@@GodzillaJawz It's not about having "more genes". That's just dumb and shows your limited knowledge about artificial life. It's about being more *biologically accurate*. The developer of the Bibites takes the idea of "survival of the fittest" to the extreme, in a way that resembles incorrect interpretations of natural selection such as social darwinism, whether it be on his tournaments or 100-hour series, it's all about competition and domination. In nature, there doesn't exist a "perfect" creature that can dominate any environment. Bibites is popular because it presents an incorrect depiction of how evolution works and how ecosystems form. They show it as a massive battle where species compete for dominance mercilessy. There don't exist ecological niches nor symbiotic relationships. Also, "at least the Bibites have digestion" it was meant to fix predation and even it failed even that 💀
@blidea9191 i wasnt downplaying this game at all it is very cool i was just saying the bibites is another one of the best ones, maybe i was downplaying it but i dont think that now
woah just as I thought that I've seen any evo sim you could find out there, this game happens to have popped up in my recommendations! This sim looks amazing, I really hope the algorithm picks this up in the near future so more ppl can see, imo, one of the more detailed evolution simulators that there is
I hate youtubes search systems! I search up evolution project content all the time and its always holding stuff back until its almost a year old, glad I got to watch this its amazing even if it took so long to be shown to me xD
This is all really cool! I've been sharing it around some other Alife groups to see what they think! Also, OH MY GOD THE TAXONOMY UI HOLY SHIT DUDE 50/10 It seems like whatever system you have going on is more conducive to predation than other Alife simulators, is it at all hardcoded? (also a discord for this would be wonderful)
Thanks for your kind words (and for spreading the word!) There's nothing hard-coded about predation or any behaviors. I very specifically designed the neural net inputs and outputs to be as "real" as possible -- each output neuron just triggers a single "muscle" like "turn left" or "bite." There are no outputs that lead to coded behaviors like "move toward object" or inputs that give it detailed information like "position of food."
@@urocyongames Thanks for letting me know! This is super interesting, in a somewhat similar game "The Bibites" we've been trying to crack predation for a while. While it has the means to evolve it has only ever happened twice/thrice (depending on who you ask) and yet in the Demo for NeuraQuarium I've already seen it evolve! I hope you keep working on this project, it's really fun as a game and learning tool :)
@@urocyongames How does the sight system work exactly then, are they given a coordinate and then can process that information however they want or do they have a direction input that they are able to know? Or might it be something completely different
Critters' visual input neurons are as follows: amount of red, green, or blue seen; the amount of movement detected; and then a neuron representing the angle to the most "significant" (as a function of size, proximity, and camouflage effects) object it sees, and finally a neuron indicating how "significant" that object is. The color inputs are weighted by significance as well, so a tiny bright green food pellet right in front of a Critter will stimulate the Green neuron more than one far away, or the same as a big beefy green Critter at a distance.
@@FireTurkey I read a suggestion about it, which involves introducing different types of plants. Predation almost never evolved in Bibites naturally, because digesting meat is simply not efficient enough to justify. A herbivore Bibite would be optimized for processing plant food, so meat will be incredibly energy intensive to digest, which cancels all of the benefits. The (theorized) reason behind predation evolving in real life is that there is more than one type of plant food. An organism can either specialize on one type of plant and extract a load of energy from one specific food source, or it can become a generalist and digest multiple plants decently well. These herbivorous generalists are thought to be the ancestors of predators. Because their digestive systems evolved to process a diverse diet, they could extract a lot of energy from meat from the moment they began eating it, simply by virtue of the inherently efficient digestion. The energy saved by the lucky digestion setup combined with meat's high energy content justified specializing in carnivory, and eventual atrophy of plant digestion to save more energy. I'm not entirely sure why hunting is a thing (in real life as well) since scavenging doesn't cost as much time and energy as hunting. Maybe it's something to do with the immune system and corpse-borne infections.
Amazing simulation ! I think that you should do simulations "ab initio" beginning from the chemical reactions let a self reproducing molecule arise, and from that let the complexity of metabolism evolve, as it was what happened on earth. Even simple cellular automata can display interesting behavior, but it stops at a certain very early stage of organization.
