Multicellular Stage: Status & Ideas?

I’m guessing that may be inspired by the fact that sponges (the simplest known multicellular animals) have the ability to reassemble their colonies from a loose collection of free-swimming cells, and sponges have long been regarded as the most primitive of animals (Parazoa). However, first of all, only cells from an original colony (starting from a single founder cell) will reassemble and not coopt random other cells of the same species, afaik. More importantly, modern insights suggest that this may be a secondary adaptation in the sponge lineage and that Ctenophores may be the earliest offshoot, and not sponges. So the simplest animals were perhaps a blastula-like vesicle akin the placozoa.

The process of “budding” as in sticking together right after mitosis is now seen as the path to multicellularity. Spherical choanoflagellates like Sphaeroeca volvox are models of these primordial animals, although other forms also exist like e.g. strings. mats or branches. In fact, it’s apparently very easy to induce multicellular forms in normally unicellular eukaryotes like e.g. yeast, which can be made to form “snowflakes”:

Nevertheless, I’m all for “Why not both?”, so -as suggested- we could have:

  1. The acquisition of “binding factors” enabling catching foreign cells.
  2. This gives the ability to catch foreign cells for extra protection/force
  3. The acquisition of e.g. “junction factors” then opens the door for multicellularity
  4. This will result in an option popping up to add cells to their design

The details of the last step is subject to further discussion.


OK, to adjust and nuance what I wrote earlier: Cellular slime molds like Dictyostelids are able to attain transitional multicellularity by loose cells aggregating into units that can form crawling slug-like forms and stationary fruiting bodies. And that does correspond with the original gameplay scenario of loose cells coming together. Also, this is mediated by cell-to-cell adhesion proteins, i.e. “binding factors”. Still, true animal-style multicellularity departs from dividing cells staying together.

But, yeah, let’s have both! :slight_smile:

OK, it may be informative with the following highlights from the article:

A. “Multicellular organisms typically develop in one of two ways, either through division without cell separation or through cell aggregation. The first mode of multicellular development is exemplified by organisms like plants, animals and fungi while the second mode, a less common strategy among eukaryotes, is nicely illustrated by the dictyostelid slime molds.”

B: “Many of the multicellular lineages present on earth today evolved from unicellular ancestors with rigid cell walls [like] land plants, fungi, and red and brown algae […]. For these cells, adhesion is a passive process very different from the intimate cellular associations found in animals and Dictyostelids; physical connections are established as new cells form and the resulting attachments between cells are stabilized and maintained throughout life. This type of multicellular development, in which cells divide and remain linked by their shared cell wall, has important implications for the developing organism as the cells cannot reposition themselves after cytokinesis.”

C: " animal cells lack cell walls, permitting them to adhere dynamically and reorganize into complex tissues and organs during development. In contrast, […] a cell wall […] provides structural integrity but prohibits cell rearrangement."

So, in summary, there are two main pathways (A) for achieving multicellularity:

  1. Division without cell separation (plants, fungi & animals)
  2. Aggregration of formerly disassociated cells (dictyostelid slime molds)

And there are two modes (B) of multicellularity:

  1. cell wall linkage (algae, plants & fungi)
  2. collagenous matrix embedment (animals)

The aggregative pathway (A1) is highly transitional and it’s hard to see how to arrive at an integral multicellular that way. The cell wall linkage mode (B1) is much more limiting and rigid and clearly only has yielded passive life forms.

So to boil it down for to something very simple in-game, it would make sense to stick to these phases:


  1. Prokaryotic phase Simple prokaryote-like cell
  2. Eukaryotic phase With the acquisition of nucleus
  3. Aggregative phase Like A1 and with the acquisition of “binding factors”
  4. Multicellular phase Like B2 and with the acquisition of “junction factors”

The latter phase will then set things up for the transition to the MULTICELLULAR STAGE.

1 Like

I like this plan, as it makes a clear path for progressing forwards in the game:

  • get nucleus, unlocks binding agents organelle
  • add binding agents and go around binding to other cells you find
  • after some condition is satisfied you get the junction (?) organelle
  • now you can design a body plan that gets filled out with cells as you divide. Once you have filled out the body plan you get to the editor and can modify the body plan. After that you start as a single cell again and start filling out your body plan again
  • after you have big enough body you can make specialized cell types in your body plan
  • then once you have big enough body plan, you switch away from the microbe view, and now place tissues instead of individual cells. This will be kind of a hard transition as the cells won’t be drawn anymore, instead the player will only see “big” creatures like themselves

I’m fine with hunting down cells to bind with, but I feel it could get pretty tedious for the player to have to repeat that process several times. Maybe after the first time or two the player will begin producing cells to fill out their bodyplan when consuming phosphates and ammonia instead of simply enabling reproduction. The player could still hunt down other cells to bind with, but they could also produce more cells themselves to help alleviate the potential tedium.
Other than that? I love this plan!

1 Like

I’m all for diversity of options and there’s no reason for the player to have to bind foreign cells, when they acquire “binding factors”. However, hhyyrylainen has a point in a stepwise sequence of abilities being a common and clear path to progression in gameplay. Nevertheless, I do recall that after every cell editor intermission the gameplay starts with the parent cell being close by. If they’re quick, a player can catch that cell and kinda have a simple “division without separation” already.

