how to mass produce this key ingredient. This is the first time that the true scale of the green circuit problem begins to show itself. It's not a hard recipe, but it's a hungry one. The second is steel production, as measured through engine production. As discussed in my last post, steel is an interesting problem in that it requires re-processing of iron plates. So it's fitting that the steel test here also requires iron, too. Thus the player must demonstrate that they have iron and steel coexisting somewhere in their factory. Finally, crucially, the player needs to have some amount of oil production figured out. As mentioned in https://factorio.com/blog/post/fff-305, this has received significant recent attention to ease the difficulty. This is particularly interesting in its interaction with the Advanced Oil Processing research. This used to be an absolute requirement to rush, as the ratios from basic processing were suboptimal for producing the solid fuel required for blue science. With the changes, a player can keep producing petroleum gas for sulfur and plastic production for as long as they'd like before moving on to producing heavy/light oil.
Purple and Yellow Science
Purple and yellow science are interesting in that they're not tiered relative to each other. They each allow the player to invest in different areas of late game improvement, both in terms of what they must prove, and what they get to unlock once they do.
Purple science focuses on raw production capacity. Of the three module types (with their identical requirements), it's fitting that productivity modules are the ones that are needed for purple science. They're the best at squeezing every last drop out of your factory, in particular due to their multiplicative effect when stacked through every stage of production. Electric furnaces and railways are both extremely useful for expanding the footprint of the base's resource input. The former to make an ore patch with no nearby fuel usable, and the latter to bring the resources back. Once the player has unlocked this tier of technology, they are largely treated to higher speed and efficiency versions of things they already have.
Yellow science, in turn, is focused on two highly disruptive areas of technology. The first is the game-changing possibilities of the logistic network. The very basics of robotics can be (and must be) done prior to accessing yellow science. But in order to access requester chests, as well as speed up robots to useful speeds, players must develop yellow science. The second is dramatically improved power armor, which can change everything from construction to fighting to simply the act of walking around.
The final science challenge functions as a post-game challenge. Thus, a rocket's purpose as the end game trigger takes center stage. The set of recipes involved here (the silo itself, the rocket parts, and the satellite parts) run the gamut of input requirements across all tiers of development. This is fascinating, because it breaks a lot of soft rules about how the recipe tree is laid out, where inputs are typically from similar tiers of development. But of course it should; it's the end of the game. Like the boss rushes in Mega Man games, constructing and launching a rocket with a satellite requires the player to produce a greatest hits album of their previous technologies. In rough tier order: iron pipes, radar, steel, concrete, solar panels, accumulators, rocket fuel, electric engines, blue chips, low density structure, and blue modules.
Thank you lauzbot, Archimageg, and dermanus for discussion that prompted this direction of investigation. And thank you n_slash_a for opening my eyes to the fascinating cross-tier inputs to white science.
Come join the discussion on Reddit!