Pacing and Combat


Originally from a Slack discussion, copied here so we can continue in public.


Tjwhale where did you get with your combat mechanics concept?


What do you mean about combat mechanics concept? I think we’re all agreed (right?) that there’ll be pilli and agents and you’ll swim around stabbing and spraying things to damage and, hopefully, kill them. Agents will be coded to only work on certain organelles in certain species (bit masks ala moopli) so you’ll need to make sure you have the right ones for the job.

Personally, and this is more about preference than anything, I think the combat should be quite fast. Think more like bullet hill twitch gameplay with pausing to sort out whether you have the right agents loaded (if that’s even how agents work) or to turn on and off internal processes (if we have that, like turning off protein synthesis in a fight to save atp - this is the basis of the movementless prototype I made so I guess it blends the playstyles well). So a life might be as little as 10-30 seconds and you just madly rush around trying to get enough compounds to reproduce. It’s a race against time more than an RPG where you spend hours playing one life and caring for it carefully.

But as I say none of this has been discussed (and I totally understand other people may want a completely different, slower, more thoughtful, style). Pilli and agents yes, how it will feel to play them (and precisely how they will work) not finalised.


I was playing Reassembly last week, and that feels a lot like how the cell stage could play. It’s fairly fast paced, both in game, and in terms of editing and improving your ship between plays. More importantly to me, the ‘galaxy’ your playing in feel very alive, there’s multiple other factions with various designs of ships going about their business, and resources are generated by the green plantlike things you see growing on asteroids, so there’s an element of territory control involved.

Not all of this translates well into a game about microbes, but if we can capture that feeling of business, and a continuous process of redesign, I think that might be fun.

Sometime last year we were discussing how what the player does should feed back into auto-evo. We never really came up with an answer, but one of my suggestions was that it shouldn’t, that playing should function as a way for the player to experience their creature, and that it should be the player’s experience that feeds back into how they evolve their creature, with what they actually achieve in game being of relatively little consequence, outside of gathering resources to reproduce, and possibly new organelles. This would work fairly well with reassembly’s gamestyle, or what tjwhale suggests above, with relatively quick games where you simply experience life as your cell.


I meant more along the lines of how pili, agents, etc. would be used. Would it be as simple as just going along to an enemy and pressing a button to stab them with a pili if they’re within range? Things like that.

Personally I’m not partial to a “rush around trying to get enough compounds” sort of gameplay, but as you say we all have different opinions so it’s difficult to say what it should actually feel like. Spore’s cell stage aesthetic (minus the eyes, squeaky voices and twirly dances) was actually pretty good in my opinion. Sometimes you’d be running away from predators but at others you’d just be collecting food, which gave a balance to the two levels of engagement. It was both “floaty” and tense, but not always rapid.

I like the idea of having an alive-feeling environment. There needs to be a lot going on within relatively short distances, but for me not all of it should be out to get you. The player could choose to get involved if they want, but I think they shouldn’t be forced into constant hair-raising situations. Perhaps certain attack or survival strategies would lend themselves to different speeds of gameplay, which would cater for both. Large photosynthesising microbes probably wouldn’t have much to worry about from most predators so gameplay would be nearly leisurely, while small agile predators with agent attack methods would have to both hunt prey and avoid larger predators, making their experience far more involved.


I agree it shouldn’t all be hostile. Some cells will be too small, they’ll run away from you, but probably not be worth the effort to go after, or they might be entirely passive if they ‘know’ your won’t bother them. Others will be autotrophs, or too big to bother with you, and yet others will be largely immobile, even though they might eat you if you get too close.

I do like the idea of having very short sessions be an option, though they shouldn’t apply to all species (it definitely wouldn’t work for proto-plants). Maybe highly r-selected species would reproduce very quickly, and in large numbers of small/weak cells, so that a gameplay session only lasts <1 minute… the risk there is that this becomes very repetitive. Equally on the longer end, if it takes 4-5 minutes of fairly monotonous gameplay (like the recent versions) to reproduce, you’d wish for something shorter. We need to find a balance of having enough to do, and it being interesting enough to spend the time doing it.

  1. I think pilli will just stab anything they touch, friend or foe, without needing to be activated or anything, it’ll just be a collision. That’s my understanding, happy to discuss it more. Agents we really need to have a proper talk about and fix something (probably through prototypes) because the bit-mask system is important and needs testing.

  2. The thing with photosynthesizer gameplay is that it’s going to have to be radically different. If you can’t move what can you do? The game can’t just be sitting around waiting to fire agents. My suggestion has been (prototype on the old forums and in the git folder) that the game should be turning on and off your internal processes in response to what is happening in the environment with time quite radically sped up.

My original idea was for a tree where a season is 10 seconds and you need to turn on and off growth and seeding at the right times, “oh noes a horse has bitten off half your leaves” etc, much like an infinite running game bizarrely enough.

