Microbe Stage GDD

Good news! The first draft of the new Microbe Stage Game Design Document is done:


…except it’s not. Anyone who wishes to read through will find one section completely empty: simulation specifics. Some of the simulation-related aspects are discussed elsewhere in vague terms, but I chickened out of writing up the exact mechanics behind everything. I’m afraid someone else will have to at least guide me through that part, or even write it themselves. Even worse, they have to make sure it complies with the rest of the document.

For all other sections, preliminary explanations are at least in place. There are lists of things in the appendix, detailed descriptions, and I even took the liberty of Americanising everything for the benefit of most of the team. Hence why there are so many z’s and not enough u’s.

In total it runs to over 27,000 words, even without the simulation specifics section. Something tells me it needs a thorough editing at some point.

Unfortunately I’ve had to make a few assumptions and invent mechanics myself for some parts, which I feel a little guilty about. This way though, the entire document is at least (as far as I can tell) self-consistent, which is more than can be said for the scattered concepts throughout the forums beforehand. I’ll run through some of these areas in a minute.

Everything in the document is subject to change. Some things more than others obviously – thermoplasts could be removed entirely, googly eyes will absolutely not be added for as long as I and many other people here are on board. Spelling and grammar (as well as clarity of expression) are also flexible, as right now I haven’t even read through anything again after adding it.

Here are a couple of the places I’ve made educated guesses or invented things, in order of where they appear in the document (I fully expect people to tell me at least one of these is completely backwards, and don’t be afraid to do so):

  • Straight off, the tagline in the summary, ‘Write your own saga of life’. This was based on a few pieces of promotional material we’ve created so far, and really was just added to make the GDD seem more GDD-like. We don’t necessarily even need one.

  • The entire game screen architecture didn’t exist anywhere other than what we already have in the current build. I think it’s mostly common sense and most people would agree it’s what was ambiently planned anyway, but you’re free to say otherwise.

  • Most of the options in the game setup interface have been briefly mentioned by people but never expanded upon. Getting those written up is one example of where the GDD is exceptionally useful: it’s forced me to consider parts of the game in implementable detail which had previously been glanced over. Obviously there are things I’ve missed or done incorrectly (NPC evolution rates and variance spring to mind, which, tied to the CPA simulation, require a theorist’s approval).

  • Free edit mode. This had been mentioned before and sidelined to a non-issue, but I’ve included it anyway. Essentially it gives the player an option to experiment freely in the editor with no MP restrictions, but they can’t use their creations in-game except under strict circumstances. Both the concept of free edit mode and the way it ties into the real editor are up for debate.

  • Some parts of the extras and tools pages have already been discussed, but never really set in stone. They’re still pretty bare in the document’s concept.

  • Every single control except those which are already in the build. These need to change based on QA feedback, but at least this gives a list which can be implemented and tested.

  • Weirdly, I’d found previous concepts which assumed the existence of rocks and surfaces, but never have I seen them ever discussed. Mostly this was the idea for lamellipodia, which I’m all for. They do require something to climb on though, which is why I added surfaces and rocks as environmental elements.

  • Some of the combat mechanics hadn’t been mapped out exactly, so I made a few tweaks (most notably the option to extend pili as an active attack method, rather than just passively swimming into something). @tjwhale since you’re the expert on combat, I’m mostly waiting for your input.

  • Everything about creating agents comes from responses to this thread and a bit of common sense. It’s never been a particular solidly defined concept.

  • Agent resistance is based entirely on the last few posts of this thread, and truth be told I don’t think I’ve entirely understood it. Feel free to correct me.

  • The agent list is a mish-mash of real compounds and imaginary ones. Gameplay-wise this might be the best choice, as it gives the player a wide variety of abilities (just as long as it doesn’t go overboard).

  • Reproduction was heavily glanced over due to its ties with the CPA system. Player reproduction is fairly straightforward, but otherwise it’s a little vague and probably not fully accurate.

