Simulating the Evolution of Courage, Fear, and Intimidation
HTML-код
- Опубликовано: 15 фев 2024
- visit brilliant.org/EightLittleBears/ to try everything Brilliant has to offer-free-for a full 30 days. The first 200 of you will get 20% off Brilliant's annual premium subscription.
---
Patreon: / eightlittlebears
RUclips Membership: www.youtube.com/@EightLittleB...
Merch: store.eightlittlebears.com/
---
Credit for Photos at the 13:36 and 13:37 - @miniminuteman773
---
Evolution simulator! Simulating natural selection! Biomes! Artificial Life Simulation!
Hi everyone! I have added a new reward on Patreon which will give you access to the data files for this simulation (CSV format) and the unedited video footage. Feel free to check it out if you are interested! Patreon: www.patreon.com/EightLittleBears. As always, please only consider supporting me if you have the means to do so! Cheers!
hi
you should add fat reserves so that the deserts will stop being instant death. Animals can't eat in deserts but an animal with a fat reserve can move further between meals.
Ah yes, the camel strat. Might also add stomach water capacity
that could make some sort of migration sistem , like creature A in tile 1 coes trougth the dessert to tile 2 to survive the winter because creature B or something
I think that would require high food vs animals in the tile. As far stores can't happen if the animal isn't eating it's fill
He should also add deses
@@TopatTom what's a des?
on why herbivory seems to increase with courage, i think what's happening is that as much as high amounts of conflict creates opportunities for meat, it also decreases the amount of rivals for herbivores to compete against for food, meaning forage ends up in the long term a much more available food source after a fight.
That is an interesting point !
Tracks with the real world too, where often the most ruthlessly violent creatures are herbivores (like hippos).
@@tiagogarcia4900 The Aurochs, wild cattle that modern domestic cattle are descended from, was much bigger and fiercer than modern bulls.
At the same time, this model still doesn't make sense for herd animals. Perhaps it's the territory system. Maybe we need to simulate migratory animals?
@@EightLittleBearsanother important factor to keep in mind is the different type of tiles change plant availability so being kicked into a savanna is a death sentence for forest herbivores while predators can find new prey there
I wasn't expecting a miniminuteman reference, but I'm okay with it.
It's interesting that intimidation doesn't cost anything, there's a lot of animals in the wild that spend quite a bit of energy on intimidation, like growing frills or antlers to look bigger, or just being extremely loud. I don't know if this is becoming too complex but maybe high intimidation should penalize stealth (to represent the extra energy investment) and bigger creatures should get passive bonuses to intimidation?
the size is relly cool idea but i think intimidation is not from the creature but form the actions like (agresivity)
@@JoseGarcia-po2pi In that case, intimidation should cost energy (since it means the creature moves more).
its getting more interesting by the day
Thanks!
please change your pfp
@@vyshaakhp6542 stop whining
it reminds me of a dark past (not accepting any questions)@@Frau_Brotchen
Gotta love the Miniminuteman reference
Something that might be fun to do in retrospect is doing a PCA graph. PCA- Principle Component Analysis- basically smooshes variables together into a graph, and tells you 'this graph explains 65% of trends, this graph explains 17% of trends...' and so on. This is usually combined with particular groups- say, time bins- to show how trends are different between those groups, with a polygon around the extremities or an ellipse showing confidence for that group. Basically, it turns a lot of data (e.g a bunch of different stats on their own axes) into a single graph of general explanation (e.g 'creatures are generally explained by some combination of bravery, carnivory and a few other covarying traits on one axis, and grass-eatery, mobility, and a few other covarying traits on another', usually with a second and third graph to cover the vast majority of variation)!
That reference to milo was freaking hilarious
Reminds me of the hawk-dove game from game theory. It's an interesting scenario, because it results in a mix of aggression (hawks) and backing down (doves) in proportion to the resources at stake and the potential cost of injury. It's a neat case where no single strategy is dominant.
I feel like I’ve seen a video about this recently but can’t put my finger on it..
I wonder if intimidation would be more favored if getting into fights had more detrimental effects. What if taking damage could decrease movement speed, courage, or increase the need for food while they are healing. Maybe taking significant damage could cause permanent debuffs.