1 Like

I do think thats something thats going to be changed. I have personally found it a bit weird that all the same cells are still around you when you exit the editor, since it represents a huge time jump

I think it has been discussed a lot already, and it has been decided that things will stick around when you go to the editor (unless you switch patches), as otherwise it is a free “get away from a predator” button.

ah i see

Stumbled upon this paper dealing with “Synthetic multicellular organisms” which actually does a good job summing matters up.

It’s still a bit technical, but I think this figure gives a good and relevant illustration for the overall steps towards multicell:


Where might we turn for a first look at what genetic and epigenetic modules are needed to engineer multicellularity? […] Genomic and proteomic analysis has shown that, remarkably, components of many of the genetic systems once thought specific to metazoans and bilaterians […] and thought to be crucial in the development and maintenance of complex forms are present in choanoflagellates […] Some of these pathways (e.g., cadherins) appear to have arisen before multicellularity, were involved in environmental and prey–predator detection, and were coopted during the transition.


A fifth observation, likely to be emergent, is that choanoflagellate colonies appear to form not due to aggregation, but due to non-separation after division (with the concomitant production of a matrix and cell junctions).

A bit more on cadherins:

1 Like

Some thoughts on how we can transition from the simple, cell-to-cell editor to dealing with a macroscopic body plan. Parts of this concept depend on talk related to how the editor will work, so it isn’t a self-contained concept. But it should serve what I think is a robust outline for how we can transition out of the microscope.


As Nunz stated, we are on the backend of development for the Microbe Stage. We are focusing less and less on implementing small pieces and are focusing more and more on the puzzle as a whole. As such, it is important to identify the general idea for transitioning away from the Microbe Stage and towards the Multicellular Stage. That way, as we start filing down the rough edges of the current stage, we can be assured that we are heading towards the same direction, and that that direction is the best possible way forward.

The biggest challenge represented by the Multicellular Stage is the transition from unicellular/simple multicellular editing of individual cells to dealing with an entire organism. We obviously can’t charge the player with having to finetune every single cell type that goes in a complex organism’s morphology; that would require a lot of individual, incredibly in-depth parts, such as contractile vacuoles and neuron sheaths, digestive cilia, etc., and would require a lot of knowledge on the hand of the player to adequately put together a creature out of these endless cellular variations. There are pretty long-established concepts regarding this need to simplify things - namely, that the editor will transition away from editing individual cells and instead editing tissues. But beyond that brief synopsis, there is barely any meat to the concept (as far as I am aware). But have no fear, for Deus has a scheme.

My idea essentially would involve incorporating germ layers into the game. Germ layers serve as the backbone of embryology, but they are incredibly important in phylogenetic analysis of Metazoans as a whole, and have very profound insights on evolution.

I’d like to note two things. First, while I am decently familiar with germ layers as a whole, embryology and developmental biology as a whole are incredibly complex and difficult topics, so I am by no means very well-versed in the topic and would really appreciate input from theorists and the team as a whole. And second, while I believe plants and fungi have analogous cell differentiation structures, I haven’t looked into them yet. So this addresses the animal-analogue part of Thrive. I plan to address plant evolution when I add on to the sessile gameplay concept I shared with you guys on Discord.


Germ layers are most commonly used in the context of embryo development. They refer to one of three layers of cells in an embryo, and are important because each of these layers specializes into different types of cells.

There are three germ layers: the endoderm, the mesoderm, and the ectoderm. The endoderm is the innermost layer of cells. The cells which form the endoderm differentiate into the digestive tract (including the mouth opening and the anus opening) and digestive organs, the basis of the respiratory tube, the endocrine system, auditory system, and urinary system. The mesoderm is the middle layer of cells, and mesodermal cells differentiate into muscles, bones, the circulatory system, and various other advanced tissues that augment other organ systems, such as developing into lung/gill muscles from the endodermal root provided by the respiratory system. And the ectoderm, the outermost layer, differentiates into cells which form the epidermis and the nervous system (including the brain and spinal cord).

The Embryo Project Encyclopedia is a great source of information for the above concepts. Here is a page for the Endoderm, and other pages can be accessed through the search bar: Endoderm | The Embryo Project Encyclopedia

Germ layers are evolutionary significant because they are important phylogenetic markers. When we talk about organisms possessing differentiated tissues, we are really talking about organisms possessing germ layers. Sponges technically have at most 12 different “types” of cells, but aren’t considered to have differentiated tissues because they don’t have germ layers - actually, they technically only have a single germ layer, but it can’t really be classified as endodermal, mesodermal, or ectodermal in the same way more complex body plans can be. The most simple animals - Porifera, the phylum which encompasses the sponges - only have a single germ layer. Ctenaphora and Cnidaria contain two germ layers, the endoderm and the ectoderm, and are referred to as diploblastic animals. All other animal groups - the remaining invertebrates and the vertebrates - are triploblastic, having the ability to develop muscles and skeletons.