This could be meshed in with the main game (swimming around as a predator) by giving all microbes the option to control their internal processes (turn on and off different ones at different times) and therefore the movement-less gameplay would be a pretty natural extension of this. Not sure, happy to discuss it.

  1. When it comes to combat I think testing should be king. We should get pilli and agents in to the game and then test lots of different speeds and styles until we find something good. I am happy with anything so long as the gameplay is good. I don’t really mind inside that exactly what it is like in terms of speed, controls etc. It’s just got to be the kind of thing you can sit down to and play some amount of without frustration and then look at the clock and think “I should really stop”. That’s what we should be aiming at IMO. That’s why I’d like to get started with testing sooner rather than later because it could be a long process.

  2. @Oliveriver I agree about not getting involved, I think often you should see one microbe chasing another and think “thank god that’s not me” and just swim away without participating. That’s cool.


I have been thinking about ways how we could make the AI better and how actual cells and bacteria fight. I decided that I would get the best results simply by looking at videos of actual bacteria eating each other. Apparently this was not as easy as I thought, since simply typing “Cell fighting cell” or “Bacteria eats other bacteria” tends to produce different results than the one I need.

Anyway, below I have compiled a list of videos I found in the past two hours that could be used to get a better idea at how the combat mechanism should work.

To begin, I would like to share a very interesting weapon I found. In Thrive we planned on having agents that we could shoot at other cells; nature went one step further and developed a gun that could be used to launch these agents quickly at other organisms.
You can read the article if you want, but if you scroll below you can find a video that shows it in action:

If the live footage makes little sense to you, just skip to minute 1 which shows a graphical demostration.

The second category of interesting fighting mechanisms I found was hand-to-hand combat.Most of these videos involve killer t-cells attacking cancer cells by injecting them with ligase. The videos are ranked by order of interestingness i.e., the first one is cool too watch while the rest are less so: (watch only the first minute, after is a commercial) (you should probably watch this one too, its good and in some of the videos you can actually see the ligase being injected) (the red one is the killer t-cell)

The third category is simply engulfment. What surprised me in these videos is that a lot of times a very slow amoeba is able to catch a ciliate, which are really fast, by stalking it. (they don’t even see it coming) (this one uses pseudopods to stalk its prey) (another case of phagocytosis) (at 0:33 you can see a cell come in feeling prey with a tentacle)

Now is a misc category of other videos I found. (really slow in my opinion) (I don’t even know what’s going on tstt) (this is a good example of phagocytosis)

As you can see most of these are either engulfment or hand to hand combat. I suggest the first one will work in the game by simply trying to swim over a smaller organism than you. The second one will involve trying to engulf something, but failing (its to big to eat that way) and then pressing the numpad keys to release agents. What I couldn’t find a single example of was a cell actually shooting agents as in the current version of Thrive. On the other hand I did find a cell using a lure that attracted smaller cells to it and I think this should hands down be an unlockable organelle later on.

I will now try to find other methods of cell vs cell combat.


Sorry for not replying to this thread, I kind of forgot. :flushed:

Straight away I’m thinking that would hamper any attempts at forming colonies if any cells have pili. Then again, the player might eventually have to make a choice between easy survival (pili) and progressing to the next stage (colonies) so it might work. I think the player should still be able to control their “spike” attack in some way though (perhaps by shooting agents out of them like the example @TheCreator mentioned). Maybe the player can extend their pili slightly by pressing an attack key, so they can attack more effectively manually but will always stab anything that comes within the normal range anyway.

I agree that managing processes should be possible, and probably the focus for photosynthesizing players. It’s been planned for a while now that the game can be paused and the player can change compound priorities for certain organelles to optimise their metabolism given variation in environmental compounds.

However, I’m wary of slowing down the game pace. For instance, what determines whether the player cell is a photosynthesizer and whether the game should slow down? Do they just have to add a chloroplast or evolve a full cell wall? The former definitely wouldn’t work because there’s a chance for the player to assimilate free living cyanobacteria to gain free chloroplasts, so the game would have to slow and change mechanics in the middle of a gameplay session.

Really I don’t think there should be radical changes in game mechanics, just a different way of using them. Both photosynthesizing and non-photosynthesizing cells could have access to metabolic management, but it should be balanced so that photosynthesizers find it the most useful. In my opinion they should still be able to use agents in the same way as other cells (even if that is slightly scientifically inaccurate) but they’ll be more inclined to use symbiosis-type agent signals like warning others about a predator.

Nice find! I think that’s basically the equivalent of what we have already - it would only take a few semantic and visual changes for it to work like that.

I believe the plan is to add a specialised engulfing edge to some areas of your membrane, allowing you to engulf smaller cells from that area of your membrane only. Like pili, we might want manual control over this, or we might not. Accidentally engulfing a friendly microbe is a bad thing, but I guess good players will realise this and won’t do it.

Good idea. It might be difficult to prevent this becoming completely overpowered though. Perhaps NPC AI could evolve to be aware of this and not fall for the trap, giving the player the upper hand for a while but eventually making it useless.