Beneficial effects of colonies and binding

With the implementation of binding agents now complete, we have found ourselves facing the question of how extensive the benefits they provide should be. I have already brought this up before, but I decided to continue the discussion here so we can refer to it more easily in the future.

The primary things we need to consider here are what benefits binding agents already provide, what they should provide, and how they should be distributed between colony members.

Present Benefits
Forming a colony in Thrive right now already provides some nice benefits that I feel quite pleased with.
First of all, adhering to a cell grants the player an extended reach for gathering resources and capturing prey without overbearing speed penalties, which as it currently stands, might be a bit too unbalanced due to the lack of physics constraints.

In addition to that, colony members provide an effective buffering armor against environmental hazards and predators, with the player being able to disband the colony and escape with the predators distracted by the player’s peers. I wonder how that strategy would hold up if player species death detracted from player populations…

Potential Benefits
In the event that the above benefits feel particularly lacking, I have come up with a few potential benefits we could implement as a feature for colonies, as well as including some ideas I have seen tossed around by others.

  • Increased healing rate due to the cells sharing their resources.
  • Decreased osmoregulation cost due to less exposed surface area
  • Increased effective engulfment size, allowing the colony to engulf larger objects using their collective power (This one might be problematic to implement)
  • Increased camera zoom, though this might become problematic with spawn radius…

This is all I have at the moment, feel free to provide your own ideas on what binding agents might provide if you have any.

Distribution Methods
The final thing I would like to go over is how the less abstract benefits might be distributed between cells, and how member count might effect them as well.


There are three primary methods I have devised through which value-based benefits might scale with cell-count, if at all.

Method A:
By making cells experience benefits based on the amount of adjacently connected members, we can encourage more strategic member placement. This would also help prevent benefits from scaling beyond belief. However, this might only be frustrating due to the lack of control players currently have on unbound cells, and how they may face difficulty in getting them positioned just right. It may also be a more difficult and time consuming method to program and debug.

Method B:
By granting a colony-wide bonus to all cells based on the total amount of members in the colony, we can encourage players to form ever larger colonies without worrying about cell placement. However, we will either need to limit the bonuses with a hard cap or diminishing returns to prevent the benefits from becoming too great.

Method C:
Applying the benefits as a flat bonus on the basis of a cell being in a bound state or not would likely be the easiest and least resource intensive route to take, but doing so would not incentivize forming larger cell colonies on it’s own.

Personally I think I prefer method A the most for balancing purposes but I fear it might be problematic to implement, as well as frustrating to handle cell positioning…

If anyone has any ideas on how else we might do this please feel free to comment.
I would also like to know which method is most desirable.

4 Likes

Thoughts thrown all together on this:

Perhaps the reverse effect as well: predators would require to be able to engulf the whole colony, perhaps? Or we might want to leave this to permanent colonies.

As long as you let the camera zoom be accessible to the spawn system, you can adapt to it, which would perhaps be even better than a fixed constant in any case. Although it could be a bit cheesy if you want things to spawn close.

An alternative is to base both the spawn radius and the camera zoom on the same metric, like effective speed or cell/colony size.

Well, extended resources storing is a pretty strong benefit as well, so you can survive much longer when compounds are sparse. This could especially be interesting for predators, as you risk less dying while chasing your pray (as long as the average net yield is positive… which I’m not sure of, given that I feel I spend more energy predating than other cells give me when playing this way).

I think this makes a lot of sense, as long as the total healing rate is not greater: one damaged cell will be quicker to heal, but if all cells are damaged they’ll all focus on themselves. Might be more complicated, but makes more sense to me and might be a bit more balanced (as the player will no longer have interest in just getting their fellow cells endangered because it will affect their own healing).

I think a good way to think of it is to think of potential drawbacks of colonies (because we don’t want binding to be the ultimate strategy, do we?). And I think an interesting idea that was thrown on Discord is the compound intake dependent on exposed surface: in a colony, the total exposed surface is lower than all cell by itself, so it both decreases compounds intake and osmoregulation costs (basically an already present dilemma with the membranes). So, an interesting yet complicated way to do this is the gradient-based method, which means that cells in the center will have less intake, but less osmoregulation as well.

The main interest of this is to prepare for multicellular specialization, allowing for weaker but protected cells in the center. The drawback is that it is more complicated from a player perspective, with no absolute benefit of binding (at least on compounds/osmoregulation). It also suffers from the same issues as method A, which might thus be a first step toward this more complicated approach.

