mumbri.ch

I recently played The Planet Crafter on a recomendation that it was similar to games that I greatly enjoy, namely Subnautica and Factorio. As you can tell from the title, I did not enjoy The Planet Crafter very much, but I think it is ‘bad’ in an interesting way. Moreover, I don’t think it is far from a good game, from an engineering perspective. I.e., I believe that it would require relatively little technical effort to rearrange the game into a much better experience. Who knows, I might even make a mod for it out of spite.

What is The Planet Crafter

The game asks you to turn this:

Into this:

This is accomplished by building drills to vaporize rock into an atmosphere, heaters to raise the temperature, hydroponics pods to generate oxygen, and eventually plant and animal nurseries to seed the ecosystem. The interface and mechanics are taken directly from Subnautica, down to the magic construction wand and aggressively constrained inventory slots.

Where the game diverges significantly is in the quantity of items you must gather to accomplish your mission. Whereas Subnautica’s Cyclops submarine (the endgame vehicle option) requires 11 items — of which only 4 are compound items requiring multiple steps, totaling roughly 50 raw resources — an endgame heater requires 9 compound items requiring 9 items each, 3 of which in turn require six items each for a total of 183 raw resources. Add to this the fact that you should be constructing dozens of heaters, in contrast to the single Cyclops you build (provided you keep it safe from the sea monsters!).

In this aspect, the game is closer to Factorio, where you try to consume and process exponential quantities of raw materials to achieve an ultimate goal. Coincidentally, in both cases you greatly modify the local ecosystem in order to escape on a rocket (although in Factorio, you are tasked with habitat destruction rather than creation).

What this critique is not

I don’t want to talk about the quality, polish, or realism of the game. For example, I am not particularly bothered that the planet is considered terraformed at a temperature of (roughly) 1 kelvin, which is colder than the cosmic microwave background, never mind below the freezing point of liquid water. Or that a small space heater is conceivably going to move the needle on the thermodynamics of a planet. It’s a game after all, and some suspension of disbelief is required to engage with it. Not to mention, the dev team consists of two people, and is quite reasonably priced at 20 Euros. This is a very impressive game in terms of execution.

What I’d like to talk about instead is how this game completely misunderstands the design choices that they took from their inspirations.

Why is it so hard to craft in Subnautica?

Subnautica is above all a horror game. The core gameplay loop is to explore dangerous areas for blueprints and resources, before returning to safety to build a fortress of solitude. Building better gear allows for exploration to a greater depth and to go for longer trips, gradually expanding the players reach mechanically, as they map out the space in their mind.

The point of the gameplay loop is to force the player into stressful and novel (but wonderous) situations before bringing them back into a calm, familiar environment. Thus the gameplay loop must be both short and should not retread the same areas, otherwise the player will either become overstressed or too familiar respectively. Subnautica accomplishes this both with the oxygen, water, and food meters, forcing the player to return to the surface or home, along with its crafting system. While the meters are relatively self-explanatory, the crafting system deserves some elaboration.

Given that the goal is for many short trips into new areas, the crafting system would undercut the intended experience if all the necessary resources for the game could be gathered in a small number of long trips to the same area. Therefore, the inventory space for Subnautica is aggressively limited, to prevent hoarding. As a consequence, most of the buildings and items require few resources and are required in small number. In addition, there is a large number of primary resources:

Many items are only available in certain biomes, forcing exploration into new areas. And since they are required in small numbers, the necessary quantities can be gathered in one or two trips, preventing them from becoming boring.

Subnautica’s Vehicles

The vehicles of Subnautica are quite important. Each augments the gameplay loop in a different way.

Seaglide/Seamoth

The seaglide is a diving scooter, while the seamoth is a mini-sub. Both speed up travel times, expanding the range for a single trip without extending the time of an exploration loop. The seamoth additionally provides oxygen, extending the depth of trips significantly, and allowing for less risky exploration of wrecks.