I really love this, I do have a few questions and might buy it. Some of my questions are: Do you name all of the different species and manually make that ancestor tree / declare speciation or is that an automatic process? Also is all reproduction Asexual or is there times when two members of the same species can combine genes to create offspring? My last question would be how long did you work on this, and do you have anyone helping with the development of this awesome simulation?
Thanks for your questions! The naming and tree building is all automatic. New subspecies split off a set number of generations down (if two Critters don't share any common ancestors after looking back X steps in their parentage, they are considered separate species.) Critters can and do reproduce with partners as well as by themselves. Partnered reproduction costs less energy, so there is an incentive to find members of your own species! I started working on NeuraQuarium in early 2021 and released the first version of it at the end of March that year. You can check out all the devlogs here: urocyongames.itch.io/neuraquarium/devlog My awesome friend wrote the music and the code that cycles instruments in and out dynamically according to how big the population is. Other than that, just me!
@@urocyongames Wow that's awesome, I wish you great success on this simulation and will definitely follow along on its journey! Also it seems that you've made so many different ideas work in harmony with each other, like reproduction, environment, and even a species tree and naming system so well, as just 1 person developing the main part of the game is just incredible especially under such a small amount of time. I've seen simulations worked on for many years that don't seem nearly as polished or have as many features as yours. Hoping you will keep documenting your progress with videos similar to this one. You've definitely earned yourself another subscriber!
Predation is absolutely a "real" thing -- see this video for a real example ua-cam.com/video/7UUSLzqEvYc/v-deo.html That said, there are plenty of reasons to evolve either camouflage or "warning" coloration even if nobody's trying to eat you on purpose -- herbivores with big nasty teeth are just as deadly as carnivores and it doesn't really matter whether your killer finds you tasty or not, you're just as dead. So if you can look like a jagged rock (blue) or a harmless weed (red), it's usually to your benefit.
OH wow this is sick! I have so many questions: Do you use neuroevolution for this? If so, what are some good sources for learning about the approach you used? How did you make the name generator? How did you "evolve" their physical characteristics? What game engine did you use?
Hi! I'll try and answer each of your questions: Neuroevolution -- I hadn't heard the term before, but that's basically what's happening. Basically neural nets as fodder for a "genetic algorithm" -- lots of fancy words that basically boil down to "successful versions of this system get tweaked and tried again, versions that don't succeed are discarded." Since both the weights and topology of the Critter brains mutate between generations this is considered a TWEANN. Name generator: This is actually multiple processes working together! The "scientific" names are new, but individual Critters have had names since I started building NeuraQuarium. To get Critter names I basically just cobble Consonant-Vowel-Consonant combos together (with some adjustments like allowing certain double consonants, or ending with vowels.) The "scientific" name is the result of a whole lot of unnecessarily complex logic examining its colors, patterns, stats, and so on to determine what's "special" about it and applying a name from a mapping (so a fast critter species might get "RAPIDUS" and a white one might get "ALBUS".) Evolving physical characteristics: This is pretty simple. When a Critter reproduces, its offspring is a copy of them (or a mix of them and their partner,) but all of those stats have a chance to be mutated by some small amount. Things like color, skin pattern, temperature preference, speed, diet, etc. (The brain mutates too, but that's another topic.) Changing some numbers is simple; what really matters for evolution is that those stats have an effect on the Critter's survival and reproduction. They aren't just numbers, they all affect energy usage, visibility to other Critters, maturing time, etc. etc. Game Engine: NeuraQuarium was built in Unity, with code written in C#.
@@urocyongames Thanks so much for the quick replay and for going into so much detail! I am surprised you go this to work in Unity. I would think all separate networks would use too much compute resources. But yeah TWEANN is a form of neuroevolution (where you use evolutionary algorithms on neural networks). This is really interesting work. I have experience in reinforcement learning and machine learning, but not in genetic algorithms so this definitely inspired me to look more into them!
The neural nets do eat up a decent chunk of cycles, especially as populations get higher. Fortunately they are hard-limited to a set number of nodes and a maximum number of synapses, so each individual net is not that intense.