An area of research involves tracing the evolution of endoderm, mesoderm, and ectoderm throughout phylogeny. Diploblastic organisms evolved from multicellular animals without dedicated germ layers, and triploblastic organisms evolved from diploblastic animals - that much is clear. There is a loose understanding of how diploblasty evolved, and I can pull up more research for that later. The gap between diploblasty and triploblasty, however, is a bit less easy to span because modern diploblastic and triploblastic organisms appear to be very different. It is likely that a flatworm-like creature was the basal triploblast. Platyhelminthes are considered to be the most simple of all triploblasts, lacking a dedicated circulatory or respiratory system.


At some point in the early multicellular stage, the player reaches a threshold; instead of individually dealing with editing types of cells and placing each individual cell down in the editor, a general body-plan is presented (what is this threshold; number of cells, an unlock?).

In their next trip to the editor, instead of the familiar microscopic interface, the player is looking at a very simple, soft-bodied wormlike blob. Although their previous cellular composition influences the abilities this worm creature has - if they had toxins they retain that ability to inject/secrete, if they had mitochondria they are aerobic, etc. - they are no longer able to place organelles/proteins. Instead, the player notices they are now able to interact with larger parts - appendages, eyespots, and more (we need to think up of some of the most basal components we’ll offer the player in the earliest macroscopic stages). They’ll also notice the editor has changed slightly: the three tabs are now “Structure”, containing the parts to be placed, “Body Plan”, which I will describe, and “Behavior”, which affects the organism’s behavior.

“Body Plan” replaces the “Membrane” tab, and it deals with widespread changes to the germ layers your creature has rather than placing specific parts. Because the basic macroscopic organism is diploblastic, that would mean there would be two “sections”: the Ectoderm and the Endoderm.

The Ectoderm deals with the “skin” of your organism, so it behaves somewhat similarly to the previous Membrane tab. Just as membrane types previously existed in the microbe/early-multicellular stages, similar membrane types also exist in the macroscopic editor. For example, a chitinous ectoderm would be analogous to flexible exo-skeleton as is seen in arthropods, a calcium carbonate ectoderm would indicate the creation of a rigid yet robust shell as is seen in clams and snails, and double ectoderm would be analogous to “basic”, uncovered yet versatile skin as is seen in most animals. Ectodermal thickness can have implications on movement and temperature resistance, and

The endoderm will represent the start of messing with the organ systems with your organism, and contains the roots of various important systems within them. Containing the roots of the digestive tract, the respiratory system, (and the endocrine system if we see the need to include it/come up with fun gameplay options), some limited customization options will be present already in diploblastic organisms. For example, the digestive tract can allow customization of diet based on membrane, and can be customized to allow more storage and somehow correspond to digestive efficiency (how quickly you extract energy from food in your gut). Perhaps more storage means less efficiency while less storage means more efficiency? There can be many ways of dealing with this.

The basal respiratory system can have a slider related to oxygen-intake efficiency. Higher oxygen-intake efficiency gives you more energy quickly but burns through your energy source quicker, meaning you’ll need to be more ravenous, and also meaning your organism doesn’t do well in low-oxygen settings, such as the deeper ocean. Slower oxygen-intake efficiency gives you less energy but burns through your energy source slower, meaning you’ll have slower metabolism while also perhaps allowing you to live in the ocean depths. There is a bunch of customization options with the respiratory system I can envision related to invertebrate respiratory systems, but let’s make sure the basic idea of this concept is accepted before we spend a lot of time on that.

The mesoderm will of course be missing at first in diploblastic organisms. Until then, the area in between the endoderm and ectoderm will be filled by mesoglea - the gelatinous material found in Cnidaria and Porifera which can be composed of both acellular and cellular material (question for theorists: is mesoglea assumed to be basal for all diploblastic animals or is it a unique adaptation to certain diploblasts?). A lack of mesoderm means the player will have limited options in regards to functionality drawn from mesoglea. However, one important function can be related to the organism’s density, and thus, whether your diploblastic animal drifts or crawls on the ocean floor: the more mesoglea, the less energy your organism needs to float above the surface of the water. As such, we can have a slider denoting the amount of mesoglea in your organism for now - more mesoglea means a lighter, more floaty organism akin to jellyfish, more agile and nimble yet more vulnerable to currents and with less health, while less mesoglea means a denser, more durable benthic organism with limited ability to get off the ocean floor. When the player evolves mesodermal cells, the mesoglea’s functions (density) will be replaced by some of the added customization options - for example, bone density and air bladders composed of mesodermal cells can fill in mesoglea’s function.

Adopting the mesoderm will replace this mesoglea with muscles, and will almost serve as the “nucleus” of the late multicellular stage, catapulting the player towards the early Aware stage. The mesoderm will allow the roots of organ systems provided by the endoderm to expand in creativity and function, increasing the efficiency of the respiratory system with an upgrade and allowing the adaptation of advanced gills and lungs. The evolution of mesoderm will allow the evolution of bone tissue, allowing the evolution of a spinal cord, and eventually, parts like vertebrate limbs, skulls, teeth, etc.; and for organisms with chitin, the evolution of advanced muscles will allow the evolution of segmentation and arthropod appendages. There’s a lot we can discuss with the mesoderm, but again, for now, I want to make sure the basic idea makes sense.