Yes, increasing a creatures consumption after a fight would be a good thing. Maybe even having a cap to the number of fights a creature can get into win or loss, before it decides to travel elsewhere. It be like HP + Courage divided by 2 and every 100 in that score is equal to 1 fight, then have it so every 100 turns the number resets back to 0 if they've had no fights since during that time.
Yeah I was going to add an “injured” status and some additional al fight logic, but I am thinking I will add that down the line along with venom and stuff (since those things are likely to share a lot of code). Thanks for the comments!
Nice. May your day be good!
❤️
I wonder where this all will go...
it will keep evolving until some sentient life forms emerge 😃😨
Nobody knows…
@@user-JLso dwarf fortress but with creatures-level ai?
I like the addition of territory sizes I think it makes things a little more dynamic. It made me think that maybe an additional variable could be added for preference for different environments which would affect how far they are willing to go into different environments/how adventurous they are. Anyways I very much like your stuff. These videos are fun thank you
Amazing video as always, it would be cool to see ways for creatures to evolve so that they could go into different temperature and moisture areas without the population quickly dying off. It seems like the creatures stay near the equator and quickly die off whenever moving away. Can’t wait to see what you add next!
Agree! That’s in the plan. In this sim there were a few brief moments where creatures survived in more northern and southern regions, but these were pretty short lived. Temperature adaption is on the todo list so hopefully that won’t take too long!
It seems, that the omnivores population drop resulted in more herbivores to be eaten by the carnivores.
I'de assume thate the omnivores were an important food source
As an ecological rule of thumb it is common to assume that the ratio of the (bio)mass of prey to hunter is 10:1 but can vary due to different factors. A 1:1 ratio is guaranteed to go distinct though
Interesting - I wonder how the ratio behaves with, for example creatures who are 90% plant based vs 50% (as opposed to binary hunter/prey)?
I feel like intimidation should have a cost. Perhaps to stealth. One of the reasons that aposematic coloration works is that it is a costly signal. It says "look at me! I don't need to hide. Whatever my defense is, it worked on everything else that saw me and I want you to know that."
Great comment. I had considered this and might add something similar in the future, but the thought was that this only describes one type of intimidation. Lots of animals have the ability so switch their intimidation on and off (for example by showing their teeth) - baboons are the best example I can think of, but really any carnivore could also do this.
It would be cool to see a map designed off of the real world. And see how things diversify throughout the continents. That obviously is a huge project tho
If temperature and the day / night cycle are still part of the simulation, they could interact with courage and intimation as well.
Having more courage in daylight, being more intimating at night.
Temperature could also effect this, likely acting as a debuff to courage when it's very hot or very cold, but in turn if the title is not their home title and is too hot / cold maybe it increases their their chance of traveling, and if the title they travel too is more suitable they get a big bonus to courage for a number of turns.
They "know" they can't go back as that title is inhospitable, but that knowledge doesn't need to be stored in memory if it's an automatic feature. This could also help with storytelling and weaving a narrative for future videos, if you choose to go that direction
Really cool ideas! I’d need to slow down the day/night cycle I think, but I was wanting to do that anyway (already slowed down the season cycle by quite a bit). This would also fit well into how I was thinking “memory” might work… with creatures high-value locations in their minds etc.
Hi, I saw you reading my old comment so hope you would see this one. I just found your channel and now after I’ve watched many more videos I have a lot I want to suggest. I would only elaborate further and give my ideas if you reply to the comment tho, I would like to hear some responses from you as the programmer as well
Anyway, first thing, I would say personally I would allow mutation in the creatures choice dynamic. Rather than letting a creature choose its action based entirely on its preferred meal type digestion wise and have to fight as a carnivore etc etc and reverse as herbivores, you could allow some creatures to choose their own path. Many “herbivores” for example would choose to eat decaying carcass even if their digestive system are that of a plant-eater, or even hunt small prey and avoid large ones. This could be an interesting dynamic and encourage the evolution of things like “wandering scavenger” which try to find meat and left if there’s none instead of attending unnecessary fights. This would also allow mesocarnivores to have a fallback plant food.
Secondly, feel like something is wrong with your berries. Often times Creatures eat leaves too much and barely anything else. Leaves are horribly inefficient and very few rainforest herbivores even prefer leaves. You made it too efficient of a food source is my best guess. Grass is a bit more efficient but it’s nothing compared to things like berries or seeds. Berries can sometimes be more efficient than meat even, and nectar definitely so, but with increasing rarity. Berries and seeds are often also important fallback food for omnivores and mesocarnivores, as they do not need a super specialized digestive system to digest.
