I Made an Ecosystem Simulation in Unity. They Evolved. (devlog)
HTML-код
- Опубликовано: 9 июн 2024
- Hey everyone! In this devlog we are making an ecosystem simulation in Unity. There are three main habitats for which we are going to create a simple AI and see if they can coexist together.
▶️ Support me On Patreon / vav97
▶️Join us On Discord / discord
▶️More videos to Watch
I Made a Game For Blind People • Can I Make a Shooter G...
Adding Tower Upgrades - Tower Defense (Devlog) • Making My Dream Tower ...
I Made my Dream Tower Defense Game! (Devlog 01) • I Made my Dream Tower ...
I Made my First Game in Unity | HARDEST GAME EVER?! • I Made my First Game i...
I Made a 3D shooter... But You Can't Shoot?! • I Made a 3D shooter......
I made GOD OF WAR but it’s a VR Game • I made GOD OF WAR but ...
I Turned MARIO Into a VR Game! • I Turned MARIO Into a ...
The idea of the video was inpired by @SebastianLague 's video on a similar topic in June 2019 "Coding Adventure: Simulating an Ecosystem " You can watch his version here - • Coding Adventure: Simu...
More videos related to this topic
Creating An Ecosystem Simulation Game In 6 Months • Creating An Ecosystem ...
Simulating an Ecosystem • Simulating an Ecosystem
My equipment (If you buy through these links it will cost you the same, but I'll get a small commission. That's a Big PP move. Consider it.)
Processor: Intel(R) Core(TM) i9-9900K
Memory: 32GB
Graphics: intel(R) UHD Graphics 630
Keyboard: Keychron K2 v2 amzn.to/3nKZHhk
Mouse: Logitech MX Master 3 amzn.to/41c2vlA
Microphone: Behringer B-2 PRO amzn.to/3pfBS1G
Headset: Meta Quest 2 amzn.to/42zmzz4
#devlog #indiedev #ecosystem #GameDevelopment #unity #GameDev #simulation #simulationgames #AI 
Развлечения
![Latto - Sunday Service (feat. Megan Thee Stallion & Flo Milli) [Remix] (Official Video)](http://i.ytimg.com/vi/9gL0F1xIt2I/mqdefault.jpg)
Check out my new video -> ruclips.net/video/4k08oBH5c0E/видео.html I Made a Sniping Game for BLIND people
You made the front deer legs wrong, that is not how they bend at all. You gave the deer human legs.....
👀
@@jar9415hey were able to reach the moon with their neck and that what is bothering you ?
@@Itzmeeeeeat Your lack of grammar...
@@jar9415 my bad I'm still learning English, it not being my native language I do make a few mistakes.
All attributes need a tradeoff, else every agent will just max out all the relevant stats.
More speed = more muscles = more mass to maintain = needs more food to survive.
More neck = visible from further away
etc.
That would make the simulation much more interesting.
Yes, we can already predict what's gonna happen which defeats the point of a simulation. It's supposed to model things that are harder to simply calculate by hand.
I think a good tradeoff for the trees would be to make how often they are eaten from be the determiner for how far the tree's offspring can spawn. Then couple that with a crowding mechanic, so trees are incentivized to keep their height in a range reachable only by some deer but not all or else they and their offspring die from overcrowding.
To build on this, while this could be more complicated to program, maybe the coyotes could have different “killing” strength. The strongest ones could potentially take down a deer in one attack but they tend to be slower while the smaller ones would be faster but potentially weaker and take longer to take them down
exactly.
to give an example for the trees. if it's too high, it needs more food and deers will focus more on them since they are more visible.
it would also need some population control for trees. if an area of trees is too dense, trees can't get enough nutrients and/or light.
one important factor would be to take into account the age of the deers. if a female becomes too old, she will be more interested in accepting a male so she can have children. after a certain age, females reach menopause and males don't ask them again to copulate.
coyotes should also be able to die of old age.
both, deer and coyotes should gain wisdom with time but lose other traits. less stamina/speed/perception but more wisdom/cunning.