A very basic and sucky Paint concept from me of how it can look like with a ctenaphore-inspired analogue. I tried to create a GUI for the editor, but I’m not very good as you can see.

Note: I didn’t address mesoderm a lot for a reason. The mesoderm really allows the proliferation of organ systems as you might see from me mentioning bones, circulatory system, etc. Many organ systems involve a root provided by the endoderm and expanded upon by mesodermal tissue. There’s a lot of flexibility there of course, but I want to make sure the basic idea is robust enough for you guys, and I want to make sure that the closer future of Thrive is more perfectly addressed.


I think this idea presents a very flexible and game-friendly way to represent the evolution of an organism from a simple multicellular organism to a complex organism with a general body plan. However, there are still questions remaining. This concept of course relates heavily with concepts for the 3D editor, so there’s a bit of a gray area where I’m not exactly sure how to deal with specific cool adaptations we see in life today.

Here are a list of questions and observations…

  1. Sponges and the Such - Sponges only have a single germ layer, but are still macroscopic organisms. This ties into the conversation surrounding sessile gameplay, and I had some thoughts about that I shared with you guys, but how could we represent such organisms?
  2. Ectoderm Details - How do we deal with, and balance, the player having multiple types of ectoderm? For example, a player who wants a shell they can retract into would probably be eying calcium carbonate or chitin, but they wouldn’t want all of their organism’s ectoderm to be rigid; they’d still want a soft-bodied part. Perhaps a specific part in the Structure tab can deal with this?
  3. Transitions, Transitions, Transitions - How will we incorporate the transition to diploblasty, and from diploblasty to triploblasty? For the former, I’d assume we’d want something else rather than “get x amount of cells and click a button in gameplay”. For the latter, it’s a matter of limiting progression fairly. I didn’t cover the nervous system here although it is a component of the ectoderm; perhaps developing a very simple nervous system will transition you to editing a body plan, and upgrading the nervous system will transition you to triploblasty? But then there’s the question of how to represent the nervous system; maybe it can be an upgrade in the Body Plan section of the simple multicellular stage as is seen in game now, where you can click a button that automatically creates a nervous system after you are generating a certain amount of ATP?
  4. Plants and Animals - Plants have analogous structure to germ layers within them: dermal, ground, and vascular tissue. How do we deal with that transition for that path of life? Perhaps we should rename “germ layers” to something a bit more neutral, and have different adaptations for organisms that have chloroplasts. Ties into the question of sessile gameplay/editors.
  5. Different “Types” of Editors - I put the word types in quotation marks because we just need different “tools” within the same macroscopic editor instead of completely different editors. I think we’d want different “tools” for soft-bodied organisms like jellyfish and worms and molluscs and the such, vertebrate organisms with internal skeletal structures, and exo-skeleton organisms with external skeletal structures such as in arthropods. For example, vertebrates would inherently edit their morphology first by adding bones for limbs and skulls and the such; it wouldn’t be realistic to have a arthropod-analogue organism be edited through similar tools because arthropods edit morphology through segmentation rather than through internal skeletal editing. The good news is that there are analogues and overlaps. For example, nothing stops a soft-bodied appendage, such as gills on axolotl, from appearing on vertebrates, limbs across vertebrates and arthropods are basically identical, etc.
  6. Worm World - I think the basic “template” organism we can expect to see in Thrive is a worm-like organism; for example, I anticipate the player’s basic animal which emerges from the microscopic-macroscopic editor transition to be a worm-like creature. From this worm, different customization options can be attached to allow the player and auto-evo to expand in morphology. A somewhat similar trend is apparent in real evolutionary history - there appears to have been a “worm world” a bit before the Cambrian.
  7. Organ System Variance - There are differences in organ systems observed across Eumetazoans. For example, arthropods have an open circulatory system whereas most other “higher” animals have a closed, tubular circulatory system. An open digestive system has an entrance (the mouth) and an end (the anus) whereas a closed digestive system only has one opening for both waste and input. We can have “variants” of organ systems with their own attached benefits and shortcomings, so that could be a future area of much discussion.

One last note: I think a good way to measure the strength of a concept related to the editor/progression question is asking yourself if the concept can realistically create a specific unique organism you have in mind. For example, if an editor concept can create a coral, a jellyfish, an insect, and a vertebrate, I feel like that’s a robust concept.


Alright, so here’s my added two cents to this idea-

It’s pretty obvious that people don’t want just straight up fully formed limbs, but more customizable limbs that can be morphed and changed over a period of time, but we also need something that is able to sense it’s a limb, so things can walk easier, so here’s my idea:

Basically, limbs are formed through connecters, giving the player control on where and how they’re placed. The could also be different kinds of limb parts. Some to bulk up limbs, some to make limbs more tentacle like or fin like, maybe textures that affect heat or toughness of skin, etc.