I recently bought Neuraquarium and am confused about something that keeps happening, I keep seeing my creatures randomly die, they will have an ideal temperature and good energy but suddenly just die. Is this just some kind of cardiac arrest lmao. Any responses would be awesome, thanks.
Critters do have a limited lifespan and will die of "old age," so cardiac arrest is probably not too far from the truth! This is most likely what's happening if it really seems to be out of nowhere. Another thing to consider is they take damage from touching walls, which you might not be expecting.
In terms of the taxonomy graph, it's simply when there are six generations of separation. Speciation (two critters being unable to interbreed) can be toggled to follow this rule, to depend on genetic similarity, or both.
@@huginn1879 Each critter knows all of its ancestors going back six generations -- its parents, and their parents, and so on. If lineage-based speciation is turned on, only critters who have at least one ancestor in common are considered the same species and can breed. The taxonomic subspecies naming system works slightly differently, though it uses the same data -- when a new critter is born and the first critter to share its species name can no longer be found in its ancestry data (i.e. six generations have passed) then it's considered a new subspecies (because it would not have been able to interbreed with the original specimen)
It's just a natural outcome of the simulation. Critters have a very basic "Diet" stat that determines how much energy they get from plants or meat as a percentage. So pure herbivores get 0 energy from meat (although they do get a Protein buff, which makes their energy last longer.) In terms of evolutionary pathways, there are a couple different things that help herbivores branch into omnivorous diets and behaviors that support it. Nuts are a big component of this -- they are VERY energy dense plant food sources but they must be bitten in order to eat them (as opposed to food nuggets which can be bitten to swallow them quickly but will also be absorbed slowly by just touching the mouth.) So a Critter that evolves biting behavior and high strength will be able to gain a lot of energy by eating nuts. From here it's a short trip to biting other things that happen to be close-by... Dead or injured critters also emit a blood scent which other Critters can detect. I haven't observed any Critters actively following their noses to scavenge but it's possible.
@@urocyongames Do you have a discord, cause I want you to meet cause me and another person are working heavily on what creates stable pred-prey systems.
I don't have a Discord server set up for NeuraQuarium, but you can email me if you want to chat. I'll give you a brief summary of my experience with predation here: "Predation" meaning opportunistic omnivores who eat everything they see is very common. "Stable predator-prey relationships" meaning separate species, one herbivorous and one carnivorous, is much harder to create and nearly impossible to maintain for any length of time. However, I have had it happen, ONCE. I detailed the experience in a devlog here: urocyongames.itch.io/neuraquarium/devlog/374603/acheiving-stable-predation-foxes-rabbits The TL;DR is that you need a very large and fast-breeding population of prey, and a SLOW-BREEDING population of predators so they don't immediately hunt their prey into extinction and then starve. In terms of evo-life sims, it can be quite difficult to achieve this setup if for no other reason than computational limits on level/population sizes.
@@urocyongames Reason we asked, is cause while me and buddy haven't done it in your alife sim, in the Bibites (a simliar alife sim) we have created one and your assessment that it takes a lot of population and sim area is not wrong, the normal defaults in that game are like 300 bibites and 5000 u areas. We have to have no cap and about 1300 bibites and 20k u in land to do the sims.
Yeah. My experience in NeuraQuarium is it takes a roughly 20x prey to support 1 predator. I have a series of autosaves of about 12 generations in a stable relationship like this but even then, we're talking 12-24 predators supported by 250-450 prey, and that's just a couple of bad mutations or runs of ill luck away from the predators dying out. And those species did not evolve together, which is a whole other bag of worms. EDIT: After thinking about it a bit I did some experimenting and have found that the Pred/Prey ratio is also really dependent on the reproductive rate. If you can balance it right, you can get away with a much smaller ratio -- the balance then becomes that if the predators get too numerous, they eat all the prey in their local area and there's a small die-off, but the faster-breeding prey can then move in and repopulate. "Predation" can mean many things to many people. The simplest to achieve is "hunter behavior" -- Critters identify targets, chase them, and eat them all the time, through natural evolution. Reaching "obligate carnivore" status in terms of diet happens, but is usually a lot less beneficial than just being an omnivore. Makes the species more brittle and more likely to die off. Creating a stable predator-prey pairing can be done, for a while at least, with some creative breeding in separate sims and then merging them or spawning them manually. Separate species evolving in the same simulation and settling into a stable pairing, naturally, without some kind of human intervention, a whole different animal.
