Up in the right corner I've added a new link to a page that is pretty bare right now. One of the things I'm interested in that I want to feature prominently in my games is generative music, and I've put a lot of thought and research into different ways of generating music out of individual elements. One of the stranger methods I've come across is something called "Bytebeats."

This method takes the concept of generative music to its purest, most minimal form. The entire piece of music is defined as a function of time, and the output is just the sound of that function applied 8000 times per second. The results have quite a lot of variation, and can become quite complicated. The functions that describe these pieces of music are so simple that people often talk about "discovering" them rather than "composing" them.

The Compositional Process

There are certain atomic operations that can be composed, which contribute different qualities to the resultant music. You can't put these together in any of the ways familiar to composers: you can't string them together into a melody, nor can you even really play two of them at the same time. It is more like you only have effects to work with, and no sounds; you apply effects to silence, and the result is somehow music.

Some of the more important elements turn up in many different bytebeat pieces, often multiple times. The one I used most often in the piece above is something called "sierpinsky melody" after the shape that the waveform makes when you graph it. (You can play around with these sounds and see them visualized here.) The basic form is (t&(t>>k)) where t is time and k is some constant. The '&' is a bitwise "and" operator, and '>>' is arithmetic shift-right.

Just with this one building block, there are quite a variety of different effects. Different values of k can have wildly different sounds, and the 't' on the left side can be multiplied by a constant to give subtler variations. (3*t)&(t>>8) was one of the first Bytebeats to be discovered.

Another important building block is the pattern t*(x&t>>k). The specific example t*(42&t>>9) was another early discovery that has a distinctive melody. It should be remembered though that these are distortions of a line, they aren't fixed melodies. They can only be combined with mathematical operations, not musical operations.

The Music

The music produced is generally harsh and electronic sounding, as might be expected. Some of it might not even sound like music at all to some people; I have always had a taste for very abrasive music, so it fits with my taste quite well, but I know that I am unusual in that regard. It also tends to be fairly repetitive, I'm quite proud of the piece I linked above because it is less repetitive than most of what is possible with this technique, but it is still a very harsh and repetitive piece of music by any other standard.

It reminds me quite a lot of the music of Masami Akita, who performs under the name "Merzbow." It's actually a little bit hard to call what he does music, but it is a deliberate arrangement of sound to create an intentional aesthetic experience, so it's pretty darn close. I quite like listening to his stuff; he's very prolific but my favorite of what I've heard is an album called "Cycle". It has the same harsh electronic sounds of Bytebeats, but with a more deliberate organization, and he often incorporates non-electronic sounds like car horns, dogs barking, and other stuff like that.

This kind of abrasive noise-music is appealing to me for a few different reasons, but the one I find easiest to articulate is the way that it changed my relationship to noise in general. Listening to these deliberately chosen noises has tuned me in to hearing the background noises of the city in a new way. Rather than tuning out things like passing cars and rattling air-conditioners, I can listen to and appreciate the sounds they make as though it were music. It isn't the same as listening to music, but it's the same kind of deliberate, careful listening to all of the sounds being made, and the way that they interrelate.

This sort of pure listening is very important to the way I appreciate music. It isn't the only way that I listen, nor is it the most important part of listening to music, but it informs both the kinds of things I listen for and the kinds of things I compose. In creating generative music, one of the things that is very important for me personally is that the sonic texture does not get stagnant. It is easy for generative music to fall back on the same sonic textures every time, because the resources from which the music is built are necessarily limited. Bytebeats are somewhat promising in that regard because the music is built up from parts that are smaller than timbre and tonality. They all end up having the same sonic character anyway, but they do so for different reasons, and they may yet prove a fruitful source for new generative sounds.
I've heard a lot of people, for a lot of different reasons, say to me that they did not feel like anyone was making the games that they wanted to play. Some people said it was because the things that they liked most in games were always treated as afterthoughts, peripheral to the heart of the game. Other people complained that they just couldn't identify with the playable characters in most games.

One story I've heard a few variations on is the father who is trying to introduce his daughter to the games he grew up with, and finds out that she wants to be able to play as a girl, which in many of those games was not an option. We got some gems out of that, like the Donkey Kong hack where Pauline has to rescue Jumpman, and I'm trying to do the same thing: spread the privilege that I have as a target demographic, share the particular sense of rightness that I get just because I am the kind of person that game designers expect to play their games.

More than a few people have said that they wanted to be a bigger part of the gamer community, but they found it so hostile that it just wasn't worthwhile. I've seen a lot of problems caused by the incredible sexism in the community, in the games, and in the industry. Sometimes games bring out the worst in us.

But I've also seen the great good that games can inspire. I've seen incredibly strong, friendly communities build around games, and I've seen those communities come together to do real good. Games have the potential to bring out our best.

I want to make games that reflect the world I want to live in. I hope that those games will encourage the world we do live in to move a little closer to the world I want to live in. I've seen how much power game designers have over the kinds of communities that form around their games, and I've seen the way that the stories and ideas in games can shape our thinking, for better or for worse. That's what I'm trying to do.
I am a big fan of games that allow or encourage unstructured play. I believe that, given the means, gamers can be far more creative than a single designer or team. I'd like to highlight a few different ways that games have been particularly successful at giving the player a versatile toolkit and enough freedom to come up with their own objectives:


 Fuel is the game that most perfectly captures the heart of what I mean by "unstructured play": give the player some physics, suggest some activities to get them started, then turn them loose in your world. It is an off-road racing game with an open-world map; races and challenges are scattered around the map, but the entire game takes place in a single area.