These are great suggestions! Thanks
The game looks really cool, but I feel like the trees have an unfair advantage against deers. Give the trees swords😤
Make a 20% chance that a tree has poison
Seems like an idea for the next episode 😁
Or some kinds of spikes.
That will weird
nah, give them grenade machine guns
"I gave them memory, but they wouldnt leave the lakes"
aaaaand you just explained why every city ever was erected near a watersource.
If you reimplement this feature, but give hunters the same ability (to remember where they last got their pray, while deer remember the last time they saw a deer die) that might lead to some really fun interactions!
Yeah because staying around water sources/ especially food sources is what deer do when they have no reason to go anywhere from lack of predators. They will just stay in one spot and eat the entire area bare if there isn’t the threat of predators to keep them moving. So it was pretty realistic they just stayed in one spot while they didn’t have predators lol
Damn that's a valid point, if hunters have memory too the prey will evolve to avoid the water sources unless necessary
@@endercake3403 yeah the wolves should have water memory too!
I would love to see this with multiple other species! Imagine other predators competing with the coyotes for deer, or other consumers competing with the deer for vegetation. This could be incredibly dynamic. Keep up the awesome work!
That would actually be so fun! Thanks a lot for the comment!
@@Vav97 Maybe also leave the carcasses as food and add scavengers, animals that only eat dead animals, creating a new type of interaction.
@@asdads3948yeah, it could also allow for some pack structure in some animals as they can actually share the food
@@asdads3948 the mushrooms will also eat this corpses, causing them to decay faster. It’ll force the scavengers to be as keen-eyed and fast as possible, otherwise the bodies will be gone by the time they reach it.
I was so excited to find out how you solved the predator problem... and then the video ended 😭. Keep going, loved this video!
Lol 🤣 I watched some of your videos when I was making the simulation! Love them!
Predator problem comes from lack of population mostly, your sim does a good job of co-evolution from a simliar stock, but it really seems for stage 1 preds to form (ones that just can barely hunt) you really need very dense pops, which is why your fish usually become preds but your land animals don't.
You guys rock! Because of your comments and activity on the video I got a notification that youtube will try to show the video to more people. Thank you and don’t stop ❤😁
Guess you cracked the algorithm with "they evolved" and "dream game" lol
@@revimfadli4666 well, 10k views is far from cracking the algorithm, but who knows 😁 maybe it will take off soon
And indeed it did it! I've just seen this one and then watched all your previous videos one by one right from the first one. I've been trying youtube for long years now (not only on the channel you can find under this account of mine) and I can sympathize with your hard work not earning too much attention early on. The sacred youtube algorithm needs some time to understand what your audience is so it can recommend it to the right people... and launch an ever-growing community IF you have a catalog of high-quality videos - and the man you have it. So... enjoy the ride, I'm totally sure that you will explode pretty soon. All those "you are insanely underrated" comments you got until now will become history! :) I'm so happy for you man. Enjoyed a LOT of your jokes and editing. I'd love to chat with you - but I don't have a 1000 subs, so I can't send you a friend request! :D :D
Btw. when people start to binge watching your videos like I've done, as far as I know, that is when RUclips just really becomes crazy about your content as you are increasing our session time on the platform by a lot. Keep it up, bro!
@@AndrisGameDev Man I saw your comment on reddit and since that I was trying to find your comment on yt but somehow I didn't get a notification. BUT I found it :DD I'm very happy to hear you liked the content on the channel! I'm available on discord all the time, join us if you wanna chat :) Indeed, it takes a lot of time and dedication to run a yt channel, but obviously I wouldn't be able to maintain it without the support from people like you, so thank you very much for taking your time and commenting! And good luck with your game dev journey as well :)
Glad you are evoluating in the youtube chain!
Great video! I have a potential solution to your giraffe problem. You could have it so that the deer eat the trees and poop out seeds, therefore forcing the trees to not grow too tall, or else they can't reproduce. I'd love to see more of this though.
Edit: You could also link the speed to the stamina, so that they'd have to sacrifice their stamina for speed or the opposite.