NeuraQuarium is available on itch.io, including a free (albeit a few versions behind) demo. urocyongames.itch.io/neuraquarium Steam release is perpetually planned "after I finish this next big feature..." Kinda like fusion power being "30 years away" for the last 60 years.
This is one of the best artificial life demo's I've seen, looking forwards to more development.
Thanks! Feel free to post clips of your own Critters, too!
@@GodzillaJawz It's not about having "more genes". That's just dumb and shows your limited knowledge about artificial life. It's about being more *biologically accurate*.
The developer of the Bibites takes the idea of "survival of the fittest" to the extreme, in a way that resembles incorrect interpretations of natural selection such as social darwinism, whether it be on his tournaments or 100-hour series, it's all about competition and domination. In nature, there doesn't exist a "perfect" creature that can dominate any environment.
Bibites is popular because it presents an incorrect depiction of how evolution works and how ecosystems form. They show it as a massive battle where species compete for dominance mercilessy. There don't exist ecological niches nor symbiotic relationships.
Also, "at least the Bibites have digestion" it was meant to fix predation and even it failed even that 💀
@blidea9191 i wasnt downplaying this game at all it is very cool i was just saying the bibites is another one of the best ones, maybe i was downplaying it but i dont think that now
You have made the best designed taxonomy UI system I’ve ever seen from an evosim project , neat!
A built-in one is a very nice touch. It's a surprisingly simple use of graph data structure
Amazing ! I’m really interested in how you designed the creatures and especially how their brains work. Could you make a video about that ?
There would be a lot to cover, but I might try and do an "anatomy of a Critter" vid some time!
I go into some more detail about the neural nets in this video: ua-cam.com/video/7UUSLzqEvYc/v-deo.html
woah just as I thought that I've seen any evo sim you could find out there, this game happens to have popped up in my recommendations! This sim looks amazing, I really hope the algorithm picks this up in the near future so more ppl can see, imo, one of the more detailed evolution simulators that there is
This is SICK! You should get more subs for this, looking forward to it.
this is such a cool project. I love to see these kinds of projects and this is definetly a standout project
Your interface and visual aids is phenomenal! The best I’ve ever seen in a simulated evolution example.
I hate youtubes search systems! I search up evolution project content all the time and its always holding stuff back until its almost a year old, glad I got to watch this its amazing even if it took so long to be shown to me xD
bought this within seconds of starting the video. The UI is really nice! Lots of great features. Haven't even played yet, but I'm a huge fan.
Cmon UA-cam this channel needs more attention
woah what the shit?? this is so good and ive never heard of this before
Amazingly done! I can't believe your channel hasn't blown up yet!
Goodness! How long have you been fermenting in your jar?!
This wine, it's excellent!
Amazing!!! You could write the mutation rate into the genes, and make it mutable itself. That way no more manual adjustment ist necessary anymore:)
What a great project! Easy sub from me!
Great video! One error I spotted was that in the "chapters", the one labeled "Day 15" should be "Hour 15"
This is all really cool! I've been sharing it around some other Alife groups to see what they think!
Also, OH MY GOD THE TAXONOMY UI HOLY SHIT DUDE 50/10
It seems like whatever system you have going on is more conducive to predation than other Alife simulators, is it at all hardcoded?
(also a discord for this would be wonderful)
Thanks for your kind words (and for spreading the word!) There's nothing hard-coded about predation or any behaviors. I very specifically designed the neural net inputs and outputs to be as "real" as possible -- each output neuron just triggers a single "muscle" like "turn left" or "bite." There are no outputs that lead to coded behaviors like "move toward object" or inputs that give it detailed information like "position of food."
@@urocyongames Thanks for letting me know!
This is super interesting, in a somewhat similar game "The Bibites" we've been trying to crack predation for a while. While it has the means to evolve it has only ever happened twice/thrice (depending on who you ask) and yet in the Demo for NeuraQuarium I've already seen it evolve!
