Expectations for Filling Important Niches in Thrive

[HyperbolicHadron]

What is our general plan for filling specific important niches for organisms in the later stages?

I see two different ways that our audience will want to play Thrive:
The player will begin a game without a set strategy in mind, and come up with what they want to do gradually over time. They will adapt to the environment more, and take advantage of the existing environment as it develops.
Or, the player will start a game with a specific goal or strategy in mind that they want to enact. This goal could be as simple as creating a herbivore that eats whatever flora exists on land. Or, it could be as complex as creating a flying ‘dragon’ that hunts ‘whales’ when they surface for air. Not only would a complex strategy like that depend on the player’s skills at designing an organism that would have success in that role, but it also requires auto-evo to have created an organism that is similar enough to the niche and behavior of a whale. (And assuming that it is theoretically possible to do this in Thrive at all, but that is not the question here)

A player can also switch between the two ways over time, perhaps they come up with a different idea halfway through the game. Still, the question remains: will a player be able to rely on specific niches or specific [real life animal]-clone existing in their game? If so, is it that auto-evo will ‘usually’ create it, or will it be 100% guaranteed? And what will the boundary be, between the organisms you can rely on existing and the ones you cannot?

For the previous example, expecting auto-evo to have created a ‘whale’ analogue might not be something we would want to guarantee. But, expecting there to be a ‘grass’ or ‘tree’ equivalent is something that would be expected in any reasonably earth-like scenario.

This mainly is a question of what we want the default or close to default settings to produce in the environment.

On the one hand, I expect most players will want to create Earth-like creatures and have success and they will expect that the environment will support their goals. I can imagine a common example to be a lot of players wanting to create a dinosaur, perhaps wanting to get a race of T-Rex to space for example. Should the player expect to be able to have a T-Rex organism ‘just work’ in a normal game of Thrive? Would auto-evo be guaranteed to create and sustain a population of large herbivores or other suitable prey organisms? I expect most player’s organisms would be at the Apex of the food pyramid by the time they reach Awakening, so the rest of the environment is basically under the complete direction of the auto-evo system here. If the player is wanting to create a huge carnivore and auto-evo randomly happens to not create any suitable prey, then the player will likely be very frustrated and would have failed through no fault of their own. I am of course assuming that the player has done everything right, and isn’t playing on a too difficult mode or going for an atypical strategy.

On the other hand, if you just force auto-evo to always converge on the same fixed set of organisms every time, with basically cosmetic differences from Earth, then the potential for interesting auto-evo generated environments will be made minimal. Do we want every single game to include an ‘alien cow’? And how important will it be for the player to adapt to the environment given to them, as opposed to the one they would have wanted?
I know that from the implementation side here, consistency seems more preferable. It also seems like it would be preferable from a game balance standpoint. But, we don’t want to make the game’s environment be too identical across different playthroughs. And depending on how the player wants to play the game they might want a more standard environment they can rely on knowing and planning for in advance, or they might want a completely unique environment to adapt to for a challenge or for a random experience.

I don’t expect us to have a final answer for this question yet, but I think it is important to figure out as we are approaching the next phase of development of the later stages.

This is already becoming relevant for my developing of the World Generator, as I would like to know what the expected properties will be of the flora typically generated, ‘grass’ and ‘tree’ equivalents especially. I have a separate post asking about that question at Flora Generation Constraints as it is more specific.

[Deus]

I will post some of the thoughts I shared with you in our conversation yesterday:

Very good questions. I think it’s best to ask for more opinions on this, especially from Thim and GameDungeon who have worked extensively on auto-evo, but here is my take on it from a design perspective.

My take is that we should preferably tie auto-evo to the environments present on a planet, which will go a long way in providing a lot of variety even without an extensive amount of special handling. So if there isn’t an environment which supports that “cow-like” organism, it won’t show up, and the player will have to make do.

The dynamic that is a bit more complicated for me to answer is, if there are environments which support a cow-like organism, should that organism be guaranteed to show up? My immediate reaction is no, or that we atleast shouldn’t specially handle the system so that organism always shows up. The ideal is ending up in a place where auto-evo is trusted to make its own decision with that; to have the possibility of that organism showing up and to have auto-evo recognize that as a likelihood, but not intentionally cooking up such an organism to appear as soon as the player shows up in a given environment.

I do consider different implications of this question though. In Earth’s history, we see different “eras” where it seems that “mega-fauna” show up, then disappear for a world generally dominated by more average sized organisms, then the mega-fauna show up, etc. (dinosaurs showing up, going extinct, rise of larger mammilian organisms, they go extinct, etc.) How do we represent those dynamics in Thrive, and is that something which should be targeted? I’m sure many players will very frequently aim for evolving a mega-fauna organism.

I think one thing to look at is pacing population dynamics in a way that ensures organisms don’t dramatically change with every trip to the editor. In the Microbe Stage, such a large time is passed with every generation, that it is natural for there to not be very consistent species that stick around for too long - which obviously can result in unstable environments. In the macroscopic stages however, the time jumps will be much less dramatic, meaning there should be more consistent organisms and species present across multiple editor-trips, allowing the player to more robustly settle into a given niche or ecosystem for atleast a few generations at most and a solid chunk of their time in the macroscopic stage at most. And since auto-evo works off the organisms that are already present, that will ideally lead to a somewhat consistent presence of certain clades, genuses, families, etc.

