The 7 month social game gamble: 2 months design, 5 months programming, but is it fun? There is a better way

Matthew Warneford
21
August
2011

The recipe for making a new game is pretty simple: spend 2 months designing the game, 5 months of late nights and weekends programming it, fix all the bugs, and release a beta.

There’s nothing wrong with recipe, but you’ve got to remember to stick your finger in the mixture and give it a taste. After all, you wouldn’t bake a cake and never taste what’s in the bowl, it might need some more sugar! Yet that’s what most game developers do. They’ll happily spend a bucket of cash building a game for 7 months, and only after 7 months do they find out if anyone really likes the game. This is the 7 Month Gamble. There is a better way.

Although we make kids games and virtual worlds, the ideas in this post are still relevant for you folks making social games.

In an ideal world you’d want to know if your game concept is going to be a hit before you start to build it – then there would be no need to delay buying that yacht. Unfortunately, while we get the time machine working again, there’s no way to know for sure if a game’s going to be a big hit. So in the meantime, our Dubit research guys have developed a couple of processes for testing the game concept well before the beta release.

The easiest way to get some feedback is to just ask potential players what they think of the game concept. Does the game sound fun? Do they like the theme? What about the narrative, does that sound exciting? The thing is, it’s a pretty bad solution. In fact, it’s not just bad, it’s toxic. The reason being: people lie. It’s not that they mean to lie, maybe they want to make you happy, maybe they think they’re telling you what you want to hear, never-the-less, the outcome is the same. Don’t trust what they tell you, observe what they do.

I heard this story a while ago and there’s no better example for choosing observation over listening. During the early planning stages for the then-new Sony Walkman, a focus group was asked, “What color Walkman do you think most people would be more likely to buy: black or yellow?” Overwhelmingly, members of the focus group responded that most consumers would purchase a yellow Walkman. At the end of the focus group, everyone was allowed to choose a black or a yellow Walkman to take home as a gift for participating. All but one chose black.

The lesson is simple, watch what consumers do, not what they say they’ll do. Only now I sound contradictory. Earlier I admonished the 7 month design and hope strategy, and now I’m saying you can’t just ask players what they think, you’ve got to observe them doing it. The trick is to observe them before the game is finished, before the 7 months is up. The sooner the better. We do this by “ghetto testing” the things that make the game unique.

With every game or virtual world we create there are three things that are going to be completely unique:

  • The theme. Is it set in space, underwater, a secret ninja society hidden among rooftops of ancient Japan? We actually thought the latter would be a great idea, so much so we even started making the 3D artwork, then we realized young girls don’t much like ninjas – this is the danger when men design a game for girls without following the testing ideas in the post!
  • The narrative. What’s the purpose of the game, is there a narrative for the players to progress through? Perhaps there is an evil blue teddy bear called Dr Cuddles who is trying to turn the world hairy? Yes that’s a real example from our virtual world we created for the BBC.
  • The golden mechanics. What are the major game mechanics, are players caring for a pet, collecting, battling, exploring?

So “What is Ghetto Testing?” It’s a simple, cheap, market test. It can be done fast – today if you want to. And it’s done before the game is built. Done properly you’ll quickly know if your theme, narrative, and golden mechanics actually work. The basic process is the same for all three.

Lets walk through testing a game narrative. Imagine you’re building a kids virtual world. Thats 5 months of development, much more if you don’t use an existing platform like ours (bad plug I know). You want to build a good virtual world, one the kids are going to love, and so you’re trying to build an exciting narrative into the game. The problem is it’s a pretty big part of the development and you’d rather know if the kids like the story before spending all those hours programming it. What do you do?

Step 1: Create three simple videos, one for each of your narrative stories. These don’t need to be fancy videos, just colored sketches and simple animations. Do spend money on good voice over. Voice over makes a narrative come to life.

Step 2: Create the game homepage and the sign up form. Duplicate the homepage for each of the three videos.

Step 3: Set up a Google Adwords account. Buy traffic for the same keywords your players would use to search for your game. Send that traffic in equal measures to each of the three home pages. Spend no more than $10 per day per homepage.

Step 4: Record how many players register after watching the different videos.

That’s it. Send traffic from Google to the three home pages. The only difference between the three pages are the narrative videos. See how many players sign up from each of the home pages. Which ever video produces the most registered users is the narrative to write into the game. Easy and powerful.

What’s so great about this approach is that players think they’re actually joining a game. They don’t know they’re part of a test, they don’t know the game doesn’t exist. They’re signing up because they want to play. This is observation at it’s best.

Where do the players go after they’ve registered? Nowhere. Don’t worry about upsetting people. Most wont remember who you are. Still worried? Just don’t reveal your brand, make up a new one!

This is exactly the approach we used for our Muddle Earth game.

It can be applied to testing the theme and testing the golden mechanics too. For the theme create a range of different home pages, each homepage should be representative of the theme. Record the ages and gender of the players who sign up, and run with the theme that converts best for the target ages. But, remember, it’s not the absolute number of conversions, it’s the percentage of conversions for each age – what percentage of  visitors who land on the homepage actually convert into players?

For the golden mechanics the process is similar, but a little more work – we like to build a virtual world’s golden mechanics as standalone mini games. Sure, it’s more work than testing the theme or the narrative, but it’s also a lot cheaper to test the golden mechanic as a mini game than waiting until virtual world is complete!

Imagine a world where players run their own pet store – they buy new pets, look after the ones they already have, and sell food and accessories to their virtual customers. Don’t wait until the world is complete to see if the pet store is fun. Try developing the pet shop as a standalone minigame. Not a big game, just enough to prove the concept.

For the minigame, the test is not registered users, the proof is repeat visits. The golden mechanics should be fun. Players come back when they’re having fun. Therefore, if players are coming back, having found the golden mechanic minigame through the usual Google Adwords channels, we can be confident the mechanic is fun!

And that’s it. A simple testing technique for avoiding the 7 Month Gamble. There are however some downsides, chiefly that innovation comes from the edges of the bell curve, not from the centre. But that’s a subject for another post.

Other Articles