Also, I’ve seen your ecosystem operates a lot more like ones that are a lot smaller than your simulation, this is possibly due to the semi-isolated nature of large parts of the rainforest which makes it act more like archipelagos than a continent. In very tiny islands predators tend to be driven extinct and omnivores dominate. Your creature in many ways also show more similarities with insects ecologically. Thus, from what I’ve seen with the imbalance here, it seems the ecosystem would benefit a lot from metamorphosis, but I’ll elaborate further if you reply
13:35 i love miniminuteman so much his content
Great project!
The division between different environments seems to be quite sharp/discrete. That probably makes it harder to creatures to explore and adapt to new environments even when there isn't anyone occupying that niche. Maybe they are stuck in a local optimal peak.
In nature we see biomes transitioning in a much smother/continuous way. Thats called an "ecotone" - a transition area where we can see characteristics of two or more different biomes in the same area. A simulation of this half-biome might solve that problem.
good work, keep up with your work
Thanks!
This is amazing, I'm surprised you don't have more subs❤❤❤
Thank you so much!!
This was not the update I was expecting. But welcomed regardless.
Your project keeps getting better and better. It's interesting to see how evolution changes with every modifications you do
❤️
I don't think courage and intimidation should be unrelated stats, as in the wild, intimidation is merely courage + ability to commit violence.
Rather, I think courage should be calculated as the maximum strength of opponent they are willing to take on, based on an aggregate 'combat power' as well as the opponent's own courage.
It’s interesting that you say this. I implemented something extremely similar in one of my previous sims (even down the aggregate stat being called “combat power”). For this sim I am planning on something similar, but I wanted a third “aggression” stat to interact as well. The reason is that (using black bears as an example) you get some creatures that are generally very passive or even timid (flee in most conflict scenarios), but in specific defensive scenarios (e.g. a mother defending cubs), they are extremely capable of committing violence. So the thought is that that build would have high courage but low aggression. A honey badger would be max courage and max aggression, etc… there’s some kinks to work out with the idea still, but basically overall I agree with you haha.
@@EightLittleBears Yeah I see that- with the black bear example though, that courage is relegated to specific situtions. Maybe it could be described as a kind of volatility? Like some creatures are consistantly aggressive or not, but others- like irl bears- can be more risky
@@tnttiger3079 interesting…
you should look into adding water as a food source. than you could get the cold regions populated as they are spares on plant life but rich in marine life
Yeah agreed. I am planning on rebuilding the plant system from the ground up, so this will be a feature (at some point)
I think it might be interesting to have intimidation/courage also depend on other stats. You'd expect a bigger stronger animals to be generally more intimidating to other animals. Or similarly, a starved animal to be more or less willing to risk a fight if they think they can get some food out of the exchange.
Just a question, but why does your grass abundance increase all of the sudden?
Love this
I feel like The Fall may have been a result of highly aggressive herbivores. Usually herbivores adopt herd tactics and other major repellents, or independents are small and flighty to focus simply on species survival as it searches for food, with territorial exceptions focusing on group defense or securing a location to nest. If the herbivores consistently picked fights it negates the benefits of stealth and size, and there was a probability that at some random point equilibrium would be broken as many herbivores get themselves killed in avoidable conflict.
That’s an interesting theory… I want to get hers behaviours into the simulation but still working on how to implement this in a way I’m happy with.. thanks for the comment!
@@EightLittleBears NP, may your PC weep for the glory of your generation.
this is fascinating
Thanks!
very nice concept, although conceptually, and I guess that might just be too complicated but, I'd like courage and intimidation to be situationally variable.
I.e., if a tile is *particularly* high value compared to the neighbouring tiles, the creature ought to show quite a bit higher courage and intimidation, right?
Whereas, if it's a crap tile, it might as well flee.
Not entirely sure how to actually model that though.
That is a very cool thought! Seems very doable as well… though it might get complicated given that different creatures are going to have different priorities.. still interesting though!
every time there’s a new video about the simulation I get giddy like a dam child
Have you considered adding in some way for environments to change? Simulating desertification or the growth of new forests and so on
Yeah environmental change is one of the big things I want to achieve this year. The first step is reworking the plant system so that we have evolving trees and grass and fungi, etc. once we have that, then more base changes (mineral levels, etc.), and natural disasters and stuff should be pretty interesting!