Thanks for the comment! True, the trees are kinda op now, as they need nothing to reproduce but time. Making the trade off is also another way to make it all more interesting. Thanks :)
Yeah that would be a good idea because then if the trees grow too tall they won’t be able to spread/ reproduce as easily so they’ll have to stay a bit smaller
Maybe consider making the trees adapt in different ways, like producing more fruit, spawning closer, etc. Also, trees could need nutrients to survive, like deer and coyote feces. Lastly, females should also consider the weaker males to mate with in case they don't have other options to reproduce. Very excited to see more of this!
Indeed, the trees ase overpowered now, I like the idea! Also, yes, if females could choose a weaker mate, they would not died 🤣 thanks for the comment!
I love this ecosystem sim concept, it would be super cool to see you explore it more.
Honestly, I think I'm done at least for now with this project. BUT I have other interesting projects coming :)
@@Vav97.
Yes ?
Look at Sebastian Lague's original version of this - also highly entertaining.
The necks are actually realistic. Giraffes evolved long necks because they only ate food from trees. Deer on the other hand normally eat grass, therefore don't need the long necks.
True, if they grew longer necks then it means they are the only animal apart from birds that can use it as a food source
dont hurt me like this, genuinely not sure if youre joking so someone will respond but giraffes evolved long necks for sexual selection. males fight using their necks. genuinely.
it was for both@@art-eroflore
@@art-eroflorefeel like it’s defo a strong mix of both
initially evolved for trees
females began to like neck boys
evolved further
Giraffes have always been giraffes. Since creation
So a big problem we have had in the bibites is that predators are very very hard to translplant, especially when they start out as a stage 2 pred (predators that can actively hunt). A few factors are necessary for survival: Co-Evolution, Large Pops, and ways for predators to store kills so they don't constantly have to find food.
The amount of detail and effort…. Respect!!
Thanks man!
Something I can think of to get the trees and deers to coexist better, would be that they require the nutrients from dead deers to grow, so the more deer they feed, the more trees exist. And then feedback loop
If you give deers food preferences and let them choose between trees, grass and bushes, you could have giraffes, cows/another grass eating animal (a good point to add a gene that changes the size of the animal depending on the food they choose) and still deers.
Repeat with more aspects of their lives and you'll have an environment that evolves a bunch of different species from the same root
That’s actually very interesting idea, didn’t think of that. I was thinking that as the deers grow their necks, make them only be able to mate with other deers with long necks. Connect that with your idea, and yeah, one species would evolve to multiple others. That would be a very interesting experiment!
@@Vav97 yeah, it's a powerful system with a lot of potential tbh. You can simplify it by making an evolution tree where you define what evolves into what and with what requirements, so it makes sense and you don't get overwhelmed by how to make a deer evolve into a mouse. With enough evolution decisions it would get there step by step
@@Vav97 You'd want a system where the chance for offspring is proportional to the difference between the genes of the parents. And open up the value a bit, 25% mutation in one generation is too much.
One idea for a future expansion on this little ecodystem is to add nocturnal creatures which only come out at night, so that way there's a larger difference between day and night, besides skydome changing
Wind could be an interesting factor. Depending on speed and direction it could boost seed spread for trees, or awareness for deer/wolves.
Maybe that'll make balancing things too tough though lol
True, I liked the idea. Thanks :)
Hey you uploaded again😄
İts nice to see you have been working on different projects in your time
Hey :) yep, spent a lot of time on this one. I hope you’ll like it!
@Vav97 we will see ig🤷♂️😊
This whole video is super interesting! I like that you charted the neck length and tree height so we can easily compare the measurements over time.
this was super fun to watch, good stuff bro
Thanks a lot! Very happy to hear you enjoyed the vid :)
This channel is going places. Never give up, please.
Nah , I’m here 🤣 don’t worry
Great video i would love to see a part two
Thanks a lot!
I’ve been waiting so long for something like this, all the other videos are so old, thank you thank you!!
I'm so happy to hear you liked the video )) Thank you!
9:31
VAV: trees survived, I guess that was a happy ending
Me: I guess that the strongest competitors was the trees all along.
