Why it took a year to make, and then break down, an amazing puzzle game

Threes is a game that is best described as being tiny.

You slide numbered tiles around a four by four grid, and you have to match multiples of threes together to increase your score. So if you have two threes next to each other, they can merge into a six. The sixes merge into a 12. And so on.

Maneuvering around the board and creating higher numbers while making sure your 24s and higher are close enough to merge becomes fiendishly difficult in a short amount of time. Your score isn’t based on the total number on the board, but on the number of advanced tiles. A "192" tile is worth much more than a "96" tile, for instance. The following animation will help explain the game.

Threes is tactile, satisfying and the $1.99 iOS release proves hard to put down once you get into the rhythm of the game. Getting to this point, where the game is easy to understand and play while keeping that simple hook, didn't take a year. Trying every possible way to make the game more complicated before throwing all the ideas away, on the other hand? That took time.

Where the time goes

"You enter a different mental state for every number. You have to be much better to get a 192 than a 96, and then much, much better to a 384, and so on," Asher Vollmer, the designer of Threes, told Polygon. The game was in development for a year, with Vollmer working with Greg Wohlwend, a developer who also had his hand in Ridiculous Fishing, Hundreds, Gasketball and Solipskier.

I’ve been hooked on the game since they sent me a pre-release code around a week ago but, once again, the game mechanics and visuals are simple, almost classical. How the hell did this game take a year to create?

"This is Asher’s brainchild, just like Puzzlejuice. He is the game designer. He came up with all these different ideas. In the beginning it sort of looked how it currently looks in terms of simplicity," Wohlwend explained. But of course a game that simple needs… well, more?

"We added like, a monster…" he said, trailing off.

This is what took so much time. The game wasn’t designed with an eye towards clarity and simplicity, and I’ve included examples of some of the art mock-ups in this piece so you can get a sense of the dead ends the two men hit in development.

Grid View

They kept trying to add new concepts, new themes and, in some cases, new mechanics. These ideas might have been interesting, but they hid what made the game fun; the simple act of matching numbers in multiples of threes.

"Every time I added anything like that, it always felt unwieldy and unnatural. It just didn’t want to be there," Vollmer explained. "Every time I took that stuff out it felt like a good game again."

Wohlwend sent Vollmer concepts for the game, based on number of ideas. There was a version where the tiles were sushi, and you got ahead by matching fish and rice. They tried it with animals. There was a chess-based version where you mixed two knights to create a king. They would often show these versions of the game to friends, who suddenly couldn’t understand what was going on.

"We were hiding the game under all these layers," Vollmer said. At one point veteran game designer Zach Gage sent them a "sobering" e-mail that said the art wasn’t working, and this acted as a wake-up call. What if the game didn't need anything else?

It took time to try every possible way to make the game more complicated before throwing all the ideas away

"It looks like the entire time we were striving for simplicity and minimalism, and that’s not true at all," Wohlwend said. "That’s just where we ended up going. The game resisted complexity because it was such a small game, it was four by four grid, and numbers, and just the four directions. It always wanted to be simple."

That year-long journey of adding mechanics and art for testing, only to strip them away and get back to the core game again and again wasn't wasted; the amount of testing and failed experiments gave them a sense of purpose and clarity when it came to the final game. It needed to be small, simple and easy to understand. The visual information needed to be clear and readable. Every other concept or theme was thrown out in deference to the original idea of matching tiles in multiples of three.

"I don’t know how I would have been confident enough to make this game as it is with just a week of working on it," Vollmer said.

Which isn’t to say the tiles don’t have some personality. You can see their eyes and faces, and they react when you merge them together. Those faces are given more curiosity as you reach higher and higher numbers on the tiles.

"There are faces that no one has ever seen in the game other than Asher and I that we can’t talk about," Wohlwend explained. "But they’re very cool, and I’m excited about people seeing them. They get more character as you ‘level up.’"

I've yet to be able to get a tile up to 384, but I've seen him. He has a single sharp fang and a pirate's eye patch. His name is "Capt. Triad." Each of the tiles has this sort of simple, easy to read personality, and reaching a new high score feels like a major achievement.

Creating a game this minimal is hard, and without trying all these other concepts it's unlikely Threes would feel this satisfying to play; the two men knew exactly what they wanted after trying so many ideas to add concepts to the game. It turns out the best thing was the original thing: the act of matching threes. They just had to gain the confidence to stop fighting their own instincts.

Instead of adding things, they had to remove everything that wasn't the game, and what's left is a brilliant little puzzler. It took them a year to learn to listen to what Threes was telling them.

"The game always knew what it wanted to be," Vollmer said.