How to Hack in Darknet
Welcome to the Darknet dev blog! I'm planning to write here at least once a week to give some explanation about what I'm making and how I'm making it. Since Darknet is an evolution of Ciess, my old 3-week-game-jam prototype, and some of the people reading this have played that version (hi!), I thought it would be best to inaugurate this blog by describing the biggest gameplay change I've made since then.
First, some background: the world of Darknet consists of a complex network of nodes, each of which is a sphere floating in space. You can zoom into any one of those nodes to hack it, which involves beating a procedurally-generated challenge. In Ciess, it worked like this:
The blue squares are security programs. You can convert a limited number of them, which makes them turn red and start expanding outwards. The remaining security programs try to destroy that red mass. This mechanic was good enough for a game jam, but it got boring quickly. After all, nobody puts on an Oculus Rift so that they can look at a flat square. There was also a more important problem: the optimal strategy was almost always to fire all of your hacks into a corner as fast as possible. Once players realized this, the tactical aspect of the gameplay sort of fell apart.
So, I threw most of it away and built something new! Here's what it looks like now:
Some things haven't changed; you're still shooting blue security programs to convert them into hacks, and the hacks still expand outwards while the security tries to destroy them. But there are a few key changes:
1) The most obvious difference is that the hacking takes place on a triangular tiling that wraps around the player, giving more of a VR-friendly 3D vibe.
2) The security programs are invincible. You can't beat them in a fair fight anymore.
3) Instead of trying to wear down the security programs, your goal is to sneak around them and get to the core before they can stop you.
The idea here was to find an aesthetic of gameplay that fits the theme of hacking. The security appears impossible to beat at first, since it always destroys your hacks one-on-one, and they move twice as fast as yours. But the key point is that your hacks are constantly multiplying, like a virus, while the security can only be in one place at a time. You can't fight them head-on, but you can outwit them, slip around them, and get to the goal before they can stop you.
Each security program makes decisions based on a simple AI algorithm. In fact, their behavior is entirely deterministic. That means that, if you were Rain Man, you could calculate exactly how the hack would play out ahead of time, which excites me in an embarrassingly geeky way. However, since the game's behavior gets complex quickly, most people will need to develop rules-of-thumb and intuitions about how the system will behave. For me, that's the mark of a good strategy game.
One nice feature of these challenges is that they can be procedurally generated and they can be tweaked for different levels of difficulty. Also, they can change quite a lot in size. In the biggest nodes, the challenges include thousands of vertices, and they get so big that they start to wrap fully around you. Check out the image at the top of this post to get an idea of how huge they get.
There are some other twists to the mechanics that I'm not talking about quite yet, but this is the baseline. There's also a lot of strategy involved in choosing which nodes to hack: the "network layer" of strategy. I'll be talking about that later! But these mechanics are the most basic, most common unit of gameplay in Darknet, and I'm pretty damn excited about how they turned out. :)