Amazing video! 😁
Thank you :)
Bro, seriously talk, it’s 2:20am and I’m drowsy af but I just had to tell you, this is amazing. I mean it seriously. I myself am somewhat of a programmer(mostly web stuff) so I can only imagine how difficult all this was but you did it!
Not only that but you did what very few people are ever able to do, you made it entertaining!
I’m writing this comment to let you know, you’ve got a loyal subscriber. Keep this up and I look forward to your blow up. Cause it’s only a matter of time
Man thank you so much for such feedback! Honestly that is very inspiring :) I hope you'll like the rest of the videos! Welcome!
There is a lot of simulation videos , but yours is my favorite! Great job with the editing.
Here's an idea for the tall tree deer population problem.
Trees need deer to eat fruit to spread the seed to grow the trees, those that don't get deer to eat them, will have to contend with their offspring competing with them for light and nutrients.
Most plant species actually evolved to fit their life cycles to the nutritional needs of various animals to help them spread. So the trees aren't competing with the deer to avoid being eaten, they're competing with each other to be more tasty.
That's how it would be in real life, I think.
If you haven't already thought of this and someone else didn't mention it, of course.
I remember seeing a similiar simulation a few years ago, but I believe it was with wolves and rabbits. It ended similiar, where they'd both grow in population then one or both would suddenly dip
love the video! it was amazing even if it was not your normal type of video. you deserve more subs!
Glad to hear that! Indeed this one was not like the others, but I guess this style also works :)
It'd be interesting if you gave the trees fruit and a nutrient system. The nutrients would be used by the trees to grow their height and also regrow new leaves and fruit. Where trees cluster together, they would fight over limited nutrients and would struggle to take full advantage of their inherited traits. The fruits can be used to spread the trees, like how it works in the video, but they can also be eaten by the deer and the deer can poop the seeds out around the map growing trees even further off. I think this would give the trees and deer an interesting symbiotic relationship like much of nature has irl
Super cool and inspiring. By adding things on top of what is here I am sure this could become a very interesting simulator!
The Moral of the Story is that Men and Women need each other to survive!!!! WOAH!!
Thank you for this video, it made my day!)
I'm very happy to hear that! thank you!
Intesting project. I like the meme/vine sounds. Keep going.
This is awesome 🤩
You are an Insane developer. Nice work🔥✨
Thanks a lot 🙏🏻😁
add ability to "urinate" i guess - mark zones and make map. Deers "urinate" more at places where they eat and drink more. other deers can be attracted there - forming places they consider a better ones. coyotes smell deers urine and hunt more "urinating" there. Coyote's urine will scare off deers forcing them to find other places to feed and reproduce.
No need for individual memory. Collective memory in nature are represented in different liquids animals (and insects) leave behind.
Hello from France, i really like your vidéos. I'm watching a lot of Devlog type videos, and and I really liked this one, worked and interristing. and the other are funny too, so...I subscribe !
Hello!! I’m so happy to hear you like the videos )) welcome!
Nice presentation, I like it. Job well done in this video (I do not mean to sound like a critique or something I just like this one very much hahaha)
Thanks a lot :) No worries, I'm always open to hear what others think
amazing video man, love to see graphs and stats!
Thank you very much! I'm happy you liked them ))
This was amazing!
Thank you!
High quality video! Nice work
Thanks! Glad you liked)
Humanity really figured out a way to make *natural* selection artificial 💀
Lol 🤣
Awesome simulation! Also top notch video editing and voice over 😃
Thanks a lot :) glad you enjoyed the video!
This is awesome! I’d love to see the results of a few specific changes: each tree occasionally sprout seeds along the path of the deer that ate from it and takes longer to reach full height, any wolf involved in a hunt can eat from the same deer (to lower the 1:1 consumption ratio), and female deer have sex drive (giving the males choice that is much less selective, to provide slow wolf-food). This might solve the issues with both the trees, and the deer going extinct. If you also include a way for the wolves to occasionally cull the older ones, their population might balance out too! 😂👍
this is crazily high quality content for a channel your size. keep it up!
my size ? it's 100k just showing 3k for some reason ... wait for a month 😁. But seriously thank you very much! I'm very happy you liked the video!
first :D (looks epic)
indeed first :D
This is super cool. I was thinking of doing something similar in python for my alevel computer science course work. Awesome work!