1 Like

I’m not really sure how cell colonies move in real life but personally my idea for one of the benefit of a colony would be physical advantage, that is the colony’s physics in the environment. I would imagine a bunch of cells huddled and bonded together is much heavier with all their combined mass and with every cell also contributing thrust, a colony (given enough cells) could easily move big iron chunks and shove through small chunks and individual cells. Basically a colony has a higher impact.

Also this could play out with water currents mechanic if it were to be implemented. An individual cell is likely to be hugely affected by currents but a colony with sufficient requirements is more likely be able to stand their ground and move against the moving current.

There are probably many more variables and things that go into this but that’s the general idea.

2 Likes

Well, it has more inertia (mass), but is also generally more affected by external forces. However, these two are not necessarily increasing at the same rate. If cells are aligned, forward-facing currents will not make the force increase much (because it depends on the exposed surface, which does not change; I scrap any dynamics in the fluid changing from the shape of the foreign body, because I don’t think it will be important enough), so overall forward-coming currents will likely be less effective. However, you’d be as much affected by currents coming from the side (because both mass and exposed surface are multiplied by the number of cells here).

Note that thrust won’t play a big role here because it increases (at best, assuming that having a cell in the rear does not affect your thrust!) alongside the mass. It will only allow for more significantly effective pushing against things that are not too massive (i.e. chunks, not a whole ocean).

The rest is about coordination: do we want bound cells to coordinate? On the one hand, not doing so will make the player less in charge of everything (although, if you have other cells choosing a uniformly random direction, they will sum to zero, and the player would effectively be decisive, even though this means they’ll be moving the colony with their only thrust, so they’ll be slower given the increased mass). On the other hand, it seems more not to do so. And more importantly, it would open a lot of challenging gameplay decisions: if your fellow cells react to e.g. signalling agents, you may choose to resist them, but you’ll fail if you have too many cellmates. So you might prefer abandoning the colony, which gives an interest to non-permanent colonies (and I find it interesting in the sense that the player is not always affected by signalling agents, but not immune to them either).

Big organisms are less affected by currents,this holds true for both air and water.
Mass rises with cube and surface with the square.Generally the bigger you are,the less viscous the fluid will seem to you.Its why cells have smaller locomotion limbs relatively to body size then other bigger fish for example.

I think method B makes the most sense for me because in a colony cells work should work as a whole.OR maybe before a certain size the cooperation between them issnt so great so method A when size <10 then method B?But this will be a pain to implement.
This has to be thought out with the transition to multicelular in mind tho.

1 Like

That’s basically what I wrote, but surface only rises with the square in cases of full symmetry. Flat fishes are precisely an example where mass and (vertical) surface are roughly the same proportionally.

Also, as it stands now, our colonies are currently 2D, so there’s a degree lost here (mass square and surface linear). But, as I mentioned, if your cells align orthogonally to the current, then both mass and surface increase similarly, so you have no effect.

To put it shortly, it will heavily depend on how cells are binded.

Now, the question isn’t which model is the more accurate, but what trade-off are we considering best between accuracy/gameplay variety on the one side and computation time/player understanding on the other side. And the “more cells less affection by exterior” is fairly simple to grasp, BUT you have less interest in placing your fellow cells (which would pretty much be limiting if brought to the multicellular stage) and it may lead to blatantly unrealistic results in extreme cases (which players may very well go for, like a “make the longest possible colony” challenge).

Do you mean a string of cells?thats such a special case ,i dont think its worth considering.Plus i dont think this is an evolution simulator as i said before.
Btw,i still think the string will be less affected then lone members.

That’s not a special case but an extreme one (and, as I said, there are reasons for this extremism to occur). Basically, you’ll have a similar effect with non symmetrical colony bindings… and, as far as I’ve seen it on let’s plays, player tend to bind cells from the front, which results in strongly asymmetrical colonies, biased in the direction of movement.

If we don’t simulate evolution in an evolution-based game, what are we doing exactly? Now again, this does not mean that we want to simulate most accurately things, but we have to find a balance. And if we end up with situations where regular players start thinking : “well this is definitely not possible in real world”, then I think we’ve failed this balance pretty badly. The whole catch line for Thrive is “Real evolution inside! Can you beat it?”. So we have to at least fool the players into thinking they are facing real evolution. That’s also why we have theorists in the team.

And fluid dynamics is one of the things that we instinctively have an understanding of, so it’s an even touchier topic, which needs to be more carefully dealt with than, e.g. quantum physics, where only a handful of players will actually be able to tell that something’s off.

Given that in such a case, thrust, mass, and surface (i.e external force) all increase linearly with the amount of cells, I pretty much fail to see how acceleration = (thrust+ext. force)/mass does change.