I did not think Milo and Eight would collide lol
Funnily I only discovered him pretty recently (in the last few months) - he’s currently my most binged creator though!
Very interesting
Please keep going. You motivate me so much to learn and someday be able to do as you do (I got as far as building a random map generator (without using the random functions, I made my own shit version!))
Thanks for the comment! And definitely keep going! I started out making my ow shit versions of stuff too! I think it’s the best way to learn 😃
@@EightLittleBears I would totally be in love even more if you could maybe explain how what works a little more. (Maybe I just missed a video of you and you did) but just seeing you solve some problems helped me so much (2 weeks ago I wasn’t even able to do hello world in python now I have a somewhat random number generation program (somehow the results tend to be a little bit smaller on average than they should be and yet also the really small and big numbers tend to show up less than they should in a real random generator but I can live with that)
i think maybe having your territory moved is advantageous in some scenarios
I think this has greatly improved the sim
Thanks!
This simulation is getting so complex that soon sentient life will be added.
What happens when a creature gets bullied into a water square? Even if they don't die immediately, it's certainly not something they can eat from. Courage might be high in those areas just because there's lots of water (dangerous tiles) in the area, where many good tiles have 4-5 possible water tiles they could be pushed into
Or if they can't move onto water tiles, it could be that they get bullied into other tiles that are also heavily populated (there not being many tiles to choose from) and bullied again repeatedly, if their courage is low, resulting in them just being bullied until they starve. But overall I think it's probably related to those particular areas having lots of chokepoints where land is surrounded by more than one water
They can’t move into water at all (even if bullied), but I think the observation about being trapped in high-population areas leading to constant bullying (or fighting) is a good one!
You should add the growing and receding of biomes
Eventually 😃
Hey, I love your videos. I'm making a evolution simulation with neural network and was wondering how you plotted all your data. I have searched a lot and also tried to code one myself(didn't work well😅). Thanks 🙏
The charts were made just using regular spreadsheet software for all but my first two videos (for which I coded a solution but it’s pretty clunky)
@@EightLittleBears so did you save the data in a json file or how and what software is it?
I save the data as a series of CSVs - the spreadsheet software I use is “numbers” (Mac version of excel) but, if you have excel, I would recommend that.
@@EightLittleBears Thank you it helped a lot☺🙏. Btw have you thought of implementing a neural network to each of them where each input has a influence of where they move or what they do(for example neat algorythm or a neural network :))
So the herbivores weren't overhunted?
you should change the color of the pink biome so its easier to understand at first glance
Could you make this public whenever it's finished?
The goal is to make it into a commercial game so it will be public at some point!
Heres a question i have, do you plan to add random events such as a tornado or earthquake? They could change the terrain, which would prevent the environment from entering a deadlock
Yeah natural disasters are in the plan, but I think they will probably be best following a rework of the plant system (so plants can evolve) since it will involve changing the environment. Thanks for the comment!
Agreed, they're certainly fun, but other parts definitely need more work still
2160p60 4k. first time i saw that setting
Thought I’d test it!
I'm curious what the impact of a risk of infection from injury would be on this data; would it make it make fights risky enough to result in different strategies?
I think it probably would, or even just injury of any sort. Thanks for the comment!
how do you visualize the simulations? and can you do a tutorial?
I use a few different tools. I run/record the simulation in unity. It outputs a bunch of data (mostly in csv format) and I then use a few different tools to analyse that data. Most of the charts just use standard spreadsheets software (any will work), and more complex visualisations (such as the hex bubbles) are custom built in Unity.
question: Does size currently affect intimidation? actually what does Size currently affect in terms of stats, because your simulations so far have had a tendency towards Mega-fauna (or maybe they are just more noticable?)
Yes! I forgot to include that, but I essentially intimation power is a combination of size and intimidation rating!
I feel dumb with all These Genius comments, and This video is Amazing, As always, Underrated
Hey! Everyone on the internet has different experiences. I feel dumb every day… it’s all good. Just remember the internet isn’t real. Everyone seems smarter and better than they actually are 😜. Thanks for the support!
@@EightLittleBears Thanks Bro, And no Problem your Videos Are awesome.
Seems odd that the choices are "scared off before battle" or "fight to the death". Why no option to get into a fight, realise it's going badly, and try to run away, which AIUI is an incredibly common event in the wild?