1 Like

@Deus I really like the overall idea of using germ layers to make the differenciation of cells more understandable and streamlined.
I was watching a video about germ layers when a question popped back into my mind: Do you know if it would be possible for alien life to develop analogous functions to our bodily functions but develop them out of a different germ layer? For example is there a reason why reproductive organs would always develop out of the mesoderm layer? Could alien life not develop reproductive organs out of the ectoderm?
I like the idsa of using germ layers, but I wouldn‘t want to restrict life to evolve completely parallel to LAWK if there isn‘t a reason why certain organs would only ever evolve from certain germ layers.

1 Like

That’s a good question, and I’m pretty sure the answer is that generally, life would evolve in that “order” of germ layers.

The big idea behind germ layers within an evolutionary lens is that organisms first evolved components that arise out of the endoderm and ectoderm because those represent the logical next steps in complexity on the path towards developing complex organs . For example: an organism would need something akin to a gastrointestinal tract to develop a stomach, a large intestine, etc. that comes out of mesodermal tissue, and an organism that becomes macroscopic would likely need some sort of tissue devoted towards skin and the outer layer, as is seen in ectodermal tissue. So I would think that organisms as a whole generally follow the same “outline” of sorts when it comes to having endodermal, ectodermal, and mesodermal tissues arise.

That said, there is some variance among many triploblastic organisms in, for a lack of a better word, the “proportionality” of mesodermal and endodermal tissue for a specific advanced organism. To be more specific and clear, certain organisms might have a stomach that has a lot of function derived from mesodermal tissue, while other organisms have a stomach that is mostly endodermal with just a few mesodermal tissues devoted to them in development. So perhaps this can indicate that the types mesodermal tissues don’t necessarily have develop in the same order - there isn’t, say, a necessity that the lungs evolve before the stomach, that a circulatory system evolves last, stuff like that. I’d like theory input on this topic however.

1 Like

In light of discussions on Discord, I wanted to post a though regarding how to approach the transition to tweaking body plans instead of individual cells.

I think we should first think of the broadest level of detail first - how editing the entire organ system will work - before we think of the finest level of detail - creating and editing individual cell types that compose specialized tissues. By that, I essentially mean instead of designing an organ system by having the players place down tissues they made to be a neuron or a vessel or a whatever, we’d essentially generate the system automatically, think of the broadest scale first, and find some small-scale areas/details for the player to tinker with.

I can foresee many problems that may occur with the first approach, of making the player design an entire system by placing individual parts:

  • Can be way too iterative, where you may only have enough MP to place maybe half of a nerve net you’ve been meaning to place, or half of a muscle

  • Requires sooooooo much work for so little output: would we be designing specific components for tubular blood cells, neurons, skin cells with oils and secretion, fat cells with lipids, digestive cells with special cilia, etc. that would realistically only be useful for a single type of cell?

  • There’s variations within the cells of a specific tissue as well - short twitch v. long twitch muscles, long neurons vs short neurons, different types of epidermis - that would be unruly to implement

  • I really just don’t see how we can assume that the player will have enough knowledge and skill to place cell layers in a way that ends up in something like a limb with bones, vessels, nerves, muscle cells, etc.

Whereas I can see a lot of benefits from the latter approach, of providing a broad organ system automatically to the player, prioritizing broad scale changes, and finding small areas to allow things like the editing of cells:

  • Can reduce tediousness and repetitiveness, as it would be annoying to have a limb that is otherwise perfect except for the fact that you haven’t perfectly synched nerve network, blood vessels, etc. Structures don’t form system by system as they probably would with a limited MP pool; that assumes too much foresight and planning from evolution.

  • We really need to ensure that we don’t overwhelm the player with designing different cell types for different organ systems at the same time. By providing a broad template for the player to tinker with, we have a steady and controlled way to introduce anatomy to the player.

  • Let’s us focus less on devoting resources to incredibly small components (specific organelles of cells and tissue types) of what will undoubtedly be a huge stage (macroscopic) and let’s us focus on making sure the game flows as well as possible. It makes it easier to envision a nervous system when we don’t worry first and foremost about the individual neuron for example.

So, what does this mean? The transition should probably involve a pretty large jump in complexity where a player is provided with the most basic forms of advanced organ systems. So instead of approaching the transition to the macroscopic stage by thinking of ways to create different highly specialized cells, like neurons or blood cells, think of ways we can involve cool broad, macroscopic tweaks to those very basic organ systems. Then, from this broadest level of detail, perhaps we can work our way down to figuring out opportunities to involve the player’s control over details as small as individual cell anatomy.

Here’s what I’m doing for the prototype regarding this. As it continues from the early multicellular prototype, all of the cell types the player has created will be carried over and can still be edited and duplicated like before. The editing is done with the microbe editor except there will be a few multicellular exclusive organelles. In late multicellular the placement of individual cells is replaced by placing down metaballs. These metaballs have a size, position, and parent which define the metaball structure. In addition to that each metaball has an associated cell type. So basically the cell types define which type of tissue each placed metaball is. I’ll keep it simple initially and probably have just two specialized cell types (designated by placing the exclusive new organelles): neurons and muscles. Placing enough total volume of neurons will be what unlocks the aware stage, and for muscles I was thinking that the player could select the muscle type when creating a joint between metaballs, which would have some kind of effect on the joint in the future.