  • Ah, good old osmoregulation. I was adamant from the start that it should be included as the method of health measurement and death, but it was only once I read the associated thread that I realised how complex it really is. Even @Moopli lost faith by the end of the discussion. However, I managed to keep it in the GDD in a somewhat intuitive way by combining it with the organelle health system. I say somewhat because it’s actually quite complicated, but makes sense after a while thinking about it. Unfortunately the player can’t afford to spend a while thinking about it, so this system probably needs some tweaking. I think it should still be included, just modified to make it less confusing.

  • The method of creating colonies was a little vague everywhere I saw it. Signal agents…bonding agents…???..colonies! The GDD’s description is just my best guess at how it might work, but as I’m not one of the theorists my opinion on it really means nothing.

  • Organelles were about the only thing fully described in the old GDD, so the rules for adding and using them were pretty easy. Each organelle type had some description, but not much, and rarely was it anywhere near rigorous. I think I’ve achieved that here (possibly overly so), but I did end up making a lot of assumptions which you’re free to contest.

  • The appearance tab in the editor seemed a little unloved, given that the membrane won’t be editable as we originally thought. I added some detail, but probably not enough.

  • Disasters have always been an assumed feature, but for the microbe stage especially they’ve never really been discussed in detail. In my opinion they should be left until everything else has been added as they’re incredibly low priority. Still, their effects on populations and the evolution processes in play should be interesting to say the least.

  • The tutorial is all my creation, but then again I’ve had its concept in the back of my mind for ages now. I think we need some sort of tutorial for both the environment and microbe editing in as soon a release as possible, but the final tutorial is completely subject to QA feedback.

  • The entire visuals section feels a bit brief, but I guess it’s up to the people actually creating all visual assets to work out the details. Adding extortionate detail would just stifle them.

  • Audio is completely fine other than a few unknown sound effects. But hey, what did you expect when you have a musician writing it?

  • The opening cutscene description is redundant but included anyway, as we already have one. Credits, on the other hand, is massively vague as nobody has much of an idea what they’ll look like. Again, it’s up to the animators.

  • Do we need more Easter eggs? Or is even considering one at this point too much?

  • I only came up with three biomes, and even then one doesn’t feel right (shallow ocean). Microbial biomes are hard to come by.

  • All the appendices are incomplete to an extent, as there are likely to be many more elements we haven’t yet considered. Agents, compounds, processes and sound effects especially.

Now, this all sounds good, but I’m still worried this won’t help progress much at all, In fact, it could even hinder it. As @tjwhale has said, he doesn’t want to make the blueprints for a building and expect other people (programmers and artists) to build it for him. This entire document is a prime example of that – I fear that by writing this I’ve become the ultimate ideas guy, which was never what I set out to do. Even more so considering the simulation specifics are non-existent right now.

Programmers could look at this and see an impossible list of things they need to do with no scope for their own input, and that’s not good by any stretch of the imagination. On the other hand, it’s better than an amorphous blob of concepts that we ambiguously wave towards and say, “Make this game, my minions!”

It also answers this set of questions at least. And that’s really what I set out to do.

I wanted to show that we did have a solid concept, it just needed to be gathered together and organised properly. Truthfully, very little of the document is my own work, I just had to make a few tweaks here and there so no concepts would go stepping on each other’s toes. Hopefully now we can use this in outreach in answer to anyone who says we don’t know what we’re doing, because from my own experience speaking to people here, everybody very much does.

Also I suggest that, if everyone agrees, we post this to all our outreach channels as proof we’re still alive (maybe once the simulation stuff has been written up). Perhaps a new website devblog and a copy on the community forums, and they could also include some explanations of what we’ve done with the forum structure and new application system.

Anyway, sorry for the massive post and even massiver GDD I’m asking you to read through.

Now, if you don’t mind, I think I’ll take a break from Thrive stuff for the next week or so. I’m shattered.


It took 4 days to read (the GDD, not your post…although you post was pretty long); however I have finally finished it. I want to begin by saying what a great job you did. To say the truth, I never really realized how much of the microbe stage is already planned out. I didn’t notice any mistakes or inconsistencies while reading it, but then again, I am fairly new here so you shouldn’t take my word for it.

