Plans for the Next Version

Alright I’m back.

It’s been a fairly small launch, but that was expected: it’s not like I advertised the game much beforehand. I’m kind of relieved: I can go back to working on it without worrying too much. I think I’m more comfortable with the idea of slow progressive growth than with rapid expansion.

And going back to working on it is exactly what I’m doing. I think I have a better grasp now of the games failings, and I’m looking into rectifying them, starting with the most obvious: the static vegetation.

The 0.5.0 release will contain an extensive vegetation overhaul: very little of the current system will survive into it’s next incarnation. This will provide three things that (in my opinion) the game needs: an gentler survival curve, an evolving environment, and a prettier world.

The gentler survival curve will be achieved by a new behavior: grazing. Rather than having to find a tree in order to get enough energy to reproduce, a creature will be able to gain energy by grazing from the ground. This comes at a cost, of course: it’s slow and won’t provide much energy, and will also reduce the local fertility. But it will mean that even creatures that don’t find a food source will still have a chance, and the deaths of most creatures will become a bit more interesting than the current average of “too stupid to look for food”.

Speaking of which, I’m well aware that the AI needs a lot of work. Unlike the other planned overhauls, I’m planning on slowly improving the AI bit-by-bit, since it touches upon all the other features I’m looking at and working on. The creatures will get smarter and more responsive, but to what extent I don’t yet know…

The evolving environment will be a load of fun to program. I should probably clarify: by “evolving enviroment” I don’t mean “evolving plants”. Sadly, with CPU usage what it is, I can’t spare the extra CPU cycles to include completely mutable plants. It’s on the wishlist, and anything I can make mutable I will, but there a hundred-fold more plants than there are creatures, so plant evolution will have to be limited to simple things like size and colour.

But even though the individual plants will be mostly static, the forests most definately won’t be. As temperature and humidity across the map changes (being influenced by creatures, rainfall, climate change, and even the player), the various biomes will drift about, and the plants native to each biome will colonise the newly exposed area’s, allowing (and/or forcing) the creatures to migrate and adapt.

With luck, this won’t be the only large upgrade to make it into the next version either. I’d like to get a top-down map in place so you can watch the simulation in a more symbolic/iconic format, and make a number of changes to the gene system so we can start Playing God ala. Jurrasic Park. I’d also like an ‘export’/’import’ creature function as soon as possible, so you can save your favorite creatures and bring them in as introduced species to devastate the local ecosystem.

Also, I’ve been playing Dwarf Fortress. It’s not exactly a bastion of good game design (in many cases it’s a perfect example of what not to do), but it has a terrifying amount of depth and complexity, and has given me a load of new idea’s for how I can improve Species (those familiar with Dwarf Fortress are either cheering or weeping right now).

I realise Species in it’s current alpha state is a very shallow game: interesting, but not for long, and with no actual gameplay behind it. All of that is coming: an idea I’ve had to start with is to allow you to highlight creatures by specific statistics, which could be used to tell you which creatures to kill and which to aid in order to promote certain forms of artifical development.

I won’t set a date for the next update just yet, but I won’t go more than 3 months between updates. I’ll also try to implement anything suggested in the forums or the wiki.

“For anyone trying to apply artificial selection in the current version, the best way is to bottle-neck the population around a mutant creature (ie. kill everything except that creature), then feed the population up again until you find another creature you like the looks of…”

Advertisements
  1. #1 by MarcoDBAA on June 13, 2012 - 5:01 pm

    I liked the concept.

    Gameplay is good, but you should be able to watch, if you like to. Everything should work on its own. Evolution needs no god xD. A real creature editor (genetic laboratory in the form of a creature editor) would be cool.

    Vegetation should promote speciation. Real species cannot feed on everything or are at least specialized to feed on certain plants

    Grazing is ok, but if you aren´t able to move around to find food, you should disappear from this world. A land of milk and honey will create dull and lazy creatures.

    Artificial Intelligence in an evolution simulation should evolve in the true sense of the word. I don´t mean tree hugging of course. Only serious options should be possible.

    Limbs shouldn´t be allowed to get reed-thin.

    Modding could be interesting too. I have done that already for other evolution simulations. I am awaiting instructions though.

    Greetings Marco

  2. #2 by Adam Benton on June 13, 2012 - 7:28 pm

    So the idiocy I was complaining to you about (not bothering to find trees) wasn’t a glitch but genetic? Interesting. Also, grazing should be a behaviour only available to vegetarians since gathering energy from such low quality plant material typically requires many adaptations to make it worthwhile.

