Many years ago I made a small "game", there was 20x20 map and up to 100 creatures (it was an ancient computer so the performance was an issue :)). These creatures were moving around the map (completely random) and breed or eat grass. Each grid had certain amount of grass (which increased/regrow each turn), the creatures had "chewing" stat which determined how much they could devour per turn. There were also other stats like "breeding cost" (how much food is required to give birth to an offspring), "max age" (after how many turns the creature will die), "virus resistance" (if you pressed 'v' key, there was a disease and all creatures that failed virus resist check died instantly), "temperature resistance" (if you pressed 't' key, there was a solar flare and all creatures that failed temperature resist check died instantly). When two creatures (of any kind, there was no gender) meet they could produce an offspring (using up stored/digested food). The new creature had average of all stats of both parents +/- 1 (mutation).

I observed many fascinating things (like the breeding cost increasing when I did a lot of disasters - under peaceful conditions the breeding cost and max age were the stats that were going up), note that there was no algorithm or AI of any kind, just pure probability that driven the artificial evolution. I was mesmerized by it.

Generally, not much of a game, nothing to do and overall boring. But... Somehow I enjoyed it. More watching than playing (end turn, inspect a creature, look at the map (full of 8x8 pixel almost identical icons), v/t key for a disaster; and that's all).

I wonder, is this idea worth pursuing? Is it worth trying to make a small game with this premise (generally, there is no problem making it more rich and complex, but the player's input would still be low)?

If you enjoy it, and it would satisfy you, why not?

If your concern is, "Will others play it?" that's a marketing matter, and you'll have to investigate that further than a handful of opinions on a programming forum.

There's quite a few experimental or niche games out there, even if they don't have mainstream appeal, and this doesn't sound too obscure. The gaming world was captivated with the evolution of the apocalyptic ten-year Civ-II game, after all.

If you can think of interesting ways to improve on your idea, and are motivated to do so, that's usually reason enough for most people.

I see two issues with this kind of game, neither of which is all that serious. First, as you describe it it sounds like a simulation with very little "game" aspect to it. Fun as a toy, but without some awfully deep simulation the emergent behaviors that would interest me would run out pretty quickly. Second, and relatedly, it can be hard to include enough variables to allow for interesting and unplanned behaviors while still being balanced enough to produce diverse behaviors. But that second piece sounds like a lot of fun to sort out

Many years ago I made a small "game", there was 20x20 map and up to 100 creatures (it was an ancient computer so the performance was an issue ). These creatures were moving around the map (completely random) and breed or eat grass. Each grid had certain amount of grass (which increased/regrow each turn), the creatures had "chewing" stat which determined how much they could devour per turn. There were also other stats like "breeding cost" (how much food is required to give birth to an offspring), "max age" (after how many turns the creature will die), "virus resistance" (if you pressed 'v' key, there was a disease and all creatures that failed virus resist check died instantly), "temperature resistance" (if you pressed 't' key, there was a solar flare and all creatures that failed temperature resist check died instantly). When two creatures (of any kind, there was no gender) meet they could produce an offspring (using up stored/digested food). The new creature had average of all stats of both parents +/- 1 (mutation).

I observed many fascinating things (like the breeding cost increasing when I did a lot of disasters - under peaceful conditions the breeding cost and max age were the stats that were going up), note that there was no algorithm or AI of any kind, just pure probability that driven the artificial evolution. I was mesmerized by it.

Generally, not much of a game, nothing to do and overall boring. But... Somehow I enjoyed it. More watching than playing (end turn, inspect a creature, look at the map (full of 8x8 pixel almost identical icons), v/t key for a disaster; and that's all).

I wonder, is this idea worth pursuing? Is it worth trying to make a small game with this premise (generally, there is no problem making it more rich and complex, but the player's input would still be low)?

I have been wanting to do some kind of simulations for a while. The initial spark was from Conway's game of life (http://en.wikipedia.org/wiki/Conway's_Game_of_Life) when I saw it a few years ago. I have a few ideas but never get to complete it. My last attempt of this kind of simulation was posted here as well (http://www.gamedev.net/topic/642857-pixel-garden-prototype/). It was a total random idea I made in 2 days. I didn't see any way to make it more interesting so I stopped there.

Perhaps you could post what you had then. I am quite interested to see what you have =)

It's a lot like the first Sim City, where you decided where people could build but the environment around it determined whether they also build something and whether it became valuable.
Sim City was a great success, although i don't think it has had any die-hard fans.
Where you still lack, compared to Sim City, is having a goal for the player, though goals can be made up during development of this idea.

(you could have different species of animals and be trying to get rid of the more parasitic species)

Players' input could be anything you want to add, btw, from disasters, to setting average temperature, to having workers build shelters(to keep the number of animals high) or adding an animal that makes the grass grow faster.

I would rethink the original system, maybe add genders and/or have the stats mix in a different way.

I observed many fascinating things (like the breeding cost increasing when I did a lot of disasters - under peaceful conditions the breeding cost and max age were the stats that were going up),

Were the different stats tied to each other in some way ? (was there a max total stat?)

Acharis, on 22 Jul 2013 - 9:54 PM, said:


I observed many fascinating things (like the breeding cost increasing when I did a lot of disasters - under peaceful conditions the breeding cost and max age were the stats that were going up),

Were the different stats tied to each other in some way ? (was there a max total stat?)

No. It was extremely primitive, each stat had 50% chance to go -1 and 50% to go +1. I guess, the "nature" stopped caring about breeding speed (which was prefered under normal conditions) when there were many distasters. Since it's all disconnected and pure random I guess the creatures with a single stat that was critical for survival were breeding (and it's statisticly easier to max up one stat than two simultaneously) and the tradational stat was neglected. Anyway, it was so many years ago I'm not so sure I remember this correctly :)

My guess is that the initial stats had a set total and that was divided between breeding and surviving.

I think the player's input/interaction (the lack of it) is the biggest issue with such a game. Also, the whole concept of "evolution" implies non ingerention (that's why it was fun, it was doing "itself").

BTW, have any of you played Spore? I didn't, but it could have some similar premise, I think; althrough I have heard it's a disappoint game. If you played it, maybe post something that could be related?

BTW, have any of you played Spore?

The idea you described is hundreds of times closer to actual evolution than Spore ever was. Spore had no evolution - you got to collect 'points' and 'body parts' from skeletons, then you could build your species in any way you wanted to. If you got the right parts, you could be a herbivore biped one generation, and then be a carnivore quadraped with two mouths and three eyes the next.

There is no natural selection (especially since you controlled a single organism) and there are no environmental conditions to put pressure on you to design a certain way.

It did make a semi-decent game, mostly because of the final stage - space exploration (if you ask me).

But perhaps one thing you can take away, to involve the player more is to do what Spore did, and put the player in the shoes of a single character.

Basically, have some system with enough complexity to simulate evolutionary pressure and mutation, but force the player to try and survive from a single creature perspective. Each time the player creature dies, make the player pick one of the creatures from the next generation, creatures that were generated through the game's evolution mechanics. If you make the environmental conditions vary randomly, the player will be forced to try and survive with whatever creature he picked, whether that creatures mutations favor the conditions or not. The score can simply compare how the player's creature is doing vs the other creatures from that generation (which the player didn't pick)

That's just a vague idea though, there's probably more ways you can implement it, but putting the player in a position where he has to anticipate a single, randomly generated, creature's survivability in future conditions sounds kind of interesting to me.

The game sounds interesting, do you know what it's called? I'd like to read a wiki on something like that.

And I've been looking into making an evolution type of game, this sounds really interesting ... It is something I'd like to play