The tutorial part was actually one of the most engaging things I read and I really liked the way you have it set up—it teaches everything in a calm manner without making it too difficult for the player. And as many people have said on the subReddit, we really do need a tutorial.

Easter eggs are definitely a great idea and one of the things I really enjoy in games, but I agree that we should probably wait until we have a finished microbe stage.

Again, great job.

Thanks! Yeah, I definitely think some drastic editing is in order to cut it down to a more reasonable length. It’s all well and good having everything finally documented, but it’s useless if no one wants to invest so much time in reading it.

I was surprised too, but there were quite a few gaps which needed to be filled. Everything to do with the simulation is still a bit sketchy as far as I can tell, which is all the more reason to get it written up properly as well.

Again, thanks! I feel I’ve got the start pretty much spot on (learning how to move, collect compounds, use the editor and attack with spikes) but after that it gets quite vague. Really after that it’ll have to be mostly tooltips and text explanations, which aren’t the best, but I can’t see a way for any proper interaction to be added once the simulation starts.

I’ve had a crack at beginning the Simulation Specifics section, and it’s already racking my brain:


I thought about the compound section alone for several days beforehand, which is where all the problems came from. Some of these might have already been addressed elsewhere, but if not they’ll need to be fixed before being added to the game.

I also finally used parts of the document @tjwhale sent in Slack for the CPA System, but unfortunately it didn’t really go into enough detail for my tastes.

As withe everything, I’ve probably made mistakes by assuming incorrect things. @tjwhale, @Moopli and @Seregon can correct me.

Great work, it’s all coming together nicely.

On the question you asked about compounds locked, if that’s the system we end up with, the way it works is continuous. So basically your species is like a sack and you have 1,000,000 protein and you lock away 10% per minute (or whatever) and so, after a while you have 500,000 free and 500,000 locked. It’s representing thousands or even millions of individuals being born and growing, but because that happens so smoothly we can call it continuous.

Then over time you also lose, say, 10% of your compounds locked per minute. That way there is a steady flow from your compounds locked pool back into the environment. This represents members of your species dying.

Finally your species absorbs compounds from the environment into it’s free pool. So there is this continual flow

environment - [absorbed through membrane] -> free pool - [growth] -> locked pool - [death] -> environment.

Does that make sense? Basically so many of these actions are happening that it looks like a smooth flow. Individually it’s single microbes being born and growing and dying in sudden shocks but with so many happening it looks smooth.

Also your point about infinite water is a good one. Given energy over time it could cause problems.

Overall interesting. As you say above I am most interested in the person who builds this having ownership of it. Whoever they end up being I’d love to help them as much as I can in any way. However until then I’m not sure I’m helping.

Nice job again!

I personally liked this idea:

“Another solution is to have reproductase transfer the compounds over. In essence, this would involve using reproductase as a label for stored compounds needed to produce offspring. Compounds could be converted to reproductase, and eventually released on death and digestion. This does raise further questions though. If organelles have a set makeup, does an organelle-heavy cell need more reproductase to divide? If more organelles are added as a mutation, where do the required stores for them come from if not thin air?”

In nature, bigger cells take longer to divide and move slower. Having reproductase be the amount of compounds necessary for a cell to split in two gives it an actual scientific meaning rather than simply it being a magical molecule that no-one can understand how it works (add reddit and the old forums are prime examples of this fact).

As for your second question, the organelles you add in the editor could already be factored into reproducase’s cost. By this, I mean that you already have a set amount of mutation points in the editor; assuming the player will always use as much of it as they can, we can multiply 100 by the average amount of compounds an organelle requires. Sometimes the player will put in compounds above this average, sometimes below, but if the player goes in the editor often enough the required stores will balance out—sometimes they will dissapear into thin air, other times they will come out of thin air. Does this make sense?

So the game world as a whole ignores the exact mechanics of how compounds would be transferred between generations? I assume that cells in the player’s field of view (which are simulated fully) wouldn’t feed back into the system in the same way then.

