Following a Trail of Breadcrumbs Off a Cliff

Our process has been to release our game very early, and to constantly work to improve it.

When we launched the game in April, there was no tutorial, the game was really confusing, and there were a lot of things that needed massive improvement. But we thought that the game was good enough that people who liked card combat games would be able to work through the confusion & find the game lurking within.

Part of the reason to do that is that with game development, it's very hard to know how players will react to the game when it goes live. Sure, you can anticipate some of these things with rigorous playtesting, but at the same time, you can get a lot farther a lot faster the more people you have playing the game. And so, we release it, and we use players' experiences as feedback, and we work to improve the game from there.

Our initial goal with the release of the game was to answer this question:

Are the core mechanics of the game good?

That is, if someone could figure out the game, would they play for a while? Would it get boring quickly? Would people find exploits that broke the game?

The result? We've got a significant number of people who've played every day since they started, many times a day, and they're not slowing down.

But there was a second question.. And that is this:

Can this be a standalone game?

And the problem with that is that the answer was "not yet". New players had massive problems understanding the game. So we spent two months working on a UI revamp and a new tutorial. We added new cards, new animations, tweaked the graphics, revised the front end, and made a ton of changes focused on getting new players "in the door".

The result? Not much. We made some improvements, but not enough to consider this a standalone game. Which was a massive disappointment, and in ways, really confusing, because the game was so much better than it was before. But there was one big problem.

We didn't ask that second question.

We asked the first question, and instead of looking at what players were doing to answer the first question, we started looking at the data to answer the second question instead. And at every juncture, that made some sense - we saw that players were confused. So we made changes to address that, and the data showed us that players were a lot less confused. But did it help new players "stick"? No. Why not? Because (something else). So we'd fix (something else), and we'd see that that wasn't a problem any more. But did it help new players "stick"? No.

And it sounds kind of crazy, but the thing that we didn't realize was that we didn't ask the second question. We knew that this wasn't a "complete" game. We released this as a test for a very specific thing, and when we got the answers we needed, we started chasing the second question.

In some ways, it was the right thing to do. We found a big problem, and fixed it. We found another big problem, and fixed it. At every stage, we looked at what the biggest obstacle was, knocked it down, and validated that we'd indeed made improvements.

But what we hadn't realized was simply that because we were trying to validate whether the core mechanics were good or not, we hadn't yet made the rest of the game. And there's no reason to think that it would behave like a complete game, because it wasn't complete.

Sounds very straightforward, but was a surprisingly difficult realization to come to.

So there are two things:

  1. Make sure when you release something early to answer a question that you really clearly know the question you're asking, and make a plan for what to do when that question is answered. Know when to move to the next question.
  2. We're working on the things that are missing. It's changed our focus a bit, and the kinds of things you're going to see in the game over the next few months. It's exciting, because this is much more player-oriented stuff. Things like character progression. But it also means we're going to be taking our first swings at the very core idea behind the game.

We built the card combat in Give Me Fuel because we needed it to do something really specific. It had to be a "good foundation" to build on. The mechanics had to support long-term evolution, a huge amount of potential content, and a variety of play styles. Over the last few months, we've learned that it does all that. Now we have a different challenge. Now we get to build the house.