Prototyping at Hyperspeed

This week, our game designer Şahin wrote an article about prototyping! You can read his article below:


Sahin_3

Filtering game ideas in an effective way is one of the challenges we have to deal with as game developers. Occasionally, companies develop different frameworks that are tailored to their needs. We hear about test launches in 6 months or even in 3 months. For smaller companies however, this is a long time. We need a more radical solution. We gave a lot of thought into how we could reduce the time between the conception of the game idea and the moment we acquire our first metrics from it. Since we are operating mostly in the casual games genre, there should be some shortcuts.

What about a month?
Too long.

What about 2 weeks?
Meh.

3 days?
Yeah, that would work!


Sahin_1

Fridays at Gram Games

Let’s have a look at how we deal with prototypes at Gram Games. We believe that the way to have a good idea is to have lots of ideas. At Gram Games, Friday is the prototyping day of the week. All of our employees become a prototype creator by setting their daily responsibilities aside and working in teams to develop on a new game idea. At the end of the day, we get together at our cafe to demo the test build. The main purpose of this day is to continue building up on what sets Gram apart from the top mobile gaming studios: our unique culture and creativity. This also allows us to work on creating a game that our employees collectively agree on building.

In short:

  • People present their ideas
  • We vote, select and proceed with the best ideas
  • We allocate our resources according to employee preference to ensure people work on what they feel excited about.

We ensure to include everyone on the business (HR, marketing, finance, and more) and studio side on the development process because for us, every idea counts.

So, 3 days of prototype production would be 3 weeks for Gram, since we are only working at Fridays on prototypes, but 3 weeks sounds nice for evaluating a game idea.

Sahin_4

Taking Advantage of the Genre

We mostly do casual games and as a general rule of thumb, casual games must have a high day-1 retention. While a high day-1 retention does not guarantee a very successful game (as there are many other variables contributing to success), a low day-1 retention is a clear sign for us to kill the project.

So, day-1 retention is the main metric we check for the first launch, and the first threshold we have to pass.

Sahin_5

Defining an MVP for Targeting Day-1 Retention

Things get a lot easier when your scope gets narrower. We will only check if the core mechanic is fun or not, and we’ll do it in a metric driven way instead of a hunch driven one.

Now that we’ll only consider day-1 retention, we can focus on features that will affect day-1 retention which are:

  • Core Mechanic
  • Main Loop
  • Onboarding (tutorial) -if needed
  • Retention Layer (enough content for 2 days)
  • Minimal polish.

Be advised that we have not done anything about the following yet:

  • Monetization Layer
  • Medium-long term retention layer
  • End-game layer
  • High polish

For the Polish Level: The game should not feel like an early prototype for the players and they shouldn’t feel like test-subjects while playing your game. In addition, minimalist and flat design can save you a lot of time.

Sacrifice Worth to Mention: Some games must be created with the monetization layer from the very beginning if it’s a fundamental part of the gameplay experience. Monetization layer should be added to the scope for those games.

Sahin_2

Introducing Fake-Launch

This is a term we’ve made up to differentiate it from soft launches. It’s is important to understand that this is not a soft launch. Soft launch will be the phase we’ll go into after we succeed in the fake-launch.

What we do in fake-launch is:
Acquire around 1000 users in 3-4 days.
Check day-1 retention afterwards.
If it’s > 45%, the prototype goes into production pipeline.
If it’s < 45% but > 35%, give it 1 more day of prototyping effort to pass the 45% threshold.
If it’s < 35%, kill it instantly.

It’s important to check other engagement metrics as well to spot fundamental problems that can cause low retention:

  • Session lengths
  • Gameplay lengths
  • Gameplay counts
  • Level funnels -if any

Going Dirty

Since we are not creating the whole experience, we’ll have to go dirty sometimes:

  • We can show false premises to keep players play the game. Example: In an endless runner, if we want to keep players play the game to unlock some stuff, we can display a chest, that can be opened for some coins, but
    we must not create the content for the chest. Yes, that will be frustrating for the players once they find out that we cheated, but these players are our playtesters anyways.
  • Do not publish those games with your company brand. Players may have unpleasant experiences after day-1,
    so it is important to refrain from any activity that can negatively impact your company reputation.

sahin_yalabik

Last Words

So far, this approach worked really well for us. It’s been 5 weeks since we began using it, and 2 games made it into production while 7 games have been dismissed. This is a living and breathing framework for us and we constantly iterate it to ensure it fits our company culture and workflow.

The key question you should be asking now is “will it work for me?”