The game does a few things that particularly favor unstructured play. The first is that the game gives you the ability to stop whatever you are doing at any time, and go do something else. This includes quitting a race midway through to go investigate some hills. There is always an option to enter free-play mode, and it is always clearly marked. From free-play mode, you can either drive to the location of a race, or you can teleport back to the base-camp and launch any of those races and challenges from a menu instead.

The entire user experience is designed to emphasize and encourage player choice, which is something that a lot of games strive for, but it does it in a very different way. Whereas a role-playing game emphasizes player choice by creating meaningful consequences, Fuel does away with consequences entirely. The player's options are always open.

Another thing that facilitates unstructured play is the physics-based driving. Vehicles are effectively indestructable, and there is always a reset button, so the player has a lot of leeway to play around with the way that different vehicles interact with the terrain. There are a lot of different combinations of surfaces and slopes, ranging in scale from large gulleys and bluffs to little dirt-heaps and creeks. The different surfaces have different amounts of friction, and the vehicles in general don't have a lot of grip, so there are some interesting things you can do with sliding around, spinning out, and of course trying not to spin out during a race.

Street Fighter IV

Street Fighter is not a game known for its unstructured play. It is a game with clearly defined goals, and there is always a winner and a loser in every match. Even so, it is a good example of a game that gives players a lot of different things to play around with, and it's a good example of how gamers can take a system like that and do incredibly creative things with it.

One Youtuber who goes by the name Desk has come up with some particularly creative videos by editing together footage of himself performing very tricky combinations of moves to create unusual effects. "Combo videos" are a staple of fighting-game culture, but Desk goes beyond just stringing together long sequences.

Here is a video where he exploits a quirk of the game that once an opponent has been knocked out, and the game is officially over, it is sometimes possible to fit one more move into the end, while the game is displaying the "K O" screen. It took a lot of work to set up each shot, because the conditions under which this happens are not particularly common, and he always figures out a way to lead up to it with a combo that is as flashy as anything else in the video.

In another video, part of a series, he takes the armor mechanic in the game, whereby a character can absorb a hit without being stunned by it, and uses it in a number of highly impractical but amusing ways.

Another group took that idea even further by creating a series of videos exploring all of the different ways that characters can avoid taking damage. Most of these techniques have disadvantages that make them useless for competition, but that's not the point: players create their own objectives and work very hard to reach them even when the game presents itself as rigidly structured.


Skyrim takes the idea of unstructured play in a very different direction from either Fuel or Street Fighter. Like Fuel, it has an open world and does not prevent the player from going where they want, when they want. Like Street Fighter, it has a well-defined structure that nevertheless encourages exploration.

The skill system in Skyrim gives the player a lot of leeway to do what they think is interesting. You have to do most of the activities in the game if you want to make a lot of progress, and most of your options are restricted at the beginning; for example, the powerful "shouts" that players gain can only be obtained by playing the central storyline. Even so, one does not need to do everything, and in fact it is not very practical at all to try to gain every skill.

Leveling up a character always involves a choice of some new tool that you want your character to have available to them, and there are a large number to pick from. This gives the game a good sense of variety, because there's always some new toy that looks interesting, and you are always getting new mechanics to play with.

Much like Street Fighter, even though Skyrim's design strongly encourages a certain approach, it does not force the player to do anything. The fact that players can modify almost every part of the game's structure enhances this. I knew one person who made it into a survival horror game where they were constantly threatened by hypothermia. Another person used graphical tweaks to change the game's appearance, the color balance, even the way light behaved; he would switch some of these settings around, look at the results for a while, then shuffle them around a bit more to get a totally different look and feel.

On top of all of that, the designers saw fit to put in a large number of tangential activities for the player to try. One could earn a living as a migrant farm laborer, or serve as a vigilante fighting bandits, or as a bandit robbing travelers. The player can join virtually any of the many different organizations in the game, or none of them.

The notion of "unstructured play" is a very broad one, and it imposes very few constraints on the game designer. As long as you give the players some physics and some control, gamers will be creative enough to find new and interesting uses of your game.
I'm still grinding away at Action Potential, but I've added a new mile-post, I'm going to put together a smaller game with the same logic. This will be a test of sorts, to see if I have what it takes to finish an entire game. It will also be a sandbox to refine my techniques and get more fluency with the whole process.

I've also taken on another side project, a big ambitious Unity project that someone proposed on Reddit. Working on someone else's project will give me another angle from which to view this mountain that I'm climbing.

I have another, better blog post in the works, about the value of unstructured play in games. I want it to be good, but I also want it up soon. We'll see which of those desires wins out in the end.

And finally, an update on the drawing tablet: I'm using it every day so far, taking a few minutes to draw circles until I'm too frustrated to sit still any longer. I have struggled with drawing my whole life, and I am not optimistic about this attempt, but so far it is going as well as I think it could go. I think the next thing to do is get a sketchbook that I can take on the bus. I have a long layover most afternoons when I could get some real practice in.