Where to next? Species 0.10.0

If you’d asked me 6 months ago, I’d have said Species 0.10.0 would be my opportunity to work on dietary chemistry. Biochemistry is the lynchpin for several major features, and I’m quite keen to get it working. Plus who doesn’t like complicated flowcharts filled with biochemical technobabble?

But then I rethought that plan…

If you’d asked me 3 months ago, I’d have said Species 0.10.0 would be my chance to add swimming and marine ecology. Swimming is more a sideline feature, but it’s big and flashy and it’s a noticeable lack in the game right now, since we have the water plane but not swimming. Plus I’d already started work on the underwater shaders, and was dropping hints about it in the forums.

But then I rethought that plan…

And now?

But I do have some idea’s.

I’ve been playtesting the game myself, trying to work out what’s missing. And there is definitely something missing. For a game with an infinite possibility space of procedurally-generated, evolving creatures and a whole variety of different biomes, it’s weirdly predictable and samey. That’s actually a word. Apparently it’s British.

It’s a problem that plagues procedurally-generated games in general. Humans are irritatingly great at pattern recognition: they can spot when entities share animations and behaviors. Different games get around this problem in different ways, to varying levels of success… but that’s a topic for another post.

In Species 0.9.0, this manifests itself in quite a few places. There’s the ‘template’ bodyplan: once you spot that all the creatures are vertebrate hexopods, their appearance becomes a lot less interesting. The basic behavior of “walk around slurping up food sources and spawning babies, then die” is also problematic. If you’ve seen P. Specium’s behavior, the behavior of many evolved creatures won’t be new to you. But how do we address something like this?

The obvious answer is “variety”, but that could refer to any number of different things.

Behavioral Variety

Like in real life, creatures really should engage in far more behaviours than just seeking food: not just social and territorial behaviours, but also various actions related to temperature regulation, seeking other important resources (water and oxygen, most notably) and finding a place to sleep. Even the generic “eat food source” behaviour could become a heck of a lot more interesting simply with the addition of different types of food sources: there’s a gulf of difference between catching insects in your beak, picking grubs out of the ground, and grazing on foliage.


Currently, there’s only really one survival strategy: “walk around slurping up food sources as efficiently as possible”. In addition to more behavioural variety, we need the environment to provide a wider variety of niches for creature’s to occupy in order to encouraging those varying behaviours to co-exist in the same world.


At the moment it can be hard to identify variety even in cases where it genuinely exists. Creatures can die from overheating, freezing, starvation, old age, injury, harmful mutations, or player-instigated robomurder, but what players generally *see* is a creature lying down and exploding. Without knowing the context of why and how it died, all death get’s lumped into the same boring “ran out of health” category.

I don’t know which of these is most vital, but my thoughts at the moment are that any of these three… features? concepts? themes? … are more important to the current build of the game than Biochemistry or Swimming would be.

I’m actually fairly far off the beaten track here: unlike previous updates, where I had completed most of the design work well in advance, 0.10.0’s development will be more reactive and involve a fair bit of improvisation. I don’t have a checklist of features to complete this time around.

At the moment I’m focusing my work on one of the items under “Behavioral Variation” (can you guess which one?), but I suspect I’ll drift between a variety of features as the update comes together. I also completed a number of smaller tasks while I was rethinking my evil plans, and I know how much you all love aimless ramblings about the technical minutiae of development, so stay tuned!


  1. #1 by Inarius on June 10, 2016 - 4:51 pm

    That sounds good !
    Lack of variety and visible patterns are indeed a plague in procedural games…

  2. #2 by Sergio González (@SergioGonzR) on June 19, 2016 - 8:31 pm

    Great ideas here, swimming and marine creatures would be also some kind of biodiversity. At his moment there is some lack of ecosystem, I can’t see any kind of symbiosis or a clear pattern of hunter/prey where population limit itself (I’m looking for some logistic function https://en.wikipedia.org/wiki/Lotka%E2%80%93Volterra_equations).

    Speaking of that, are you planing to add some graphic with the different populations and evolution in time (like the ones you can see in wikipedia)

    And last, is there any ETA for this update?

    Great work.

  3. #3 by ASamaritakis on June 19, 2016 - 11:42 pm

    Maybe you should make a low-poly style version, it is easier to handle because you don’t have to do the textures and stuff, and overal you can work pretty much a lot faster 😉 Have a look at the game Astronomer for example.. Also low-poly.