I hadn’t considered using assumptions of how a cell would mutate next generation to calculate an average requirement for reproductase. That sounds like it would work until you realise that Mutation Points don’t just cover adding organelles - they can also be used for upgrading organelles, removing organelles or changing behaviour. None of those actions leave extra organelles to represent the locked up compounds, so there’d still be a compound deficit.

It might still be possible to calculate an average taking these things into account. We’d just have to accept that, for the player’s patch at least, compounds involved in reproduction would eventually deviate wildly from initial values.

Yes, I guess that is true. If the parent has a mass of X and then has an offspring and afterwards both the parent and offspring have a mass of X/2 then the total amount of compounds in the species has not changed. We’re not keeping track of any individuals or anything like that, just the blueprint/genetics of the species and the amount of compounds they are holding in total.

I think the plan when it comes to relating the swimming around with the CPA model is that the swimming around will have no effect on the underlying CPA (if you kill 10 out of 1,000,000 members of a species in your patch that is closer to 0 than anything) however what you experience as you swim around is slaved to the CPA. So if the CPA says your biggest predator is species Y and they outnumber you 50 to 1 then as you swim around you will see 50 of species Y for every one of your own species you meet.

The idea behind this is that when you mutate your creature in the editor you will have to optimise for two things, 1) While swimming around what organelles do I want and 2) in the CPA how can I help my species the most. If the swimming around is slaved to the CPA tightly enough, and the CPA models the swimming around well enough then these two problems should be the same. Ie if you add a flagella and that makes you better able to flee your predators then also your species does better in the CPA for the same reason.

However this is a super complex problem so we’re going to have to wait for some programmers who want to work on it and then let them build what they can and then accept that. I don’t want to get too pie in the sky about it, you could easily do a PhD on this problem (many people do, modelling microbial interactions, for example).

There’s a simple solution to all the problems with reproductase: remove reproductase entirely.

I’m pretty sure that’s already the path we decided on – now that we’ve figured out how to model the-stuff-a-cell-is-made-of (with ‘locked’ compound pools), we no longer need a magical compound to represent the-stuff-we-need-to-make-a-new-cell, because those are one and the same. If an average, healthy cell has a certain amount of each compound, then two of them would have twice as much. Therefore, for a cell to be able to split into two average, healthy cells, it must have twice as much, and then a bit extra to cover the energy needed to reproduce. Note my wording – cancer cells, for example, generally reproduce too early.

How strict we should be about having double of everything is another question entirely (for example, if we have a glut of food protein, can we split earlier even if we don’t have large enough glycogen reserves?), but it provides interesting tradeoffs for gameplay.

The other question, of what we do with changing cell mass through evolution, is simple: The population still has the same amount of each compound, so if each organism is bigger, there are now fewer of them.

I’ll fork off a discussion before this gets even further off-track.

Since I’m about to get tied up in other things again, I’ve added a list of things that need changing to the start of the GDD. Things like replacing reproductase, changing the way external organelles are placed, etc.

That’s now done, so once again everything but the Simulation Specifics section should be fully up to date.

I replaced the internal/external organelle system to one with three classes: internal, external and periphery. Internal and external organelles are placed in the same way (as hex arrangements, no more fiddling around with edges), but external organelles can have their orientation changed in the Appearance tab. Peripheral organelles were discussed here, and cover an entire cell with a shader coating. There are currently only three peripheral organelle types - cilia, lamellipodia and cell walls.

Reproductase is now gone everywhere but the Simulation Specifics section (which I’m yet to update). Hurrah! Instead, compounds are now converted to a locked-up mode, determined by how many organelles a cell has. These locked-up compounds can’t be accessed by any means except when a cell dies, when they’re distributed amongst floating organelles (which already contain the locked-up compounds used in the previous generation to produce them.

1 Like

Since I’m unlikely to have much time for extensive GDD updating for a while, I’ll to strive to leave it in a state where other people can happily edit it without making a mess. That means adding links between relevant sections (so that when something is changed in one place, other locations where it’s mentioned can be easily found and adjusted accordingly), which I’ll try to get done within the next week.