Thanks a lot! good luck with your project :)
Wow, really good editing and video in general, will watch more😊
Thanks a lot! 🙂 happy to hear you liked the video!
7:12 even the short deers get friend zoned
Really freaking cool I absolutely love it and would like to see more of this
Thanks a lot!
Nice video I would love to see more of this kind with these dynamics
Thanks a lot :) I'm already working on the upcoming video, so expect it soon!
This is cool! But the trees should be using the dear to spread their seeds. The dear should be eating their fruit or seeds instead of the tree. I’d be really interested to see what that looked like too! Great work!
There should also be a cost to stuff like faster speeds or taller trees, or they will obviously all max out their stats and stay that way because why wouldn't they? Higher trees should be more likely to fall down and die or something like that, and taller necks and faster speeds should require more food.
The neck gene is great in put I'm so invested
What is the cost function here? What's to stop the gazzelles and cheetahs just maxing out their speed? In real life increased speed means increased food needs, and possible other costs such as reduced ability to hoard fat reserves. Likewise long necks have the risk of snapping too easily or having poor blood flow from the heart to the brain
MOST UNDERATED VID AND CHANNEL!!
I've been dreaming of doing a game like this some day. I love watching other people's attempts... makes me want to do it more and yet terrified to try it at the same time. 😂 Will you be continuing this project?
Honestly not as difficult as it may seem. Maybe i’ll make other simulations in future, but I guess this one is done for now :)
@@Vav97 ohhhh come on I was looking forward to the giraffes developing spider legs and scorpion tails!
@@CoreyFarr 🤣🤣 I won’t disappoint with the next project!
This was so interesting! Please more ecosystem simulation ❤
Thanks a lot! I'm very happy to hear you enjoyed the video!
Looking really nice. I've just stumbled upon your channel, and you seem pretty cool. Keep up the good work
Thanks a lot! welcome :)
stunning work.
Thanks a lot!
The amount of work... Keep it up! :D Btw youtube recommended me this video so it should pop soon! o.o
Thanks a lot! Fingers crossed 🤞😁
This is pretty awesome!
Hey man! thanks a lot :)
Can't wait!!!!
Pretty Nice I've had the idea for ages now but sadly lack machine learning skills great to see it happen!
Thanks :) It’s a simple behavior tree, not that complicated as you may think.
Amazing video ! the video reminds me of the graph where the deer number increase -> then the coyotes have more deer to eat so many are killed -> now there are more coyotes which can reproduce and feed their children however now they need more food to feed the higher coyote population but since so many coyote's can hunt many deer, now there is a small deer population -> coyote's numbers reduce since they can't find enough deer to feed the increased population -> deer numbers increase since less coyote's to hunt them, rinse and repeat.
It was a very well-made simulation, however it felt like there were no drawbacks and the ideal way for things to survive quickly became a maxing solution rather than minmax, and it kind of took away any opportunity for balance. I think having drawbacks to characteristics would be a massive improvement and lead to much more interesting results which would really let the natural selection / genetic algorithms shine!
You should add cannibalism to the wolves with a variable for "time spent without food" where they will start considering other wolves as food.
It would be interesting to see the end game after the deer are dead. I wonder if two different species would basically evolve or what would happen
Fantastic video! Interesting and funny xd
Thanks a lot :)
Very interesting, thank you.
Thank you for commenting :)
Great vid, deserves more exposure
Thanks :)
Thanks :)
What a chill devlog😅
Thanks :)
You should keep at this it's kind of fun and cool to see what happens
love the vids m8
Thanks a lot! :) glad you like them!
You should make it so that the trees contest each other for nutrients so the tree that the deers can reach will be eaten and they evolve into the different subspecies of tall and short trees, the deers would also have different subspecies of tall and short. And there should be both leg and neck length, neck length also influences vision radius, and leg length influences speed. there should also be a cone of vision, and the current circle should be a mix of vision, sound, and 6th sense.
i like it. I’d like to see animals communicating with one another, perhaps 3 calls each. and have them tied to inputs and outputs through genetics. And maybe there should be trade offs for the various stats;
trees have height and the downside could be a shorter life span (doesnt make realworld sense but would limit offspring)
deer have stamina, fov, speed, and neck height, downsides could be food required, water required, stamina used, and speed reduction respectively
coyotes could have similar downsides to deer.