Prawn Suit

The prawn suit is a deep-sea exosuit that essentially trades oxygen management for freedom of movement. It can’t swim, but it has a grappling hook and an infinite supply of oxygen. It also provides a degree of psychological protection, as the suit’s armor acts as a second health bar. It therefore permits longer trips into deeper, more dangerous areas.

Cyclops

The cyclops is the midgame goal, and acts as a mobile base. It permits the gameplay loop to be undertaken from new locations, avoiding the issue of navigating the same path repeatedly. Crafters can even be built inside the Cyclops, completely obviating the need to return ‘Home’ as a matter of routine. This gets around the limitations placed on the game by its design choices, once it has reached the point where the drawbacks outweigh the benefits.

How Factorio Works

Factorio, in stark contrast, is all about consuming exponentially increasing quantities of raw materials and converting them to finished products. Rather than exploration, this is accomplished through the engineering of processes to feed inputs and draw outputs from crafting machines. Great care was taken in the design of the interface to allow the scaling of production from 100-1000x to be roughly as easy (or difficult) as the scaling from 1-10x. This is principally done with the copy/pasting of building arrays and robot constructors. The biggest challenge of the game is around managing the evolving logistics and complexity of the factory as a whole.

Manual gathering and crafting is explicitly discouraged. Manual mining very slow and unscalable, while midgame items are gated behind forced automation recipes, most notably the engine unit which cannot be hand crafted.

Meanwhile, the automation tools at your disposal are varied and powerful. Trains can carry thousands of items at high speed across large distances, and can be fully automated and networked. Everything can be routed by transport belts. Inserters automate the feeding of inputs and outputs. Robots can fly items point-to-point. Et cetera.

How The Planet Crafter screws it up

The Planet Crafter tries to combine the exploration and wonder of Subnautica with the scaling of Factorio, and fails quite badly. I think that this is a very difficult task, so I won’t denigrate the dev team for trying. Surprisingly, however, I think that they have most of the ingredients for success.

The real charm of The Planet Crafter is watching the barren rock you’re on come to life in front of you. The team at Miju Games have absolutely nailed the beauty of this transformation. The problem is that every gameplay system that they put on top of it conflicts with this core thesis.

Exploration

There are two distinct exploration phases to The Planet Crafter, before an oxygenated atmosphere, and after. Before, the exploration radius is hard limited by the seconds of oxygen in your tank (plus the extra oxygen bottles you bring with you), which can only be increased to approximately 5 minutes by midgame and 8 minutes shortly before oxygen stops being an issue. This ends up being the principal reason to return to base, which can feel frustrating if you have not filled your meager inventory.

Unfortunately, the rover vehicle does not provide oxygen, so does not serve to extend your radius of exploration in practice. It does expand your inventory somewhat, so it can be used to some effect once the atmosphere is oxygenated. Once this happens, the principal limit to exploration is inventory size. Since the crafting system is lifted wholesale from Subnautica, this is a very hard limit which comes close to the 5 to 10 minute limit previously imposed by oxygen.

The problem is, there is no need to shorten the gameplay loop! Since the game is not terrifying (no monsters), the player is not under stress when exploring and therefore does not need a periodic reprieve. The limited inventory is an unnecessary obstacle that only serves to interrupt the flow of exploration. I would go so far as to say that the oxygen meter itself is antithetical to the core game experience.

Given that the point is to view the change in the world, it is important that the player becomes familiar with the world at all stages of progress. This is the opposite of the experience objective of Subnautica’s gameplay loop. By starting with a small exploration zone, the player can only see the change of a tiny fraction of the game’s world. Thus the point of exploration in the context of The Planet Crafter is undercut by its mechanics.

Gathering and Crafting

As alluded to previously, the aggressively small inventory is detrimental to the game’s thesis. It shortens the gameplay loop when it should be extending it. But the problem is actually much worse.

Because the mechanics of the game rely on exponential growth, massive quantities of raw resources are required. Whereas Subnautica restricted its recipes to using trivial quantities of inputs, The Planet Crafter demands increasingly expensive infrastructure to progress.