I hope you keep working on this project, it's really fun as a game and learning tool :)
@@urocyongames How does the sight system work exactly then, are they given a coordinate and then can process that information however they want or do they have a direction input that they are able to know? Or might it be something completely different
Critters' visual input neurons are as follows: amount of red, green, or blue seen; the amount of movement detected; and then a neuron representing the angle to the most "significant" (as a function of size, proximity, and camouflage effects) object it sees, and finally a neuron indicating how "significant" that object is. The color inputs are weighted by significance as well, so a tiny bright green food pellet right in front of a Critter will stimulate the Green neuron more than one far away, or the same as a big beefy green Critter at a distance.
@@FireTurkey I read a suggestion about it, which involves introducing different types of plants. Predation almost never evolved in Bibites naturally, because digesting meat is simply not efficient enough to justify. A herbivore Bibite would be optimized for processing plant food, so meat will be incredibly energy intensive to digest, which cancels all of the benefits. The (theorized) reason behind predation evolving in real life is that there is more than one type of plant food. An organism can either specialize on one type of plant and extract a load of energy from one specific food source, or it can become a generalist and digest multiple plants decently well.
These herbivorous generalists are thought to be the ancestors of predators. Because their digestive systems evolved to process a diverse diet, they could extract a lot of energy from meat from the moment they began eating it, simply by virtue of the inherently efficient digestion. The energy saved by the lucky digestion setup combined with meat's high energy content justified specializing in carnivory, and eventual atrophy of plant digestion to save more energy.
I'm not entirely sure why hunting is a thing (in real life as well) since scavenging doesn't cost as much time and energy as hunting. Maybe it's something to do with the immune system and corpse-borne infections.
I wish I could make an artificial life simulation, it’s just I don’t know how to even go about doing it. Nice video.
It would be cool to one day see the concepts of this and “thelifeengine” merge
Amazing simulation ! I think that you should do simulations "ab initio" beginning from the chemical reactions let a self reproducing molecule arise, and from that let the complexity of metabolism evolve, as it was what happened on earth. Even simple cellular automata can display interesting behavior, but it stops at a certain very early stage of organization.
This is bibite level stuff ; maybe even better!
I really love this, I do have a few questions and might buy it. Some of my questions are: Do you name all of the different species and manually make that ancestor tree / declare speciation or is that an automatic process? Also is all reproduction Asexual or is there times when two members of the same species can combine genes to create offspring? My last question would be how long did you work on this, and do you have anyone helping with the development of this awesome simulation?
Thanks for your questions! The naming and tree building is all automatic. New subspecies split off a set number of generations down (if two Critters don't share any common ancestors after looking back X steps in their parentage, they are considered separate species.)
Critters can and do reproduce with partners as well as by themselves. Partnered reproduction costs less energy, so there is an incentive to find members of your own species!
I started working on NeuraQuarium in early 2021 and released the first version of it at the end of March that year. You can check out all the devlogs here: urocyongames.itch.io/neuraquarium/devlog
My awesome friend wrote the music and the code that cycles instruments in and out dynamically according to how big the population is. Other than that, just me!
@@urocyongames Wow that's awesome, I wish you great success on this simulation and will definitely follow along on its journey! Also it seems that you've made so many different ideas work in harmony with each other, like reproduction, environment, and even a species tree and naming system so well, as just 1 person developing the main part of the game is just incredible especially under such a small amount of time. I've seen simulations worked on for many years that don't seem nearly as polished or have as many features as yours. Hoping you will keep documenting your progress with videos similar to this one. You've definitely earned yourself another subscriber!
Just started the game and allready have some preditors :)
How do they evolve colour specific when there is no real predation?
Predation is absolutely a "real" thing -- see this video for a real example ua-cam.com/video/7UUSLzqEvYc/v-deo.html
That said, there are plenty of reasons to evolve either camouflage or "warning" coloration even if nobody's trying to eat you on purpose -- herbivores with big nasty teeth are just as deadly as carnivores and it doesn't really matter whether your killer finds you tasty or not, you're just as dead. So if you can look like a jagged rock (blue) or a harmless weed (red), it's usually to your benefit.
everyone reading this will all be dead by the time this genre innovates on it's own premise.