This ties into your consideration of balancing what the player desires to create vs. strategically reacting to what is present in the environment. If the player wants to create a T. rex for example, there should be two things that act as important factors to consider. First, it should take some time for the organism to, say, get bone-crushing teeth, a larger size, an efficient respiratory system, etc. if the morphology of the organism is petite and delicate. Second, their environment should be a limitation; if they evolve a T. rex during the Permian Extinction analogue in their planet’s history, well, good luck with that.

And that also ties to auto-evo “seeding” forms of life. The Thrivian cow-analogue shouldn’t always be forced to evolve, but if and when it does, the player should be able to see this organism and its related organisms be a reliable food source, thus allowing them to respond with an upscaling or downscaling of their organism.

As such, in a standard playthrough, I think the player should generally be made to think strategically of their evolutionary path, but should be given both variety in ecosystems across time and some form of “steadiness” in the different ecosystems. In other words, the player might not always be able to create and maintain their mega-fauna, but should be given chances to do so once they reached a sufficient level of morphological progress in the stage.

Of course, this should be tempered by difficulty. With easier difficulties, Thrive should encourage creativity, allowing the margin of error to, for example, put a T. rex on desolate planet without some negative impact, but not enough to absolutely make you miserable. In normal difficulties, players shouldn’t be absolutely denied a T. rex if their ecosystem isn’t perfectly ideal for such an organism, but their performance could be damaged and their success reduced due to a lack of food sources, and some periods in your planet’s history just might not be it for that sort of mega-fauna. In harder difficulties, evolving a T. rex would be really punishing, and could easily result in your organism going extinct; as such, you must prioritize fitness to the current ecosystem above your creative desires.

---

Also my thoughts on how Earth-like we want evolution to play out in Thrive, which has implications on auto-evo:

This is something I’ve spent a lot of time thinking about as well. I think the “standard” Thrive playthrough, and what the planet will default to without further customization, is essentially an Earth-like story of evolution, but with the chapters potentially reorganized chronologically, crossed out, or expanded on. I think in this standard playthrough, the environments present on a planet would be pretty similar to those on Earth, and as a result, the organisms to evolve will show some similarity to those on Earth.

In that standard playthrough, I think the variety will shine through a different sort of evolutionary story when it comes to the type of organisms that are dominant. For example on Earth, you saw a period early on where arthropods showed dominance over numerous ecosystems, then the rise of fish and the early advent of amphibians on land. Then the rise of reptiles over synapsids/proto-mammalians due to the heat of the planet at that time, then a crescendo of life on Earth’s most (imo) awe-inspiring organisms in the form of dinosaurs. Then their extinction, allowing the rise of mammals, and the rise of a mammalian sentient lifeform.

But what if on Thrive’s standard planet, the rhythym was much more different? What if the early fish-analogue evolved at an unfortunate time for their success, leading to a much longer period of arthropod/invertebrate dominance and a much larger presence of giant arthropodic organism throughout all of life’s history? What if the planet was much cooler in the “Permian” of Thrive’s standard planet, meaning the mammal-analogue actually had more fitness over the reptile-analogue, thus resulting in their crowning dominance? What if the reptile-analogue was who benefited from an extinction event, thus resulting in the sentient organism of the planet having scales?

This isn’t me signifying that there must be a reptile/mammal/arthropod analogue as we know them in Thrive, though the structures we offer in the editor and the timing of “unlocks” will probably be dependent on Earth’s lifeforms. What I’m trying to say is that the timing of extinction events, and the evolution of the planet’s climate, could have really dramatic results on variation. And ideally, we can structure our “parts” and tools in the editor to allow that sort of variety to be natural. Though we should concisely ensure that there’s a plus and a minus for all adaptations, we should try our hardest to untether our perceptions of what traits are successful on Earth at the current moment from what actually could be successful in diverse environments.

With customization, things would probably be more directed from the player. For example, a colder planet would likely result in larger organisms with less surface area on extremities and some sort of shaggy covering, while warmer planets would likely be dominated by organisms with features which retain moisture, like scales, and planets with very limited area that isn’t swampy or close to the shore would likely feature more amphibious organisms on land. A planet with more oxygen has larger arthropods, meaning exo-skeleton bearing organisms are much more dominant as large organisms in different ecosystems. etc. etc.

[hhyyrylainen]

I remember there being a discussion about this way in the past, though it probably was just on Discord.

My opinion on this topic is still that we should avoid hardcoding auto-evo making any specific creatures as long as possible. And I hope we can make auto-evo good enough without these kinds of fixed guide rails on what it should make. So I don’t think auto-evo should be made to explicitly try to create creatures there ought to be. I think with the miche model auto-evo has been very good in basically always managing to fill all of the available miches within a couple of generations.