Game of Life as a Game?

[Jonax]

Fellow game programmers!

Game of Life by Conway is an ancient computing concept. For over half a century clever minds have played around with the possibilities of the cellular automata 'Life'.
They have discovered tons of amazing patterns and developments. I hear the the Conway Life in theory can even emulate a computer. For example running Doom. Though I don't know if anyone has really pulled off that trick.

In fact I am not aware of the concept being used for gaming at all.

I suspect that though fascinating the Game of Life not very well suited for gameplay. But perhaps something can be done..

Below is a link to my unoptimized and unpolished but playable alpha version of Game of Life as a Game?

There are a few game modes available.

In sandbox mode you can do whatever you like as long as you like.

In the Immigration mode there are the two teams, red and blue. Winner is the team that has the largest numbers of squares when status quo is reached.

As the supreme ruler of the GoLaaG universe you can favor whatever team you like and interfere at will. Just pause the game and click where you want to extinguish or recreate.

The Target mode is a bit like curling. The invading blues win if they occupy the target square. Your job is to stop the invading blue. Preferably with minimum interference.

Butterfly effect is just a special case of immigration.
If running undisturbed the campaign ends after 359 iteration with blue victory.

Your job is to make just one little change and produce a red victory. I'm afraid I haven't tried that myself but I assume it can be done.

Then you have some further options to play around. You can choose between three different sizes of your universe. You can also save and load game postions as .bmp image files. Highly inefficent but it gets the job done. I'm afraid the .BMPs won't scale well in high DPI mode though.

Feedback is highly welcome. I'd be happy to see some discussion about the Game of Life but I can't promise I'll make efforts to implement new ideas.

The program has very modest system requirments and runs well in Virtual Machines.
Compliled with Lazarus on 64 bit Debian Linux for Intel/AMD.

Good Luck and enjoy the GoLaaG!

https://www.jonax.se/linux/golaag.php

[davido]

Even in alpha, your game seems to offer an original and playful approach to Game of Life. In addition to allowing an exploration of cellular automata, you have added game mechanics that make the experience interactive and competitive. If you continue to refine and add features, it could really become an engaging and rewarding game to play. This is a great initiative that could interest both Conway fans, as well as those looking to explore game mechanics based on simple but powerful rules.

snaptube vidmate

[Ñuño Martínez]

Did I miss this one?

I did a GoL for a Jam in just one hour time ago (actually it was a weekend jam but my project didn't work so I had to improvise) but never find a way to make it an actual game.

I should test it.

[Jonax]

Thanks for your kind words and feedback Davido & Ñuño.

That Ñuño just made a GoL shows the GoL-concept is alive and kicking.

I guess one way to make it a game could be like a having a tournament or series where sevelal players get to fill opposite halfs of the playground and then having them combined for an immigration style match. A bit like of a correspondense chess match. Very slow.

Though this could be a challenge for us humans I bet modern AI would excel at such games and quickly find an optimal strategy..

[SilverWarior]

I guess one way to make it a game could be like a having a tournament or series where sevelal players get to fill opposite halfs of the playground and then having them combined for an immigration style match. A bit like of a correspondense chess match. Very slow.

Or perhaps give players ability to draw several smaller starting patterns. These can then be provided to game host as one larger bitmap split into multiple smaller sections where each section provides one pattern. Finally you go and spawn limited number of these patters from each player into one large world map and start simulation. During simulation these patters (provided that they are good) will start moving or spreading throughout the map and sooner or later interacting with patterns from other players.

Though this could be a challenge for us humans I bet modern AI would excel at such games and quickly find an optimal strategy..

Actually scientist spend a lot of time searching for most powerful starting patters using various methods even including various AI approaches but I don't thing anybody managed to identify this "one best pattern" yet. And bare in mind they were doing this on original game where you have only one organism type. As soon as you introduce multiple organism types finding best pattern basically becomes impossible due to shear number of possible paths the simulation can progress.
Why so many possibilities? Let us just say that you can draw starting patter that is 10 x 10 cells in size. Each cell can have any of the two states. So just the number of possible starting patters you can draw is 2 to the power of 100 which is 1,267,650,600,228,229,401,496,703,205,376 or one nonillion two hundred sixty-seven octillion six hundred fifty septillion six hundred sextillion two hundred twenty-eight quintillion two hundred twenty-nine quadrillion four hundred one trillion four hundred ninety-six billion seven hundred three million two hundred five thousand three hundred seventy-six
I believe this already exceeds the number of atoms in the know universe.

[Jonax]

So many possibilites. As you hint it's perhaps best for us humans to start with smaller patterns. The GoL is totally deterministic but very hard/impossible to foresee where the system will end. A bit like those chaotic weather simulations.

If making a game of this it should as you say benefit from a game host or a dedicated web-site.

Have you tried the 'butterfly-effect' challenge? I haven't tried that myself. In fact I just randomized that start pattern. I assume it can be done but haven't tried it.

Can anyone suggest a better/more interesting starting pattern for the the butterfly challenge?

[SilverWarior]

Have you tried the 'butterfly-effect' challenge? I haven't tried that myself. In fact I just randomized that start pattern. I assume it can be done but haven't tried it.

I have not heard of "butterfly-effect" challenge until now. I guess I might be giving impression that I know a lot more about Game of File as I do
While I did do some researching about Game of Life years ago I must admit that one I realised how "deep rabbit hole" learning about GOL is I actually gave up. I had no interest in spending days if now weeks learning about this.
Instead I shifted focus in learning about other games where user input have much more "direct" effect. For me personally those are much more interesting. But I still understand how some people can have huge interest in GOL from programming perspective as it teaches you the basics for simulating cause and effect in a game. That is one of the reasons why many computer science programs do have an exercise on recreating GOL.

[Jonax]

Well.. I invented the butterfly-effect challenge for this alpha toolbox version of GoLaaG. Though the idea is not very original. It wouldn't surprise me of other people already have done things like that decades ago.

In this case you press the top right button with the the text butterfly effect. Then you get a starting position which I admit to have once generated randomly but is always the same. If you press the on button the GoL does its thing and "If running undisturbed the campaign ends after 359 iteration with blue victory."

Very boring to wait 359 interations. Suitable to go get coffee or something. Anyway the challenge is to make as small a change as possible to make the red team win instead of blue.