A string of cells as far as im aware is possible in the real world.Maybe not evolutionarily viable
but possible

I suppose that’s fine as long as someone is willing to try coding it.

Due to resource sharing, your effective storage capacity is already increased, next time you play Thrive pay close attention to your compounds bar when you are playing around with colonies.

The problem with this downside is that right now cells equally distribute resources between themselves in a colony, at this scale, any delay before getting your share is negligible.

I love this idea, nothing really to add to it. Cell colonies in real life move in many different ways depending on their arrangement. Try looking up volvox colonies and slime mold colonies, two vastly different locomotion methods.

Strand-shaped colonies is very much a common strategy in the microbial world, you’ll often see it in algae. More surface area for each cell to perform photosynthesis while still benefiting from the structural support and safety of a colony.

We could perhaps implement this in Thrive by using the A method, thus discouraging photosynthetic cells from forming large blobs and instead having just one or two connections to form a chain.

I meant, it is already a strong benefit, which was not mentioned in the initial post.

Yeah, it does not handle cell differentiation, but it still changes the total exposure (i.e. income & osmoregulation) of the colony. It’s not very different, however, from what would have with single cell.

And I think this is the key here: almost every other benefit (except your 3 methods and perhaps faster healing) can be obtained with a single, larger cell. Why would we have, in such a setting, cell colonies? That’s a tough nut to crack.

A tentative answer could be that it’s easier to achieve (make two independent parts and join rather than create everything from one part), or falls under Nash equilibria (doubling the cell is beneficial, but you intermediate steps are not), so that’s just “auto-evo” at this point.

One could also invoke disposability : if a part of a colony is damaged beyond repair, it only kills x% of it, with a single cell it’s 100%.

More temporarily, there’s also the advantage of being able to split when needed (like chasing predators), but that can’t be applied to “permanent colonies” (i.e. multicellular organisms). Are temporary colonies a mere step toward it though, more likely to emerge, and fitter than loneliness, but less fit than permanent colony in every setting?

I can’t really tell more at this point, but I have the feeling that many things we discuss here should actually already be taken into account for single cell physics, upon which colony physics would be built.

Dont you mean method B?

Oh, my bad

In regards to colonies versus larger cells, unless you are a photosynthesizer, managing a large cell can be difficult and resource intensive. Thus colonies can be a superior method of reaping the benefits of being large, without actually being large.

There is most definitely benefits that colonies experience that true multicellular organisms don’t. Being able to disband at will being one of them, as well as individual cells being able to live and survive on their own should the colony be disrupted. Players will likely experience this in Thrive, but should they want to reach later stages they will need to be able to become a multicellular organism.

No I meant Method A, as through that method cells could have negative effects based on the amount of adjacent cells attached to them, rather than based on the amount of cells in the colony as a whole which is method B.

So, this basically mean we’d have a kind of penalty to larger cells (like total osmoregulation actually being x% higher for hexagonal size y%, added to the base cost for each hex)? That seems like an interesting way to incentivize colonies.

Question is, how do we get this to happen in the environment of Thrive, in a way that makes it noticeably more interesting for the player?

1 Like

Oh,i thought there would only be benefits not negative effects too.

Quickly whipped up some demo branches for each of the big ideas: Branches · Revolutionary-Games/Thrive · GitHub

2 Likes

Oh wow you work fast. What method are you using to determine how the benefits scale with colony size? If at all?
(Not counting the engulfment thing of course.)

1 Like

Just eyeballing it. I recalled at somepoint someone saying that about 10 cells is when we wanted to transition to multicellular, so I wanted that to be the “max level” point where you’re really good, but not goofy broken. At that point, you are either 3x regen rate, or half osmoregulation cost. Seems strong enough to get the feel for it.

1 Like

Finally got around to reading (most of) this thread. I just want to say one thing: for moving to multicellular it is a must that a colony can get pretty big. So maybe the first benefit should be a moving speed boost. I have the following in mind: colonies get slower with added cells because they have less thrust compared to the total organelle mass. To offset this, the colony leader cell should be able to get the base movement and flagella thrust from colony members (for now let’s assume that all flagella are pointing the right way around, otherwise this could get really complicated). Though, to not be overpowered that extra thrust needs to consume ATP like normal. Additionally maybe it should stop consuming ATP if there is less than 50% ATP remaining, this way colonies would still be a bit mobile even if all the calls can’t produce thrust at the same time due to insufficient ATP production.

1 Like

Any more thoughts on this? If nobody’s using them it’s probably about time to delete my experimental branches.