That’s basically how I imagine the base body shape will be done even once the late multicellular is no longer a prototype. This model doesn’t address blood vessels or any kind of systems like that. I think we should probably leave designing those parts of the game until the basic metaball based editing is done and we get feedback. We should carefully balance how many systems the player must juggle at once in the late multicellular stage, it should not be overwhelming.


Fleshing Out Hydrostatic Skeletons, Endoskeletons, and Exoskeletons

In an effort to create greater clarity on our plans for the macroscopic stages and seeing as the base of this concept seems to be well-received, I would like to jot out some thoughts I’ve had building on this base concept. Insight and opinion is, of course, very much welcomed.

Looking at metazoans, there are various methods through which organisms “create” structure in their body plan. A large majority of diverse animal groups are soft-bodied creatures, typically relying on a hydrostatic skeleton with no solid objects to root things like muscles on. Vertebrata utilize an endoskeleton, with an internal bone structure as a root for muscles and more advanced organ systems. And there are exoskeletal creatures, such as arthropods, who utilize an exoskeleton to give themselves form. Note that another major structural group are shelled organisms, such as snails and other molluscs, that utilize a hard substance as a structural anchor as well; but those are technically considered to be soft-shelled creatures with unique adaptations.

Those different types of structures come with their own costs and benefits, and define heavy implications on morphology. Organisms with an exoskeleton demonstrate a heavy level of segmentation due to their rigid bodies. Their genome has adapted towards this explicit segmentation, allowing them to replicate limb structure much more easily than a vertebrate genome can - the latter needing to ensure that multiple resource-intensive organ systems coordinate to create a functioning limb. However, organisms with exoskeletons face limitations with their size on land, need to molt, and require special adaptations to ensure a proper exchange of gas and adequate sensory information. Organisms with an endoskeleton, such as vertebrates, are much more resilient against the square-cube law, allowing them to deal much more easily with size.

Exoskeleton (Arthropods)


Segmentation Means More Replicable Parts - Segmentation is a lot more explicit in arthropods than in other organisms, with each segment being its own defined part of a body (thorax, head, abdomen, etc.). The arthropod genome has structured itself around this segmentation, meaning genetic material for a limb can essentially be “copied” much more easily than it might be for a vertebrate, allowing arthropods the ability to have a large number of limbs and appendages.

Greater Protection From Environment - A tough external coat of armor surrounding the entire body protects from blunt trauma and predation.

Easier to Keep Moisture In - It is much easier to keep moisture within a solid and rigid object than it is for a continuously exchanging medium, such as skin. This effectively means it is easier for arthropods to make the transition from water to land, and vice versa.


A Lack of Flexibility - Because the entire body is covered by a chitinous coat of armor, arthropod joints tend to have much less dynamic flexibility than a ball-and-socket joint in vertebrates have. This can limit speed and agility.

Molting - A rigid exoskeleton cannot continuously grow alongside an organism, so it must be shed every once in a while. Not only does this leave an organism vulnerable to predation, but it requires intense effort and can potentially take away from efforts to gather food and other resources.

Limited Size And Less Force Potential - Compared to an endoskeleton, an exoskeleton is much more subject to the square-cube law. Not only does this mean that arthropods cannot grow as large as other organisms, but should they theoretically reach that size, they would not be able to exert the same amount of force on their skeleton as a vertebrate might.

Gaseous Exchange Limitations - The same property of a chitinous membrane that allows them to retain moisture better than other organisms also limits gaseous exchange between an arthropod’s internal systems and the environment. This means less efficient respiration, which essentially means limitations to sizing up on land.

Endoskeleton (Vertebrates)


Happy Mediums - Organisms with an endoskeleton receive the benefits of having a robust skeletal structure, but are not as constrained than organisms with an exoskeleton in terms of movement and flexibility. They are also able to maintain gaseous exchange through skin, meaning a capacity for efficient respiration remains.

Greater Capacity for Size - Internal skeletons are much more capable of accomodating the square cube law than an exoskeleton, meaning a greater capacity for attaining size and preserving strength at said size.

Greater Force and Flexibility Capability - Being less rigid than an exoskeleton, endoskeleton-bearing organisms are able to exert force at a more dynamic joint range. This allows for greater strength, size, and other forms of force projection at size.


More “Set” Body Plan - Especially at larger sizes, it takes a lot to make an endoskeleton work properly. Because of this, rather than “simple” replications of segments as seen in arthropods, stem cells in a vertebrate embryo have to carefully coordinate expression so that development proceeds as needed. This essentially means that appendages aren’t as easily replicable in vertebrates as they are for exoskeletons.

Vulnerability of Exposed Skin - A greater capacity for gas exchange means a greater need for protection against the elements. This means the initial transition to land is harder for vertebrates than it might be for arthropods.

Hydrostatic Skeletons (Soft-Bodied)


Much Use, Little Resources - Soft-bodied organisms are able to conduct a great amount of their biological processes through simple membrane gas-exchange. They are also able to maintain structural integrity without need for dense skeletal structures, relying only on the presence of fluid/water. This allows great regenerative capabilities and requires little structural complexity. It also means a great amount of diversity in terms of soft-bodied body plans: jellyfish, octopi, molluscs as a whole, and all sorts of worms utilize soft-bodied skeletons.