Automated mines do exist, but are available frustratingly late in the game. Furthermore, they mix in common ores, complicating the gathering process. On its own, this could be an interesting complication, but with the very small inventory spaces and the lack of mid-game automation, it is needlessly frustrating.

Transporting these raw resources to your main base is similarly taxing. Due to the player’s small inventory space, junk resource outputs, and distance to the mining sites, much of the midgame revolves around going to and from miners to fetch resources. Logistic robots are available quite late in the game, and are rather expensive to boot. They resolve the transport issue to an extent, but have relatively poor throughput. Fortunately, they are sufficient for the game’s crafting requirements.

Automation

The Planet Crafter requires sufficient crafting that automation is necessary, certainly if the scaling is increased or the issues with gathering resolved. Automation tools exist, but are so late in the tech tree that the game is essentially over by that point. The autocrafter and logistics bot network are extremely expensive, and would be uncraftable early game, even if they were available.

They are also too powerful. Logistics bots move items from anywhere across the map to requester chests anywhere in the world. Autocrafters grab inputs from anywhere in a large radius. Production chains are as simple as putting a dozen autocrafters and requester chests within a radius and configuring them. But, each chest can only request a single item, otherwise it will get clogged due to the small inventory.

Scaling

Scaling is unsatisfying in The Planet Crafter. Tech levels separate machines by roughly an order of magnitude each, and use mostly the same ingredients. Therefore, it is best when reaching an unlock to scrap your existing machine array and build the new machines for a 10x improvement. These can then be scaled by multipliers, either from satellites (somehow), “optimizers” (beacons from Factorio), or species bonuses for oxygen and biomass generators.

There isn’t much depth to this system. There are no complex interactions, no bottlenecks, no real difficulty. If the planetary outputs become unbalanced, it often makes sense to focus on the overproducing elements because the multipliers will stack faster than on the under-producing elements. In fact, the only thing tying the different outputs together (aside from some secondary production from certain buildings) is the tech tree, where a next-level heater is gated behind a pressure milestone.

This is a far cry from Factorio, where (for example) a green chip fab is bottlenecked by a shared resource input, and while a second array can be copy-pasted, a new logistics method is required, etc., etc…

How to Fix Everything

I believe The Planet Crafter is frustratingly close to greatness. Here are my recommendations for fixing the game, using only existing systems:

• Allow automation and logistics much earlier, and make them much cheaper. This should be the primary means of gathering resources.
  • All chests should be suppliers by default!
• Soften oxygen requirements, or get rid of them entirely. The Rover should provide air, at a minimum. The player should be familiar with the entire map before liquid water arrives.
• Expand the inventory system by a factor of 10 or 100. Allow for item stacking. Full inventory is not a fun reason to go back home, and shouldn’t happen in an automation game.
• Mining contamination is only interesting if the automation, logistics, and storage issues are significantly streamlined. It should be scrapped until those elements are up to an acceptable standard.

The above changes should be more than enough to significantly change the feel of the game for the better, and they require no new assets. If new systems and assets are on the menu, however, the following changes will probably make a positive impact:

• Trains! Trains are always a good thing.
• More developed crafting machines and inputs. Transport belts into crafters, like in Satisfactory, would be both more challenging and rewarding than the current system of teleporting inputs and clogging outputs.
• Copy-pasting building arrays. Probably quite technically challenging, but it would be cool.
• More ‘map magnets’, i.e. context clues to investigate particular elements in the world, ideally with a deadline. Miju made wonderful scenes go from barren to full of life, but the player will never see the transformation of most of them.

Conclusion

I unfortunately cannot recommend The Planet Crafter, and I somewhat resent the people who recommended it to me. The game is, however, frustratingly close to greatness, and could probably be fixed with a single mod. If that happens, I’d be excited to recommend it to everyone who enjoyed Subnautica and Factorio (i.e. everyone with taste).