We essentially have a simplified version of this, it’s just that “it’s going badly so I’ll flee” is lumped together with being scared off (since being scared off during battle is effectively the same as being scared off prior to battle). That said, in the future I’ll be adding an injury system, so we can get more of what you are talking about!
good night or day! :D
Thanks 😃
be carful, you could get sued for how underated you are
This simulation is so cool how can I do stuff like this but for... free...?
I use Unity, which is free as long as you aren’t making like $200k per year from your project
It would be cool if Courage affected the territory statistic.
Interesting thought.. it possibly might. If creatures have a small territory stat I in a high-population area, then courage should (in theory) become more valuable.
i have a theory that i think could be true.
1high western courage ovnibores are alwayas in figths that makes them efectibly carnivores.
2they totaly obliterate all the hervivores
3they die because they have so much courage that they die instead of eating plants and when carnivores eat the ovnivores, they also die.
4in a butterfly effect that destabilices the savana biome cousing it to die.
5a domino effect makes the east creatures intimidation unuseful because they dont have to figth territory because the savana is emty
5 the equilivrium brokes and then they eventualy go estint.
(i hope you can undertand it ,i am spanish and my english may be bad)
That’s an interesting theory, and certainly a possibility!
Whats the pink region?
Desert
Add breeding animations
EightLittleBears is primer but better
Let’s not get ahead of ourselves 😅
I watched the part at two minutes explaining the system several times. Listening and watching your graphic was wrong based on what you said. I believe the roll needed to be on creature B's intimidation. So it means I think it means if creature a wins there is no round two and creature a wins. And if creature b wins both time there is no fight, b wins. But if both creatures win courage then no one was scared off and there is a fight. If some one wins intimidation which in the example looked like the bigger stat box they win without fighting. Okay just yeah.
Thanks for the comment! I think the way I worded it is right but it’s confusing for two reasons:
- because I refer to “when the bully loses” and then point to the victim’s courage as winning. In reality the bully losing and the victim winning is the same thing.
- I refer to “creature 1” and “creature 2” but the graphic says “creature A” and “creature B”.
Basically the rule is:
- if the bully wins round 1, the victim flees
- if the bully loses round 1, the victim becomes the new bully
- if the new bully wins round 2, the original bully flees
- if the new bully loses, then that means courage won twice, no one is backing down, so fight.
@@EightLittleBears I did understand it but you rip through it fast and the visuals don't aid as much. Just feedback. Thought the overall video was great, I don't think I'm subscribed yet but I also think I've only seen a scattering of like three of your videos. I'll get there. Seems like quality content. It's at least fascinating to observe and partake.
I wished people would use "conspiracy theory" properly.
To put simply: It's a theory about a conspiracy. I know. It's at real difficult concept.
A conspiracy is "An agreement to perform together an illegal, wrongful, or subversive act."
Hi friend! That is one (I believe, US legal) definition, but just a note that there are lots of others from other jurisdictions or from non-legal authorities which are all correct. For example, the Cambridge Dictionary defines “Conspiracy Theory” as: “a belief that an event or situation is the result of a secret plan made by powerful people”. 😃
@@EightLittleBears Yes. A group of powerful people. Who are supposed to be *Conspiring* to create the event. A conspiracy.
The definition provided is too restrictive, requiring specifically "powerful" people. A conspiracy can be made up of multitudes of less-than-powerful people... which is why they would *conspire* together to bring about events that they couldn't do themselves. (Though powerful individuals might conspire to bring about the events, while limiting negative repercussions.)
Without that group of people, that you theorize is attempting to bring about an event, you don't have a theory of a conspiracy.
The definition is what it is 🤷♂️ (one of many), if you feel it’s too restrictive you might need to take it up with Cambridge - we can start a petition! With regard the group of people.. exactly! A group of powerful people are conspiring to influence the outcome of or simulation! We need to stop them!
@@EightLittleBears OK, you are taking it to mean "some nebulous group of powerful people, somewhere out there, doing something that I will in no way theorize about"?
Workable enough use of Conspiracy Theory. Have fun.
@@SangoProductions213 - tbh I was just having a bit of fun with it. Sorry if I came across as rude or otherwise. It wasn’t my intent ❤️
I have a non zero chance of winning against a bear I’ll have you know
I guess that’s true… it could rip your arm off and, try to eat it but swallow slightly wrong and then choke to death?
hi
intimidation may be bad for predators tbh.