Incredible Flexibility - A soft-bodied organism is able to fit through any crevice as long is said crevice is bigger than any potential hard-body structure (octopi beak, shells, etc.). This allows strong tunneling abilities, and allows organisms to hide in very discrete areas.

Lightweight Buoyancy - Lacking a dense skeletal structure, soft-bodied organisms don’t need as many adaptations should they be filling a niche near the ocean surface.


Limited Force Projection Capabilities - Because they have no anchoring point for muscles to utilize, soft-bodied organisms have limited force protection capabilities. As such, they don’t move very fast, and aren’t very strong.

Heavy Dependence on Water - Organisms with hydrostatic skeletons require a constant fluid source to maintain structure. Furthermore, because they have limited muscle strength, soft-bodied land organisms have very limited movement capabilities.

Very Limited Protection - Unless they have a shell, soft-bodied organisms are very vulnerable to abrasive force.

Limited Structural Complexity - Without skeletal units, fine structures, such as advanced graspers and jaws, are incredibly difficult to evolve. Soft-bodied organisms must rely on more nuanced structures, such as tentacles and suckers.


I notice that much of our discussions regarding how the editor will work are centered around organisms with endoskeletons. If we wish to adequately represent metazoan diversity, it is important that, depending on the player’s choices, the way the editor works will work differently depending on the morphological choices the player has made. While this means more work, I do think the underlying plan - metaballs and convolution surfaces - can work across all three forms of metazoan structure.

What I think will have to happen is essentially having different tools be available to players in the editor depending on the morphological choices they have made up to that point. This doesn’t necessarily mean that we will need a variety of parts - instead, the basal parts we offer to a player will have to behave slightly differently depending on if a player has an endoskeleton, an exoskeleton, or a hydrostatic skeleton.

I won’t go into too much detail with limbs since I feel like that’s a bit of a different topic, and so that the ideas contained within this thread can be reviewed first by you guys. For now, I’ll cover potential general differences and how editing the torso will work.

I also attached some boot-leg concept arts for some ideas I want to make sure are made a bit more clear than they might be just in text. Once again, apologies for their relative lack of quality, but I think they illustrate whatever point I am trying to emphasize.

Hydrostatic Skeletons (Diploblastic or Triploblastic)

Editing the Torso - Soft-bodied organism metaballs will probably operate a lot like vertebrate metaballs, but will have a lot more freedom in terms of playing with the properties of each metaball. To clarify more, remember how in Spore, increasing the size of a certain metaball along the spine would also make other metaballs increase in size a bit? For soft-bodied organisms, that effect will be a bit less emphasized so that one metaball will have less of an effect on another. This will allow more creativity with shapes.

Editing Limbs - One unique thing with soft-bodied organisms I think would be cool to see reflected in game, while also giving them a unique trait for the player to play around with, is for the ability for the anterior side of the creature to not necessarily be where its mouth is. I’m thinking of jellyfish and octopi in particular. Perhaps we can attach an ability for soft-bodied limbs to move backwards rather than forwards?

Regardless, the initial investment for limbs for soft-bodied organisms should be rather inexpensive, but modifications for these limbs should be somewhat expensive.

Here are some concepts for various fake organisms, inspired by real-life Earth analogues. Note that for organisms like jellyfish, we would probably need to incorporate a special modification of some sort to create their “caved-in” appearance.

Exoskeletons (Only Chitinous Triploblastic Organisms)

Editing the Torso - Each metaball within the torso represents a segment. To make a certain segment larger than another, the player can highlight specific metaballs and designate them as a group, which will serve an aesthetic purpose (what functional purpose should we attach?). For example, a player who wants to make a millipede like organism wouldn’t want to group any metaball together, while a player who wants to make an insect-like organism would want to group metaballs into 3 groups representing the thorax, head, and abdomen.

The segment with the mouth opening within it will be considered the head segment, and the segment with the anus opening within it would be considered the posterior.

Editing Limbs - Limbs should be rather easy to add for organisms with an exoskeleton, with both the initial investment for the joint and the modification of joints being rather inexpensive.

Concepts to illustrate segmentation. Note that different color metaballs belong to different segment groups. So for the centipede, each metaball is its own segment, for the ant, metaballs are split into three segment groups, etc.

Endoskeletons (Only Triploblastic Organisms)

Editing the Torso - This will probably be rather similar to Spore’s editor, so as of now I don’t think it needs too much explanation. Each metaball within the torso represents a vertebrae along the organism’s spine. You can manipulate each vertebrae to define your organism’s shape.

Editing Limbs - Limbs will be the costliest for organisms with endoskeletons (limb parts will cost more MP for endoskeleton organisms). I think we should make the player specify where they would like to place a joint alongside the spinal cord instead of just having limbs be free dragging akin to Spore, although I am not too adamant about this of course. Once the player specifies this, a single jointed appendage can be placed. From there, players can add additional limb segments.