Very cool!
OH wow this is sick! I have so many questions:
Do you use neuroevolution for this? If so, what are some good sources for learning about the approach you used?
How did you make the name generator?
How did you "evolve" their physical characteristics?
What game engine did you use?
Hi! I'll try and answer each of your questions:
Neuroevolution -- I hadn't heard the term before, but that's basically what's happening. Basically neural nets as fodder for a "genetic algorithm" -- lots of fancy words that basically boil down to "successful versions of this system get tweaked and tried again, versions that don't succeed are discarded." Since both the weights and topology of the Critter brains mutate between generations this is considered a TWEANN.
Name generator: This is actually multiple processes working together! The "scientific" names are new, but individual Critters have had names since I started building NeuraQuarium. To get Critter names I basically just cobble Consonant-Vowel-Consonant combos together (with some adjustments like allowing certain double consonants, or ending with vowels.) The "scientific" name is the result of a whole lot of unnecessarily complex logic examining its colors, patterns, stats, and so on to determine what's "special" about it and applying a name from a mapping (so a fast critter species might get "RAPIDUS" and a white one might get "ALBUS".)
Evolving physical characteristics: This is pretty simple. When a Critter reproduces, its offspring is a copy of them (or a mix of them and their partner,) but all of those stats have a chance to be mutated by some small amount. Things like color, skin pattern, temperature preference, speed, diet, etc. (The brain mutates too, but that's another topic.) Changing some numbers is simple; what really matters for evolution is that those stats have an effect on the Critter's survival and reproduction. They aren't just numbers, they all affect energy usage, visibility to other Critters, maturing time, etc. etc.
Game Engine: NeuraQuarium was built in Unity, with code written in C#.
@@urocyongames Thanks so much for the quick replay and for going into so much detail! I am surprised you go this to work in Unity. I would think all separate networks would use too much compute resources. But yeah TWEANN is a form of neuroevolution (where you use evolutionary algorithms on neural networks).
This is really interesting work. I have experience in reinforcement learning and machine learning, but not in genetic algorithms so this definitely inspired me to look more into them!
The neural nets do eat up a decent chunk of cycles, especially as populations get higher. Fortunately they are hard-limited to a set number of nodes and a maximum number of synapses, so each individual net is not that intense.
@@urocyongames fantastic, do you use Dots or OOP?
My code's all object-oriented. Some portions such as critter vision and heat/gas diffusion are multi-threaded through the Unity Jobs system.
i see this game for the first time, that is what i was dreaming about
Interesting, still very hard to understand who belongs to who even if you say the name, as it very complicated sounding. Though love the content.
Woah this is really great
I recently bought Neuraquarium and am confused about something that keeps happening, I keep seeing my creatures randomly die, they will have an ideal temperature and good energy but suddenly just die. Is this just some kind of cardiac arrest lmao. Any responses would be awesome, thanks.
Critters do have a limited lifespan and will die of "old age," so cardiac arrest is probably not too far from the truth! This is most likely what's happening if it really seems to be out of nowhere. Another thing to consider is they take damage from touching walls, which you might not be expecting.
Nicely done!
What if the artificial lifeforms could evolve to modify their own probability of mutation?
How do you decide when a mutation becomes a new subspicies?
In terms of the taxonomy graph, it's simply when there are six generations of separation. Speciation (two critters being unable to interbreed) can be toggled to follow this rule, to depend on genetic similarity, or both.
@@urocyongames Interesting. "six generations of separation" -> Is it correct to assume that not every offpsring is considered a separation?
@@huginn1879 Each critter knows all of its ancestors going back six generations -- its parents, and their parents, and so on. If lineage-based speciation is turned on, only critters who have at least one ancestor in common are considered the same species and can breed.
The taxonomic subspecies naming system works slightly differently, though it uses the same data -- when a new critter is born and the first critter to share its species name can no longer be found in its ancestry data (i.e. six generations have passed) then it's considered a new subspecies (because it would not have been able to interbreed with the original specimen)
How does your predation work?