  4. #4 by Goncalo on June 30, 2016 - 12:31 am

    I think that different water courses should have different components. It would be amazing to have fresh water (rivers, lakes).
    Each course of water should have tides with water speed (ideal to create natural barriers to certain species as mountains do in solid ground), different water pH, different nitrite values, oxygen pressure. All of this could contribute to diversify the species.

  5. #5 by UndertaleNotUndertail on June 30, 2016 - 3:39 am

    How about you make plants evolve too?

  6. #6 by Silver_King01 on June 30, 2016 - 1:53 pm

    I think you should add water creatures and air (if you can) and special occurrences what I mean when I say “special” occurrences is for example. A volcanoe(s) erupting in a certain(or random) time. Another example would be a meteor crashing down on earth that impacts a chunk of area which becomes a radiated zone where it will have an increased mutation effect on them. IF you can please also add water and air creatures in this or later versions. It is a MUST to have these types of creatures in the game. I know this is long but thank you for even listening. P.S even if you ignore this good job on the game i’ll try to keep up with the news ;p

  7. #7 by Gamaran Sepudomyn on July 21, 2016 - 2:22 am

    Filter feeding (or, more difficultly but more importantly parasitism) would be a nice alternative feeding mechanism to have.

  8. #8 by Gabriel on July 29, 2016 - 12:15 pm

    I like you game Qu, keep going on!

  9. #9 by mscottveach on August 9, 2016 - 8:14 am

    I don’t know if this will be helpful or not but the one game that I am aware of that does a remarkable job of solving this problem is Dwarf Fortress. It might be worth studying through this lens and trying to see why it seems to work so well there and as you say fail in so many other PCGs.

  10. #10 by Michael Aguilar on August 9, 2016 - 9:09 am

    Is this game still alive?

  11. #11 by Adesh Maharjan on August 27, 2016 - 2:18 am

    GO QUASAR! Exiting things, being brought up here. Make ’em a reality!

  12. #12 by Creature on September 27, 2016 - 1:22 pm

    This is a unique program. This is that is great idea, but difficult. I’m glad that now will increase the diversity of the behavior of the creatures. It is unfortunate that now to experiment with a large number of populations is difficult, because when creatures more than 400 my computer runs very slow

  13. #13 by jabber texting on October 15, 2016 - 2:25 am

    I am buying a new macbook pro. Here’s hoping there is a wine bottle option for XNA that will play nicely

  14. #14 by Shura Kuhn on October 24, 2016 - 12:36 am

    Hi! I think the hardest part of evolution games is this: whatever efforts you make to “improve” the lacking details, you are almost always comparing your planet with earth. That is exactly what is meant by “trying to work out what is missing”. The ultimate result will be a live wikipedia. But that is good enough. At least for educational purposes. To make people interested in life science and evolution. To move kids from violent games to games like this.
    I think the most elaborate evolution game would be like this: you give all the initial conditions in great detail, and see how they evolve, from inorganic molecules to organic, from ocean to land, from stupid to intelligent. Players erupt an ocean vent, and life begins. They set the sun’s spectrum differently, and plant leaves evolve blue. They can change all the variables. And when life evolves with a different set of variables from our planet, there’re gonna be a lot of new things to observe, and these probably happen now far apart from us in a different galaxy.
    But these are hard to achieve now, since we have only little understanding of nature. And there are a lot of details to take into account.
    I am planning to work on a life simulation game, about plant behavior. When in darkness, they grow their stems strenuously so as to get rid of it as soon as possible. That’s easy. When a player takes control of a plant and faces the same situation, he can think up the same strategy. But there are many other kinds that have not been fully discovered, and I will rely on human intelligence to take place of mother nature’s. Maybe this is one way of creating something new, something more than just modelling after what we already know.
    (Sorry for grammatical errors. I’m not a native speaker)

  15. #15 by Someonesomewhere on November 7, 2016 - 3:15 pm

    I think what would add to what is missing is what you’ve stated, the animals interacting with the world. Part of why people love to watch and study animals is seeing their behavior, social patterns, and their ways of slurping up food.

    After all, adding the water branch would be neat, but it would still have the “Same-y” feeling after the novelty wears off. But watching as herds group together and packs pick off loners, and seeing animals with little versions of themselves following, or even watching as animals hide or seek out shelter would reduce the “Same-y” effect.

