Branching and Experimentation in Design

Post by: Grant Rodiek

I’m curious how many designers are aware of sunk costs, or are willing to build things, admit failure, and completely try anew? There is an urge to preserve an idea and tinker with it until something emerges, but often times if there’s no spark, there won’t ever be one. Designs, like dating, should be given a few tests and then discarded. It’s not me, it’s you and your fiddly rule set.

I’m not advocating you completely discard an idea, though I’ve done that quite a bit and I think it’s healthy. What I want to discuss is the notion of branching to take a cool idea, revise it, and approach it from a new angle.

Branching is when you put a fork in your design, you zig (instead of zag), and pair a new layer with your original kernel. This kernel can be a thematic one, like “I want to be a pirate on a ship,” or a mechanism, like “I want to assign workers, but each can be promoted to have a higher output,” or a combination of that, or you might be trying to improve upon a previous design you’ve played. There are countless origins and far worse sentences I can offer. Don’t tempt me.

The choice to branch or toss really comes down to the quality of work you have at this moment. Ask yourself, and do so sincerely, “Is this game worth saving? Is there something here?”

I would say yes if you can match a few of the following:

  • The idea is unique. There is nothing like it.
  • There is an obvious, great moment. Not the game necessarily, but there is something very cool you’ve witnessed.
  • People get it. They aren’t fighting the game’s mechanisms, but they get the conceit.
  • You’ve thought of 1 or 2 brilliant thematic abstractions.
  • It matches a market/business opportunity beautifully. This one is tough to know as a new designer without great reach. It’d would have been tough for Love Letter to catch on without AEG’s muscle.

If you don’t match any of those, you probably want to toss the game. Just move on. Return to it in the future, or just do something else. But, if your game has a spark, but isn’t quite worth developing, consider branching it.

When branching a good idea marred by bad friends, your game is like a snail seeking a new home: highly averse to salt and in need of better surroundings. A recent design of mine needed a branch. I’d tested it 3 times. Each test I’d changed a few things, but the core remained quite steady for 3 plays. I liked the game. People were getting it. It had a neat hook. But, the parts weren’t coming together, there wasn’t enough tension, and the decisions weren’t as meaningful as they needed to be. I also didn’t feel the level of interaction was up to snuff.

I re-examined my core hook, which is letting players dynamically create their action menu. After that? I’m throwing everything out. I’m distilling it to its essence and trying new layers. As it turns out, there are 2-3 ways for me to express my core desire, which is to give players a dynamic set of decisions. The one I had? Totally works. Is FINE. Not great, but I have the rules set aside to return to.

How do you know how to branch? If something wasn’t working with the first pass, how can you make productive steps forward to improve it on the new version? Firstly, you need to know your goals for the game. What is most important to you? If you know your goals, you can identify what serves that goal and what doesn’t from the current game.

Next, seek to identify not what worked, but what didn’t. If you cling too much to what worked (or what you think worked), you may have a difficult time making a new branch. For example, technically, my core mechanism in the first iteration of my game works. But, it might also be the root of the problem!

Therefore, examine what doesn’t work. For me, this included:

  • Low interaction.
  • A lack of arc. The game felt highly repetitive.
  • Low tension from decisions. Nothing was nail biting.

Begin examining how to address each of those problems in any way, ideally through the lens of what you already have, but being willing to set those things aside. This will help you identify the root causes of the issue and allow you to re-examine things through a new light. This is how you figure out the new direction to take.

This is a difficult topic about which to blog, because every game is unique and the circumstances in which you might use this process vary so wildly. I don’t want to beat this point into the ground. Therefore, let’s quickly review.

  1. Give your game 2-3 tries. Does it have a spark? Is it singing? Keep going. If not, it might be time to branch.
  2. How do you know whether to branch or discard? Well, ask yourself whether the game is worth fighting for.
  3. If you think it’s worth branching, re-evaluate your goals. What is sacred in this design?
  4. Once you know what’s sacred, take a magnifying glass to what’s currently broken. How are you currently not hitting your goals? What’s falling short?
  5. Reconsider your mechanisms and choices through the lenses of the shortcomings. Fix your shortcomings however possible, while being flexible to revise and toss things away.
  6. Test again.

I hope this was useful. Really, just leave knowing that it’s not a mistake to throw things away, try again, and branch. You should be experimenting as much as possible in your design until you find something that really, honestly, sincerely, awesomely works.