How to start a game design?

Last week, I was listening to an episode of Marc Rosewater’s Drive to Work podcast episode #469 (for those that don’t know he leads game design for Magic the Gathering) where he talks about how to start a game design. Or at least how he starts it. It came on a very convenient time to me, as we had to give up a couple of game jams a while back cause we could not get to a first design/prototype in time.

One of my favorite cards of my one of my favorite games

On the podcast, Marc describes his idea about The Golden Mechanic. A mechanic that excites you as a designer and a player. He talks about how game design is really complex and abstract and the Golden Mechanic works like a float, bringing you back when you start to dive too deep. It helps drive you forward and stay focused during moments of doubt.

This idea was very inspiring to me as I could relate to it. Some successful games I made had a Golden Mechanic. But, since other great projects did not, relying on lore constraints or even tech ones like: “Let’s make a game that uses genetic algorithms?”, I decided to do a little research of my own, and gather how other game designers start their processes.

A little research of my own…

***DISCLAIMER***: This is an non comprehensive opinionated list. It’s also not in any specific order.

1. Start with player’s feeling

This one date way back, as it was Mark Venturelli, Critical Studio lead designer main strategy. The idea is to start on the intended feeling the game should arise on it’s players and then design from there. Scared, Rushed or Surprised are common examples. And the key is that every aspect of the game (mechanic, lore, etc..) should support the intended feeling.

I have to admit that this never really worked for me. It’s just a bit too abstract and ends up getting me stuck more than helping.

2. The world building trip

What about starting with really deep world building? Thinking about the characters personalities, back stories and goals. Environment details and stories. Writing the story should get you comfortable enough to understand better how the game should be played. Is it a platformer? An fps? Strategy? All decided after the story.

To be honest, this seems super hard to me. Probably because of my Computer Science background. System design are more of my thing than storytelling. Anyway, sounds really interesting and I will give it a shot for sure.

3. Build a polished toy

This one feels really Nintendo-y to me.

I heard from MANY game designers that this is their main strategy for designing video games. The idea is you start by building a playable prototype with a core set of mechanics and then polishing it to the point where it’s fun to just play with, even though there are no real goals or challenges (definition of toy).

This relates back to the Golden Mechanic idea, adding polish and a lot of playtesting on top to nail down the game feeling before you move to other relevant decisions like: “What is the game world like?” or even “How do you lose?”.

One example I like here is Holodrive from Bit Cake. Today, the game have a nice player base and it all started with a simple toy about: “Shooting each other on a 2D platform online arena shooter. WITH JETPACK.”. Check out it’s first videos and it’s current version:

First recorded video
Current version

4. What about the moneyz?

When you are a professional game developer, it’s hard for you to run away from the fact that you want to make money with your games. Really hard. So, why not embrace it and start from a market opportunity? Or maybe a innovative monetization strategy?

Call me purist, but I think there is value on starting the game design searching for fun, not revenue, but I heard stories of games that started from a player base insight, like: “There is no game of this genre on this platform” or “There is no game of this genre in this theme”. All that could be opportunities.

At Rock Milk, it’s a bit hard to keep both things apart as Diego and I are involved in all steps of development. We end up mixing up business intelligence with game design frequently, but I admit I try to make a conscious effort to avoid it in the beginning.

5. Mechanics Tinder

Another interesting approach I heard is to have a list of interesting game mechanics and mixing and matching them looking for inspiring combinations.

It is the gamiest way of designing games I ever heard and I like it. It probably works well as a prior step to Golden Mechanic. Another one I plan on trying in the future. A website that comes to mind here is Three Hundred Mechanics, an 10-years old page looking to list 300 different game mechanics, creating a dictionary for game designers, what is proving really hard as they are on 250s 10 years in.

In conclusion..

As expected, there is no one right way to start a game design. There are plenty success stories with all the strategies I mentioned before and even more failure ones. They are all nice ways to create constraints that will help you move forward and design fun games. The key is picking one that makes you comfortable and start creating. You will be using many of this strategies at some point anyway.

Aaah, almost forgot, you should test and iterate a lot. Every designer I know says you should get to a playable fast and test a lot. It’s a big deal!


P.S.: I would like to thank: Marco Portugal, Lucas Thiers, Eduardo Lahmut, Thiago Baptista and all the other game designers for sharing their opinion and strategies that contributed to this article.

If you read so far, thanks very much! If you have questions or any other strategy that works for you, I would be thrilled to chat. Please reach out on Facebook and/or Twitter.

Matheus Almeida, Best Half and Head of Sorting Algorithms @ Rock Milk