Yes, the trade offs is a must to implement, cuz with this kind of logic, they will always evolve ALL their stats, or die out. Thanks for the comment :)
Ps,
Deer is both singular and plural.
"Deers" only works for multiple groups of different species of deer.
For ex: 10 antelope would be ten deer, but 5 antelope and 5 mule deer would be 10 deers.
Same applies for fish, sheep, etc.
this is really cool
this is amazing!
Thanks :)
I HAVE BEEN WANTING A GAME LIKE THIS FOREVER. I WANT ONE THAT HAS SO MANY MORE SPECIES, AND COMPLETED
Mine is not a game tho 🤣. What you are searching already excists. It’s called equilinox.
I love ecosystem simulations!
This is amazing.
Thank you ❤️😁
When you get here early, you already know it’s gonna be amazing
Damb that was pleasant to read 🤣 thanks a lot! I’m glad you’re enjoying the videos
this was a breathe of fresh air to see!
Cool 🤣 thanks!
I can think of two things that would be interesting to add!
The first is that trees could evolve some kind of defense against getting eaten. Like some trees and plants in the real world do they could maybe produce a very bad tasting bitter flavor that most herbivores don’t like! Or they could evolve some kind of other defense mechanism like thorns or poison.
The second idea is to go back to the memory thing you tried earlier but instead of it being something that everyone has it could be another genetic trait. Basically some of the dear and some of the coyotes will have better long term memory than others. To prevent them from just getting stuck near the water we could use some kind of discouragement. Ether make it so that they are more vulnerable if they are close to the water hole or (the more fun thing in my opinion) we could ad a new contender in this little evolutionary experiment! I’m thinking maybe alligators or crocodiles that live in the water and will ambush the animals that come to drink. So now the animals that go to the water will have to be extra careful so they don’t get eaten!
I would also like to see if it could be possible to make a completely self sufficient system eventually. What would it take for the ecosystem to be stabilized and not collapse after a week or two?!
Another idea for two more additions to the ecosystem is owls and mice! Owls live in trees and hunt mice at night (giving night an actual purpose for survival) and mice will live in underground tunnels and will only come out to find food or water!
I just saw a similar test on an older video. Nice concepts and very well executed as well. I will use something similar running for my Survival gamethat way players have a direct impact on the ecosystem.
Thanks for the comment :) good luck with your game!
nice video and editing
thanks a lot :)
Wow nice video bro
Thanks man :) glad you liked it
bro the deer ramming its head in the leafs 😂 I love it
Amazing explanation of evolution 🤔🤯
Thanks :))
it would be great to make a hunger level for predators. So that he does not chase food all the time. Like, the higher the satiety, the less chance that he will attack someone.
wait why does this not have more views?? its awesome!
Thanks a lot! 🙂 it started to get views actually 🎉
This video is sooo cool, should have more views
Thanks a lot! I’m glad you liked it! Check out my new video :)
This is missing 3 things:
1-insects to aid the normal/small trees (speeding up their reproducing factor or scaring deer off)
2-make grass a food for deer (less nutricional but its always everywhere)
3-give the females also a need to breed
This is really cool wish I could play it
Sounds like a really cool sim
Thanks :)
You should add groups of animals, so the deer are together and help each other, and the coyotes pack hunt lone deer.
damn, if they could add this evolving traits to npc as well
This is a really good concept! Would love to see multiple different predators hunting each other or hunting for the same food. Normally I would like this content but you make it very interesting and fun! Your gonna blow up one day. I can tell. And I’m gonna be here early. +1 Subscriber.
Yep, already got a suggestion like that and I like the idea! Thaks a lot for the sub :)
This is not a very realistic mechanic. There are a handful of scavengers, but most predators see other predators as a very high risk of injury and actively avoid each other.