Yeah I think you are right. I am planning to add resource stealing at some point (I.e. the brown bear strategy), which will (probably) make it more valuable.
Oops wrong video
5th
Re: zero percent chance against a black bear?
Any human adult has a higher absolute bite force than any black bear. There are parts of the black bear's natural range where an adult female human is on average larger than an adult blackbear. A black bear's claws are not sharper than your nails and with so much as a sharp stick you both outreach it and have a weapon capable of doing lethal damage.
Don't get me wrong, even under the best of circumstances a fight with a black bear is an uphill battle, but it is not zero percent, especially not if the human gets access to weapons they would realistically almost always have and that have existed since before the evolution of our species.
I guess fighting a bear wasn’t the point of the video, but I suppose I should have said “you can scare away a black bear even IF your chance of beating the bear is 0%”.
That said, in all but the most extreme circumstances (e.g. Brian Shaw vs
What I don’t understand is why the entire map doesn’t get covered in animals. Evolution excels at finding non competitive food niches.
On earth even in the most hostile environments, as long as there is food, animals find ways to live in them.
In your simulation the differences between the biomes must be too great for the animals to evolve into.
The animals under more stress should evolve more while animals in stable environments should evolve less.
Also, I am not seeing emergent behaviors. Are there pairing of traits between animals or even triplets?
Why would a die off in one jungle create another in another jungle? It should be the opposite as this is new territory for new types of animals to inhabit. This leads me to conclude that the die offs are inherent in your code instead of emergent. Something is affecting all your animals at once.
So I think, with the regions, some of it is a matter of time. Creatures on earth earth everywhere, but they had billions of years to get there. The move into the savanna region is good evidence of this I think because this simulation was significantly longer than any of the ones before (and I actually cranked up the rat ore of evolution, so there was much more scope for change). I suspect that, with enough time, some builds would make it into the harsher biomes, but who knows.
With regard to the extinction, I too found this curious, but given that it has only happened once at this level, I don’t think it is a code flaw. My best guess is that creatures on both sides of the map were moving towards similar ends but at slightly different rates (noting that the extinction took place over around 100k turns), and that “end” was teetering on the edge of something unsustainable. The only evidence I have is that we saw, for example, koala builds emerge at similar times even though the populations had been entirely isolated from each other. In other words, my best guess is that the mass extinction wasn’t so much a chain reaction as it was “convergent evolution” which ended in something unsustainable. Anyway thanks for the comment!!
@@EightLittleBears
I understand your point about time, I just disagree that there hasn't been enough time in the simulation. Time matters to us but not to the creatures in the sim. It is just marking that things have changed. Time spans between events don’t really exist, only the order in which things happen. Why a particular event occurs is the more interesting question.
This is an experiment I propose. Have one biome and one creature ideally suited for it. For instance, green food and animals that eat green-only food. Add another biome, with blue food, where that creature would die swiftly in it. Add a gradient corridor between the two biomes where there is some blue food and some green food (or a food that has a percentage of both green and blue). (As a control, run the experiment without a corridor but just an abrupt transition between the two biomes.) The animals should be able to move between the two biomes without any hindrance except the intrinsic differences of the biomes. The animals should also have small random variations to the tolerance of green and blue food and a separate ability to gain nutrients from the food independently of each other. Add these attributes one at a time between experiments.
My hypothesis is you will see animals eventually fill both sides. I also think you will see a bouncing effect where animal types will oscillate between the two biomes with the availability of food.
It could just be a coincidence that two similar animals in similar biomes died off. It is more likely the die-offs share a common cause. Convergent evolution doesn’t explain extinction in two isolated areas. It is like saying a die-off of river dolphins and sea sharks is caused by them both having similar shapes. Can you save the sim at a midpoint like right before and right after a die-off and rerun it to see if it has the same outcome? If it is the same outcome that could give you a clue as to the root cause of a global die-off.
The second experiment would test the global nature of your sim. This time have two identical biomes with the identical type of creature in them and an impenetrable wall between them. Run the sim to see if they die-off the same way (regardless of the time between). (A control would be having one creature type unable to evolve. A true control group would be one outside the sim which is not possible.) Try the experiment again, this time with a difficult but permeable barrier (one in a thousand creatures can get through it).
BACK DOWN