  3. #3 by ququasar on June 14, 2012 - 9:21 am

    Gameplay is good, but you should be able to watch, if you like to. Everything should work on its own. Evolution needs no god xD.

    I totally agree with this: the player in Species will always be an agent who interacts with the simulation, not a god who defines it.

    A real creature editor (genetic laboratory in the form of a creature editor) would be cool.

    I want to avoid any similarities to Spore because Species is a very different game, so there won’t be a graphical creature editor. Player-motivated creature editing will take two forms: artificial selection (kill off the creatures with small legs, promote the ones with huge legs) and direct genetic manipulation, where you make changes and then create a creature and compare it to it’s parents to see what the gene you just played about with was.

    Actually that gives me a really good idea: what if the game records this information for the future, so you could modify that gene again without having to remember where in the genecode it was? Rather than just earning some abstract “experience points” or something, the game rewards you directly by showing you what you actually discovered and making it available for the future.

    Vegetation should promote speciation. Real species cannot feed on everything or are at least specialized to feed on certain plants.

    Agreed! I’ll even append that to say that everything in the environment should promote speciation: the more niches available in the ecosystem, the more interesting the game will become.

    Grazing is ok, but if you aren´t able to move around to find food, you should disappear from this world. A land of milk and honey will create dull and lazy creatures.

    Surprisingly, this isn’t true. Making things easier on the lower-end of the creatures means more creatures, which means more competition and interaction with each other. Try comparing a grass world to a heavily forested world: the forest world creatures will have high selection pressure for things like mating and combat, while the grass world creatures… well, they probably won’t survive (which is why I want to implement grazing 🙂 ), but if they did they’d have high selection pressure for energy efficiency and storage.

    Artificial Intelligence in an evolution simulation should evolve in the true sense of the word. I don´t mean tree hugging of course. Only serious options should be possible.

    Well that’s where things get tricky: where do you draw the line? Many real-life animals are promiscuous enough to ‘try to mate’ with inanimate objects, or are curious enough to try out a variety of different actions with an unknown item.

    Hopefully later in the game, improvements in the physics and AI will allow for a more detailed behavioral system: maybe even a mutable if/then/else tree, following the best precepts of Genetic Programming. At the moment though, it’s just impossible for them to survive without a fairly high level of intelligence: the ecosystem simply doesn’t have enough ambient sources of energy for brainless creatures.

    Limbs shouldn´t be allowed to get reed-thin.

    I completely disagree: I think they should be allowed to get that thin, but if they do they should then be unable to hold up their own weight, with obvious results.

    Modding could be interesting too. I have done that already for other evolution simulations. I am awaiting instructions though.

    I’ll get those to you as soon as possible: next post will be all about the modding. I was also planning on uploading a template file with a few examples in it, but I haven’t made that yet.

  4. #4 by ququasar on June 14, 2012 - 9:40 am

    So the idiocy I was complaining to you about (not bothering to find trees) wasn’t a glitch but genetic? Interesting.

    Err… semi-genetic. Part of it is influenced heavily by the creatures Attention Span, a genetic variable which determines how long a creature is willing to do one thing before doing something else. There is also a genetic variable called “Interest in Tree’s”, which I’ve been meaning to fold into the Carnivore/Herbivore distinction. Both of these can have a strong influence on a creature’s ability to survive.

    Mainly though, they’re just dumb.

    My biggest plan for the AI is to remove Attention Span and Idle entirely so creatures will always be doing something, and have them make decisions in response to actions and the ending of actions.

    For example, a creature eating will make a decision the moment they’re full, a creature attacking will make a decision in responce to losing more than 50% of their health, and so on. Add a bunch of behavioral signifiers which have a strong effect on their decisions (“Hungry”, “Wounded”, “Horny”), and it should be quite the intelligence boost for them.

    Also, grazing should be a behaviour only available to vegetarians since gathering energy from such low quality plant material typically requires many adaptations to make it worthwhile.

    Agreed: I’m thinking of changing the environmental diet spectrum, so grazing is most efficient for herbivores but omnivores can barely graze at all, and the tree’s are reduced in number and changed to be “fruit trees”, making them significantly more efficient for omnivores, and letting carnivores eat them. Currently there’s not much incentive for a creature to be omnivorous or carnivorous, and the ‘point source’ tree’s lend themselves well to the high-risk high-reward strategy of carnivores (while not being very efficient as as a low-risk strategy, as the alpha demonstrates). The tree’s are sort of like carnivore-lite.