Apologies if this concept appears a bit messy; I intended to present a rough outline of how we can address the morphological diversity of all metazoans as easily as possible. So I want to see if the basic ideas are sound before potentially spending more effort on the finer details of these concepts. I would like input on if you guys think this system can cover a wide diversity of body plans while presenting a fun way for players to interact with the decision they have made in their organism’s evolution up to that point in their playthrough.

And I would also like some input on whether or not this approach is possible in the first place. I do realize this could be a lot of work to implement, but I do feel like it’s the best way to represent a diverse amount of biota through the same underlying editor system. Much of our discussions have focused on the idea of vertebrate evolution, but I feel it would be inadequate to only provide an endoskeleton and call it an all-expansive simulation of evolution. Regardless, I’m sure we can simplify this further.

We would first need the soft-bodied tools to atleast work properly at the beginning of the macroscopic stage. So if we want to focus more heavily on this aspect in the future, that should be the first stop.


Really interesting work with all the science you’ve pulled up on this thread.

I totally agree with everything in regards to the tradeoffs of soft-bodied, endoskeletal, and exoskeletal structures. Those all seem to be the exact same tradeoffs I’ve seen when reading about the different body types. In fact I will need to bookmark this page for future reference once we get to implementing this. The point about segmentation is also very true, and as you say we need to find a way to make it easier for exoskeletal creatures to segment more easily than endoskeletal ones.

In terms of germ layers, I’ve thought about using them as a way to categorize the Organism Editor before. What I ultimately struggle to see is how such a system will meaningfully contribute to the gameplay for the amount of complexity it adds. What’s more, from what I understand it is mostly a categorization of how different animal organs have evolved (and how they develop as embryos) and does not directly apply to the evolution of plants or fungi. For example there’s no reason to believe life on other planets will also have three layers and evolve the same tissues and organs from the same layers. As such I keep coming to the same conclusion that it just doesn’t seem to be universal enough and game-effecting enough to design the Organism Editor around that (especially since it would add a lot of complexity for the player’s experience in the OE). But I’m curious as to what you guys think.

Since I’ve been away for a while I apologize if there have been newer discussions on this, but the latest concept for early macroscopic (aka the 3D transition) I remember is that once the player transitions to 3D, we assume that they are still so small and basic that they will not have any differentiated tissues yet, and that the specialized cell types they placed in multicellular don’t count as separate tissues. As such, they will spawn in their first 3D generation as a tiny 1cm wide blob of membrane. The blob will resemble the shape that their colony was. They will be only comprised of a single, all purpose tissue called “Mesoglea” or “Mesenchyme” or “Mesohyl”. This initial 1cm blob of Mesoglea represents the entire colony that the player was just editing the generation before. The way that the Microbe and Multicellular decisions translate over is that the starting stats of your Mesoglea tissue are defined by your design as a colony. So for example a colony with many aerobically respiring cells will have an initial Mesoglea tissue that performs a lot of aerobic respiration. If the colony had a lot of contractile fibers, then the Mesoglea tissue will have a higher “Strength” stat than otherwise. If the colony had a lot of adipose vacuoles, the initial Mesoglea will be very energy rich to any predator who feeds on this creature, etc.

Differentation then occurs once 3D has already started. The player can evolve this Mesoglea tissue to have different stats, they can evolve new tissues out of the Mesoglea, they can evolve entire organs, they can evolve to become longer or wider or both, etc.

Not married to this idea or anything, this was just the latest concept that I can remember of what the early 3D stage would look like that seems to cover a lot of the questions that need answering.

1 Like

Good points listed here.

Regarding germ layers; I actually find myself coming to a different conclusion regarding their universality, as even complex plants have 3 “layers” of cells (dermal, ground, vascular). I put layers in quotations because unlike animals, plant embryos don’t fold in on themselves, and these differentiated layers continue to develop heavily throughout the plants entire life cycle rather than primarily as an embryo Chapter 12A. Plant Development

A brief google search tells me than fungus development doesn’t feature the same level of cell differentiation, but for complex multicellular organisms, I would assume atleast some analogue to germ layers is needed to meaningfully differentiate cells.

As to their implementation in Thrive, I would think their only purpose would essentially be a way to pace out progression. I think they would essentially just serve to make sure we don’t see an incredibly rapid jump in complexity from simple jelly blobs floating in the ocean to an organism with a fully developed skeleton in just 3 generations. So I would think they would just be a sciencey way for us to justify slowing down game’s pace a bit when a player gets to macroscopic.

Atleast from my memory, there hasn’t been much discussion about the transition to macroscopic gameplay since that initial idea of a mesoglea glob. It seems the biggest difference between that and what I detailed was that for this newer concept, the digestive tract would be formed as you made that jump. I certainly would like a very seamless jump, but I worry about how much we would be asking of the player to form their digestive tract from piecing cells together. Of course however, we could make it as simple as a sort of upgrade however, and that’s what can introduce cell differentiation. It definitely is something to consider.

Regarding the current implementation (as of 0.5.9) the late multicellular editor picks off from where the player left off from the early multicellular. Which means that all of their cell types the player made before are converted into tissue types and all of the placed cells are converted to one metaball each (at least for now, also the conversion is not super good yet).