Given that we’ve finally made some inroads into overhauling the reproduction system and health/combat system, I think we should probably update the GDD to reflect everything that’s been decided on since the last revision. By the time we get to post-0.3.2 outreach, it needs to be as up to date as possible. That includes the simulation page if anyone’s willing to write it. I’ll try and fit in the new game mechanics, as well as updating everywhere else in the GDD they appear.

Given people have still had different assumptions about mechanics, I’d like to ask everyone who considers themselves a contributor to microbe stage discussions to read it. All of it, if you can. Take your time of course, but it would really really help to get everybody on the same page, even if that page is subject to change.

Ignore the parts that are out of date after recent discussion (health, reproduction, the new CPA lists), but otherwise if there’s something drastically different to what you thought was planned, SPEAK NOW (or whenever you finish reading). This is meant to be a self-coherent concept, though still subject to change. If somebody starts a discussion on a topic that doesn’t fit what’s decided upon for the GDD, and they don’t explain what else should be changed to make it fit, I retain the right to be slightly annoyed.

There are a couple things I could point out, but they’ve all been discussed and there hasn’t been any clear consensus on them (such as pilli). I think the GDD at the moment is very good at showing what is planned, but I don’t expect that it should be a final and perfect copy. The final details should be rediscussed when we will be implementing a particular feature—we might run into some problems while actually coding it or it might turn out to detract from experience.

I vote @moopli or @tjwhale for writing the simulations page.

I’ve got quite limited Thrive time at the moment so I’m going to spend it working on the prototypes rather than writing up stuff.

Another issue is that the CPA system might change quite radically when it gets built. We need to play around with it and see what works and what doesn’t to get it to feel right. After that I think it’s worth writing some nice wiki pages so people who are playing can understand it. However before that process has happened I’m not sure it’s worth it. Or at least anyone else is very welcome to write as much as they like.


I had a suggestion for relatively minor changes based on some research I’ve done in the past 24 hours. Just some name changes to make certain game elements more representative of real life.

  1. Agent secretors could/should be called extrusomes. The way @TheCreator described the agent secretors to me is a sphere that buds off smaller spheres filled with agents the player wishes to secrete. In real life, extrusomes are membrane-bound structures that discharge their contents outside the cell. Some notable examples are mucocysts (mucus secretors basically, which are in the GDD) and nematocysts (think tiny venomous harpoon guns. Could be a way to introduce the pilus to the player?).

  2. Rename vacuoles to vesicles. Vesicles is more broad and allows for some specialization early on. Evolve (upgrade) certain vesicles into lysosomes to digest other microbes, secretory vesicles to help with waste removal, transport vesicles to help with protein synthesis, etc etc.

  3. Change “Mitochondria” to “Mitochondrion” in the editor, as mitochondrion is the singular of mitochondria.

I was also curious about how likely it’d be for entirely new game mechanics to be added at a later time.

1 Like

This is all just preliminary. We can change and chop bits just as much as before, but my intention is to give everyone a clear sense of exactly what the starting point of those changes is. And it won’t hurt to have a comprehensive (if ever-changing) concept when selling ourselves to prospective new developers.


Sure, we can fiddle around with the semantics. Extrusomes could work as long as the player knows what they are (most people know what chloroplasts are for instance, but the more obscure organelles might need explaining). Revamping vacuoles (as vesicles) is a possibility, but from your description that would introduce some new mechanics which we’ll have to review before adding. And yeah, it should definitely be mitochondrion in the editor.

1 Like

For the extrusome: That’s why we have tooltips!

For the vacuole(vesicle?): Yeah this conversation would be more relevant when the game gets to a point where players can actually upgrade their organelles.

Mitochondrion: Sweet!

I looked through the GDD again and noticed there was a suicide option? Is that still planned?

I have an idea for an alternative if you guys are interested. Most microorganisms, when conditions are unfavorable, have the ability to go into suspended animation as cysts.

In game, instead of committing suicide when your ATP levels drop, the player could be given the option to go into suspended animation until the compound concentration around the cell is enough to “start up” their cell again.

Would the player really want to wait around doing nothing waiting for their cells to gain enough compounds?

1 Like