It's just a natural outcome of the simulation. Critters have a very basic "Diet" stat that determines how much energy they get from plants or meat as a percentage. So pure herbivores get 0 energy from meat (although they do get a Protein buff, which makes their energy last longer.)
In terms of evolutionary pathways, there are a couple different things that help herbivores branch into omnivorous diets and behaviors that support it. Nuts are a big component of this -- they are VERY energy dense plant food sources but they must be bitten in order to eat them (as opposed to food nuggets which can be bitten to swallow them quickly but will also be absorbed slowly by just touching the mouth.) So a Critter that evolves biting behavior and high strength will be able to gain a lot of energy by eating nuts. From here it's a short trip to biting other things that happen to be close-by...
Dead or injured critters also emit a blood scent which other Critters can detect. I haven't observed any Critters actively following their noses to scavenge but it's possible.
@@urocyongames Do you have a discord, cause I want you to meet cause me and another person are working heavily on what creates stable pred-prey systems.
I don't have a Discord server set up for NeuraQuarium, but you can email me if you want to chat. I'll give you a brief summary of my experience with predation here:
"Predation" meaning opportunistic omnivores who eat everything they see is very common. "Stable predator-prey relationships" meaning separate species, one herbivorous and one carnivorous, is much harder to create and nearly impossible to maintain for any length of time. However, I have had it happen, ONCE. I detailed the experience in a devlog here: urocyongames.itch.io/neuraquarium/devlog/374603/acheiving-stable-predation-foxes-rabbits
The TL;DR is that you need a very large and fast-breeding population of prey, and a SLOW-BREEDING population of predators so they don't immediately hunt their prey into extinction and then starve. In terms of evo-life sims, it can be quite difficult to achieve this setup if for no other reason than computational limits on level/population sizes.
@@urocyongames Reason we asked, is cause while me and buddy haven't done it in your alife sim, in the Bibites (a simliar alife sim) we have created one and your assessment that it takes a lot of population and sim area is not wrong, the normal defaults in that game are like 300 bibites and 5000 u areas. We have to have no cap and about 1300 bibites and 20k u in land to do the sims.
Yeah. My experience in NeuraQuarium is it takes a roughly 20x prey to support 1 predator. I have a series of autosaves of about 12 generations in a stable relationship like this but even then, we're talking 12-24 predators supported by 250-450 prey, and that's just a couple of bad mutations or runs of ill luck away from the predators dying out. And those species did not evolve together, which is a whole other bag of worms.
EDIT: After thinking about it a bit I did some experimenting and have found that the Pred/Prey ratio is also really dependent on the reproductive rate. If you can balance it right, you can get away with a much smaller ratio -- the balance then becomes that if the predators get too numerous, they eat all the prey in their local area and there's a small die-off, but the faster-breeding prey can then move in and repopulate.
"Predation" can mean many things to many people. The simplest to achieve is "hunter behavior" -- Critters identify targets, chase them, and eat them all the time, through natural evolution.
Reaching "obligate carnivore" status in terms of diet happens, but is usually a lot less beneficial than just being an omnivore. Makes the species more brittle and more likely to die off.
Creating a stable predator-prey pairing can be done, for a while at least, with some creative breeding in separate sims and then merging them or spawning them manually.
Separate species evolving in the same simulation and settling into a stable pairing, naturally, without some kind of human intervention, a whole different animal.
auto clicker with the one that generates heat is fun
you should look at "Thrive"
When will this launch on Steam? I'd love to be able to play this.
NeuraQuarium is available on itch.io, including a free (albeit a few versions behind) demo.
urocyongames.itch.io/neuraquarium
Steam release is perpetually planned "after I finish this next big feature..." Kinda like fusion power being "30 years away" for the last 60 years.
When will iOS version come? Please evolve your game! Very cool! I will buy when evolved
Fantastic, please post sources on github
Why did you let the simulation run so slowly? Why not speed it up, and only slow it down when you want to look at it?
That's exactly what I did! Any references to days/hours are in simulated time, not real-time.
Its alot like the game creatures :) and i have a copy its good :)
AMAZING
Nice
I was here to be subscriber 423
well you love your latinisms
Amogus sugomasus
3D