  5. #5 by Adam Benton on June 14, 2012 - 11:09 am

    I believe the key advantage to carnivory is that you can get a rather large amount of energy with relatively little digestive effort. Perhaps a sliding scale, like the “fondness for trees” or “aggression” already in the game is what is needed. At the one end you can process vegetation very efficiently (although ultimately less efficiently than a carnivore can process meat) and grazing is a viable strategy. At the other end you have meat efficiency. Again, this is more efficient than a peak herbivore, but the strategy is more risky since it involves hunting rather than just placing your face in a tree.

  6. #6 by Swufty on June 15, 2012 - 1:20 pm

    Dwarf Fortress? Definitely weeping ;D

  7. #7 by ququasar on June 15, 2012 - 1:32 pm

    Dwarf Fortress? Definitely weeping ;D

    Heh heh, it’s not quite as bad as that: I’ve probably got as many Things Not To Do from Dwarf Fortress as new idea’s. DF manages to be awesome and terrible at the same time. But I definately want to dig into the underlying physics a bit deeper, and I think there’s things I can borrow from DF’s rediculously complex and gory bodypart/damage system. It would be a lot of fun to get more of the “red” in “red and tooth and claw”.

    Tell you what though, if you ever catch me defining the ignition point of eyeball fluid, then you can kick me.

  8. #8 by Skylimit on July 5, 2012 - 7:53 am

    Hey,
    Just discovered you sim, and spent like the last 10 years for something like this. It’s not finished yet, but definitely like the direction you’re taking, and curious about your next release. Keep up the good work!

  9. #9 by ququasar on July 5, 2012 - 8:51 am

    Thanks! Sounds like what happened to me too. When I first played Primordial Life, I thought there just had to be something-like-this-but-bigger already out there. The idea that it was actually a unique concept never even occured to me, so I went looking for evolution games.

    Most of the things I found were either abstract, micro-scale evo sims (Biogenesis being my current favorite, though I’ll always have a soft spot for Primordial Life since it was my first), or directed evolution games, where you just followed the evolutionary path(s) laid out by the developers. Credit goes to Creatures for being something else entirely, but it was simulating pre-built intelligent life, so evolution in it didn’t go much faster than evolution in real life. As I understand it, Noble Ape is similar (though I haven’t had a chance to play it).

    Anyway, the end result was realising that my assumption was wrong, and this simply hadn’t been done before. That was when I started designing what would eventually become Species. (and I’m still not entirely convinced it hasn’t been done before. Part of me believes the next link someone sends me will be to a genuine obscure first-principles macroscopic evolution simulator that beat me to it. But it’s looking more and more like that’s just paranoia: I really am the first! Yay!)

    (EDIT) Fixed HTML fail.

  10. #10 by MarcoDBAA on July 5, 2012 - 5:56 pm

    Biogenesis is my favorite too ( your simulation might take the top spot, after one or two updates ). That´s why i created and uploaded a mod for it. Biogenesis has the GPL license, so i haven´t done something bad. http://sourceforge.net/projects/biogenesiscolor/

    Hope it is ok for you, if i link it here (already using the link as homepage), and hope that it doesn´t slow your work, although i would feel flattered xD. I am not a real programmer, only wanted to change some details, but it was becoming somewhat bigger. Extensive enough to upload it.

    And yes, you are the first to create a large scale macroscopic evolution simulation.
    Scientists don´t need a “real” graphical world, it is too much work for hobby programmers , and doesn´t pay for big game studios.You have to be crazy… 😉

  11. #11 by ninjakitty on May 27, 2014 - 3:22 am

    when are you planning to make a mac version, after most of the stuff is finished, or earlier?

  12. #12 by ninjakitty on May 31, 2014 - 12:39 am

    are you gonna add herding behavior, where basically if one creature sees another creature go somewhere there’s a ten percent chance of it following it, if there’s two animals going there’s a twenty percent chance, and eventually if there’s ten creatures following a creature any creature that sees them going will have a hundred percent chance of following, so large migrating groups can form perhaps if conditions get bad in one area creatures will migrate to a different area

  13. #13 by ququasar on June 2, 2014 - 7:29 am

    I do have plans to migrate to the Monogame framework at some point in the future, which would allow us to make a Mac version, but I need to buy a newer version of Visual Studio and upgrade to XNA4.0 first. And then I would need to buy or borrow a Mac to test it on! So Mac support will likely come later rather than sooner.

    As for herding behaviour, that will likely come with some of the other behavioral upgrades we’re planning: specifically, altruism. A creature with an altrustic personality will occasionally ‘help’ other creatures, either by joining them on a hunt, defending them against attack or simply looking for food with them. Of course, it remains to be seen if altruism would be a viable survival strategy in the game, but an altruistic species would automatically herd together as they help each other with various tasks.

  1. Material System | Species Development Blog

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: