A Possible Problem with Niches...

Poisson wrote:

So we seem to agree that a hybrid of Lamarkian and Darwinian evolution is required?

Not required, but important for getting what you want, I see most people playing this game having a certain goal in mind. It might as well be required. Let's say it is required, then.
But for those who want something completely random, an 'off' (or 'auto-evo') option is also required. This would be the non-player species' evolution.

The Uteen wrote:

for those who want something completely random, an 'off' (or 'auto-evo') option is also required. This would be the non-player species' evolution.

would this mean it would randomly pick the 'yes' or 'no' on the evolve tap

or would it pick wich ever one is needed to survive (although im not sure how a computer would know such a thing)

yeah something would be a helpful thing like "stronger venom for killing prey"
but "faster creature" to avoid predators would be needed for survival

Not necessarily. There's a little wiggle room, honestly - instead of getting faster, it could ambush prey, and instead of "stronger venom" (Note to self - there's a lot of variety in the venom department, do a data dive unless someone gets to it first) the organism might be fine with the level of venom it has, or develop an additional strategy. (Venom's not as straightforward a situation as speed.)

However, a "select all" button would be problematic, as I see multiple options coming up (devote more percentage muscle mass to back legs (speed, jumping, etc.) versus devote more muscle mass to front legs (for digging, wrestling, etc.) would just end up balancing the creature out.)
I think the best way to code our random evolution button would be to have it make choices based on what we spend the most time doing (say, digging) without telling us.

Mysterious_Calligrapher wrote:

I think the best way to code our random evolution button would be to have it make choices based on what we spend the most time doing (say, digging) without telling us.

But then we again run into the problem of how do we tell a computer that this is good for digging while that is good for running? Just a random yes to one would be better if we want more Darwinian evolution.

Random, or yes to all, whichever works better; but deciding based on actions would make it Lamarkianish.

I don't think that would be an issue, so long as the other creatures were random, and we divided the species up correctly.

The problem is, we have many of each species making these events. We might solve this by making one organism representative of their species, and the next generation is effectively determined by them.

roadkillguy wrote:

The problem is, we have many of each species making these events. We might solve this by making one organism representative of their species, and the next generation is effectively determined by them.

This seams to be the only reasonable way to do things.

While we are on the topic of other organism evo, how are we going to determine behavoir of other organisms?

If we have one "stock" organism for each seperate species, it seems like it would make programming a lot easier. As far as predator/prey, we can have our old friend population dynamics be in charge of shifting the food web, which would determine where creatures would be found. (Due to the fact that herbivores are found where they can eat plants, and carnivores are found where they can eat herbivores.)

Location shouldn't be too hard, but specific behaviors other than predation might be a little iffy.

@OP

So...is this now the official autoevo procedure?

specialk2121 wrote:

@OP

So...is this now the official autoevo procedure?

Ah... no. I really need to put up a thread to finalize this.

I read this:

specialk2121 wrote:

The Uteen wrote:

for those who want something completely random, an 'off' (or 'auto-evo') option is also required. This would be the non-player species' evolution.

would this mean it would randomly pick the 'yes' or 'no' on the evolve tap

or would it pick wich ever one is needed to survive (although im not sure how a computer would know such a thing)

and it gave me an idea.

What if each niche had two lists. A list of every minievolution (mutation, I guess) that's good for that niche, and a list of every bad minievolution (okay, I kinda like that) for that niche.
This way, the organism CAN know what's good for it.

Let me know if I'm wrong, but I think that's the solution we've been looking for. That right there is what completes the circle.

Of course, whenever new minievolutions are added, we'd need to update all the niches. That's a lot of updates, and the computer is blind as to what goes where. Thus, we'd need to have niche artists "adopt a niche" and look through each new minievolution for a match.

Eh?

roadkillguy wrote:

I read this:

specialk2121 wrote:

The Uteen wrote:

for those who want something completely random, an 'off' (or 'auto-evo') option is also required. This would be the non-player species' evolution.

would this mean it would randomly pick the 'yes' or 'no' on the evolve tap

or would it pick wich ever one is needed to survive (although im not sure how a computer would know such a thing)

and it gave me an idea.

What if each niche had two lists. A list of every minievolution (mutation, I guess) that's good for that niche, and a list of every bad minievolution (okay, I kinda like that) for that niche.
This way, the organism CAN know what's good for it.

Let me know if I'm wrong, but I think that's the solution we've been looking for. That right there is what completes the circle.

Of course, whenever new minievolutions are added, we'd need to update all the niches. That's a lot of updates, and the computer is blind as to what goes where. Thus, we'd need to have niche artists "adopt a niche" and look through each new minievolution for a match.

Eh?

I have no Idea how we could make that work. If I'm understanding this right, that means had coding good and bad mutations, right? I'm open to the idea, but it seems pretty complex.

I see how we could hard-code some adaptations as being good for certain conditions (thicker fur = good for cold temperatures) but there's going to be some adaptations that tailor the niche to fit them. Hawaiian honeycreepers, for one - their niche is to feed on a certain species of plant, because their beak is shaped specifically for sipping nectar out of it. This is good, so long as the plant thrives.
Then there are going to be some that have benefits and drawbacks. I posted somewhere about muscle mass allocation - speed versus strength, essentially. If a greater percentage of muscle mass is in back leg positions, the creature can jump/run faster, and if a greater percentage is in the front legs, it can dig faster. (Note that this is percentage, not amount - there's only a certain amount of muscle fiber a body can take before the bones need to be scaled up, and then the body mass is greatly increased, which means that the extra muscle that was handed out is now required just to move around, so you haven't gained anything, really.)

...so if the niche calls for a fast organism, it will choose to be faster, and if it wants a digging one, it will choose to dig faster. I don't understand.

~sciocont wrote:

I have no Idea how we could make that work. If I'm understanding this right, that means had coding good and bad mutations, right? I'm open to the idea, but it seems pretty complex.

It's just as complicated as niches themselves. Each niche would list which mutations it wants to be successful.

We already have lists of biomes and niches yes?

roadkillguy wrote:

~sciocont wrote:

I have no Idea how we could make that work. If I'm understanding this right, that means had coding good and bad mutations, right? I'm open to the idea, but it seems pretty complex.

It's just as complicated as niches themselves. Each niche would list which mutations it wants to be successful.

We already have lists of biomes and niches yes?

But mutations and niches are dependent on other organisms. The only way to not break the game is to let the game or player decide if the creature is fit or not.

Now, each biome could have traits which would be useful to survive in it, like heavy fur as mentioned above. That would help quite a bit, but giving definite lists of good/bad will ruin diversity and probably break the game.

How else is the game going to answer: This creature just unlocked stronger claws - Should this creature accept them?

roadkillguy wrote:

How else is the game going to answer: This creature just unlocked stronger claws - Should this creature accept them?

We don't have a part-based system. Remember, in the OE, everything is pretty fluid. there are no spore-like "parts". That creature will get those stronger claws whether it likes them or not, because mutations are random. The problem is how the claws will affect its survival.

Forgive me if I missed something, but aren't mutations random? Therefore who cares what the environment is, a creature will get who knows what kind of [perk?\trait?\ability?\part?].

Gorbachof wrote:

Forgive me if I missed something, but aren't mutations random? Therefore who cares what the environment is, a creature will get who knows what kind of [perk?\trait?\ability?\part?].

If what you're suggesting is to simply randomize each organism slightly, and then let the best adapted creatures survive, then no. They have already said doing complete Darwinian evolution would be too much for the computer to handle and that it would not allow the player to deliberately progress toward sapience. This method of the environment altering mutation is a matter of playability.

Those that prove beneficial, however, will eventually show up in the majority of the species population and then either create a new species or change the existing species. /evolution on speed

Since we don't have the play time to run through 1,000,000,000,000,000,000,000,000,000,000,000,000,000 insignificant or fatal mutations before we find one that changes your species or creates a subspecies/new species, we ignore them. Also, keep in mind that this is a game, so you can't really cut out the player entirely.
/OT

As stated by many people above, hardcoding is tricky at best, game-breaking at it's worst. Only a few niches might have certain traits that could be hardcoded - and those wouldn't necessarily be traits, they'd more be a whole set of potentialy beneficial traits that would conquer one specific challenge of the ecosystem.

Ex:

Quote :

Biome: Arctic Tundra
condition: extreme cold
adaptational factors extreme cold: insulation, warm blood.
Traits creating advantage "insulation": thick fur, layers of fat, large/flattened respiratory orifices, coloration absorbing stellar radiation

Gimme a bit to think of how we'd swing "warm" and "cold" blood by the OE. Basically, all of the above are ways that LAWK deals with the real arctic tundra. Polar bears come to mind - even though their fur is translucent to camoflage them in the snow, their skin is black to absorb as much solar heat as possible. Larger/fatter noses warm the air faster as it is breathed in.

Other conditions of the arctic Tundra, such as a high ratio of secondary consumer species to primary consumer species, could not necessarily be hard-coded for, though certain things such as "Growing season x% of year" could potentially be hardcoded. And anything less general than an adaptation just to tolerate the biome present is going to be fracking belgium to hardcode.

Quote :

Those that prove beneficial, however, will eventually show up in the majority of the species population and then either create a new species or change the existing species. /evolution on speed

Well that makes sensce, like here on Earth you have a ton of different animals that in some way or another resemble each other.