The Origins of BEAT DRIFT: How Bad Prototypes Can Be Good

Digital Uzu’s first game, BEAT DRIFT, was released for iOS on December 20th, 2013. It was done in collaboration with a small group of friends under the team name LunarPixelGames.

BEAT DRIFT gameplay

Screenshot from BEAT DRIFT

If you’ve played BEAT DRIFT (and of course you have, right?!), you will know that it’s a fast-paced, twitchy, block-dodging, evasion game. The feedback has been extremely positive, and we have gotten great ratings and reviews. I was thrilled to hear some people even say that it’s the best evasion game they’ve played since Terry Cavanagh’s amazing Super Hexagon.

The thing is, though, the initial prototype for BEAT DRIFT was very bad. We weren’t sure what type of game we wanted to make, and the prototype felt like a random collection of convoluted ideas thrown together. The first prototype played more like a twin-stick shooter than the twitchy-evasion game BEAT DRIFT has become.

Through this post, we’ll take a look back at the origins of BEAT DRIFT, and show how we were able to gradually refine our prototyping process, which in turn allowed us to test our ideas more quickly and efficiently. In the end, we were able to take something “good” out of our initial “bad” prototype, and use this to make a quality game that we are proud to call our first release.

The Released Game

For those who haven’t played BEAT DRIFT yet, this is what it looks like in action:

The video cannot be shown at the moment. Please try again later.

BEAT DRIFT release trailer

The gameplay concept is simple:

  • You control a pulsating, sphere-like object that is traveling at high speeds down a 4-sided tunnel made of blocks.
  • Solid blocks appear in front of you in various patterns.
  • If you collide with a solid block, you die and restart the level.
  • You can touch the left/right sides of the screen to move the player clockwise or counterclockwise to dodge the solid blocks.
  • The goal is to survive for 60 seconds.

In addition, the camera rotates in sync with the trancelike background music, and the solid block patterns get increasingly difficult, making it a deceptively challenging game.

The First Prototype

Jumping back in time a couple months, this is what BEAT DRIFT originally looked like in the first prototype:

The video cannot be shown at the moment. Please try again later.

Twin-Stick Shooter version of BEAT DRIFT

BEAT DRIFT was originally a twin-stick survival shooter, inspired by games such as Geometry Wars and Minigore. The only thing we were certain about at the time, was that we wanted to use blocks (similar to Minecraft) for the visual style and level construction.

Working with blocks (or cubes) has a lot advantages. Not only do they present interesting visuals, but more importantly, they can be created programmatically. Working on a small team with only one part-time artist means you have a very limited art budget. This forced us to use our resources wisely, and do as much as we could via code.

In this first prototype, the left joystick was used to move the player around the block world, and the right joystick was used to shoot and destroy blocks. We added all sorts of things to the prototype: enemies, different weapon types, collectible items, procedural maze level designs… We spent a lot of time experimenting and trying out new ideas, but it never felt quite right.

There were a couple presentations a few years ago at GDC (one by Chris Hecker and the other by Jonathan Blow) that discussed what makes a good prototype. Basically, a good prototype should strive to answer one clear and concise question.

In hindsight, the initial prototype for BEAT DRIFT was not very good. By adding so much “stuff” to the prototype (enemies, weapons, power-ups, …), we were trying to ask a lot of different questions at the same time, and none of the questions were particularly clear. Like many failed experiments, though, this “bad” prototype gave us inspiration and led us towards better ideas.

We knew there was something fun within the prototype – we just weren’t sure what it was. As the team played it more and more, we realized that the most interesting thing was actually the simplest thing: smoothly moving the player around the world and destroying blocks was a lot of fun, and it was something we wanted to pursue further.

The Second Prototype

Based on what we learned from the first prototype, our next one would aim to answer a single question: “What would it feel like to focus only on moving through and destroying blocks?”. This is what we created:

The video cannot be shown at the moment. Please try again later.

Vertical-Scrolling Block Dodger version of BEAT DRIFT

We wanted the game to be fast-paced, so we kept the player constantly moving forward. We also wanted lots of block destruction, so we destroyed a block whenever the player came into contact with it. We turned the prototype into a simple vertical-scrolling block dodging game. Through reducing complexity and focusing on our core mechanic, we were able to drastically simplify and refine the gameplay.

Our current prototype wasn’t perfect, but it felt good and it answered our original question of “What would it feel like to focus only on moving through and destroying blocks?” (Answer: It feels great). We were headed in the right direction.

The “Final” Prototype

The second prototype had invisible walls on the left/right edges of the level to prevent the player from moving outside of the level boundaries. This had the unintended consequence of making the player feel trapped, and drastically slowing down the pace of the game. We knew we wanted to make a fast-paced, smooth game, so we had to fix this problem.

We tried wrapping the player around when he went off the edges of the level (i.e. if you move off the right edge of the level, you appear on the left edge and vice versa), but this felt very unintuitive and confusing. It was hard to visually understand from the player’s perspective.

We needed to try something different, so we asked the question, “What would it feel like if we wrapped the level around on itself (connecting the left/right edges)?”, and this become our next prototype:

The video cannot be shown at the moment. Please try again later.

4-Sided Tunnel Block Dodger version of BEAT DRIFT

This prototype (which is almost identical to the gameplay of the released version of BEAT DRIFT) proved to us that moving through a 4-sided tunnel at high speeds while dodging blocks was very fun. Since the level no longer had invisible walls to constrain the player, it fixed the “slowness” that we were feeling in the previous prototype, and we were very happy.

After this, we continued prototyping and asking many questions. The 4-sided tunnel was made semi-transparent so that it felt more open and looked better from an artistic perspective. Because of this, the destruction of blocks became hard to see, and unfortunately we ended up removing it. We experimented with and refined a lot of different ideas, but the core mechanic (from our first “bad” prototype) of smoothly moving through a world of blocks at high speeds remained the same.


Released version of BEAT DRIFT

Game design is a non-scientific and very iterative process. Even though our initial prototype could be considered “bad” since we didn’t focus on answering a single question, it definitely was not a failure. It led us towards many good ideas and taught us how we can make better prototypes more efficiently – not just for BEAT DRIFT, but for all our future projects.

Tagged with: , ,
Posted in Development

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code class="" title="" data-url=""> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre class="" title="" data-url=""> <span class="" title="" data-url="">