    The problem I see would be how the AI is handled, but I’m confident you’ll figure it out.

    Keep up the great work!

  16. #16 by jeroen on November 11, 2016 - 11:48 pm

    i hope the game will do well

  17. #17 by Kalem on November 20, 2016 - 12:13 pm

    Hey, um…….. is either this game or are you dead?
    It’s been over 5 months since the last update; I’m getting worried.

  18. #18 by Martin on November 30, 2016 - 11:14 pm

    Hello. I am potential buyer. But i am worried that game is not in development anymore.
    Could you assure us that you are not dead or dropped that game development ?

  19. #19 by Penguin on December 12, 2016 - 10:55 am

    Maybe you could include a simplified version/option that displays only the important stats and in more basic terms? It can be a bit overwhelming having a screen full of stats when you only know what half of them do…

  20. #20 by talgie22890 on January 3, 2017 - 7:48 am

    I think it’d be cool if I had the exact same environment 10,000 times and 10,000 times its different from each other. That’s variety and it could be small differences to large differences. Part of that is not just the creatures forming, behaving differently but also communication, their love or hate of light, water(more than what’s necessary), etc. Small things that altogether add up to big difference.

  21. #21 by talgie22890 on January 3, 2017 - 7:56 am

    And to make i worse, and I’d hate to compare to spore, but the idea of these creatures forming a pack(if applicable to that creature) and later a tribe is a great idea.

  22. #22 by dranorter on May 6, 2017 - 7:54 am

    Hello! I realize I’m commenting on an older blog post but I totally thought I had written a comment at the time.

    This problem of procedural content feeling “samey” is interesting, and there has been considerable discussion of it elsewhere. I can’t seem to find the links which I ought to include… the first would be a discussion which took place during a NaNoGenMo a year or three ago. NaNoGenMo, ie, National Novel Generation Month, is this thing where people write code which writes a novel, during November. There’s a tendency for any novel-length generated text to start feeling samey very quickly. Generally you can read a page or less and feel like you know what the whole thing is going to be like. People trying to combat this directly have gotten content which remains interesting for a few pages. Specifically, they manage this by using a large number of templates; so they’re basically getting the variety by actually writing content.

    The discussion in question linked to a similar thread over on ProcJam, with people talking about the overall problem of samey content. Someone actually made a tool to help fight sameyness in auto-generated levels etc. The idea was that sometimes you can write code to actually quantify what feels samey about your output; and then what this tool does is generate and rate hundreds of examples and create a nice 2D visualization of how much variety you have.

    Of course, Species is in a different category; the variety one hopes for here ought to come from evolution, not just from random generation. But a lot of the visual variation of the creatures doesn’t connect in any way to their success, so in that respect they’re “just” procedurally generated.

    Nonetheless, variation gained through evolution is the more important aim. Now, I haven’t looked in detail at how Species works, but clearly it’s based around trying to make various life-strategies viable, right? Like, you want both carnivores and herbivores to be viable niches which these guys could occupy. You also want huddling behavior to be viable, you want specialization to different food sources to be viable, etc. So it seems like a major risk is that the game ends up with just a handful of distinct strategies, with evolution within each niche reducing to a short genetic walk to the optimum.

    When presented with problems of this sort, my usual solution is to take a Steve Grand sort of attitude. (Surely you’ve heard of Steve Grand’s Creatures games? They’re more focused on the brains, but still, a major a-life landmark.) Steve Grand tends to talk about figuring out what sort of system to create, such that the desired behaviour emerges on its own; as opposed to programming in the behaviour.

    Obviously Species has some of that going on. Glancing through recent posts I see that, for example, you found existing avoidance of cold was probably a good enough way to encourage avoidance of (cold) water. But it doesn’t evolve, right? You don’t end up with mutations where they seek the cold (and die quickly because of it). So I guess what I’m proposing, purely as an example, would be behaviour genes which aren’t initially attuned to the organism’s needs; the evolution would take care of that part anyway.

    Better examples and arguments can be found in Steve Grand’s book “Creation”. In the case of Species itself, I think a lot of complexity would be unlocked by the hoped-for plant evolution feature. But imagine if plants could evolve into animals and vice-versa!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: