Candy Adventure

Play video game together with your family and friends

VIDEO GAME DESIGN | HIHEX | 5/2014-6/2014
with Heng Song


In March 2014, I joined a startup in Shenzhen, the hardware innovation center in China. The startup aimed at renovating the overall user experience of smart TV OS. At that phase, the company was working on a mobile remote control app named HexLink(video demo), and we wanted to design TV applications that use this remote control.

In a small agile team of 2, we create this video game called Candy Adventure, which inherited the mechanics of the classic game Puzzle Bobble , and incorporated brand new features to adapt to the Chinese video game market. Applying an agile development method, we created this game through several quick sprints and released it after 2 months. The number of active users reached 1,000 within 1 week of release, and it has over 100,000 ​downloads as of June 2015.

A short demo of the game

Problems & Opportunities

We wanted to design a game that uses our remote control app, HexLink. HexLink well solved the problems of the current widely-used smart TV control methods.

TV Remote Joystick HexLink
Continuous Control NO YES YES
Directions 4 4 All directions
Number of Players 1 2 Unlimited
Extra Cost YES YES NO

First, HexLink offers continuous control, which means users don't click on one button to move one step towards one direction, but rather control the movement continuously by moving fingers in the app. Joystick partially solved this problem, but it still has only four directions to go and causes extra cost.

Second, HexLink brings the opportunity for multi-player entertainment activities on TV. TV originally is a shared entertainment tool which is watched by family members sitting together in a living room.

This is how we came up with our idea. What if there is a collaborative video game that multiple family members can use our remote app to play together intuitively without having to pay for extra devices? After doing competitive analysis, we found with delight that there was no one single game that does this in China's android TV app store. So we quickly started our plan.


As a startup team, we had to move fast to be the first one in the market, so we decided to build on some classic game mechanics. After brainstorming, we came up with two rough initial ideas, formed two agile group of two, and got started!

As the designer for the second group, I was responsible for the whole product design, including features, mechanics, interaction, visual, sound, animation, and everything. I collaborated closely with the other developer. I was also part of the user testing, marketing, and operation.


In the first sprint, we quickly created an MVP that only support the key mechanism of the game, and meanwhile incorporated the sharing nature and intuitive control methods. One design challenge is how to design the layout to support multiple players. I tried different versions, including users in the center shooting outside, users staying in parallel at the bottom shooting up, etc. I picked the one with four players at each four corners because more interaction could happen with a shared workspace and bouncing walls.

We mocked up two distinct visual versions in order to test users' general expectation and feedback on such game. We tested the MVP at an industry conference with some experts in this area as well as some random users.

The participants enjoyed our game and easily got addicted to it. Some interesting dynamics emerged which informed our further iterations:


In our second sprint, we not only enhanced the two major strength of our game - continuous control method and collaboration, but also put effort in creating a unique aesthetic identity for this game:

Game Mechanics

From the user research, we found that players often applied the strategy to kill others first and then focus on clearing the bubbles. And the people who are killed complained about feeling boring waiting for other people to finish the game.

So we added the mechanics of saving other people and going back to life. If one clears the bubbles that kill one player, the player would go back to life.

Mechanics of die and reborn

We also included the feature to share the result on Wechat (China's most popular IM app), which gives people the opportunity to show their achievement to friends. Click here for the responsive sharing page I designed and implemented for this feature.

2 versions of mobile sharing page

User Experience

The control method is what we devoted most time to. We adjusted the ratio of the vertical and horizontal movement when we translated it from mobile remote to the TV screen because people tend to move horizontally more. We also mapped the TV location to the phone screen because people at the top two spots were confused by the flipped control direction.We created an interactive tutorial for this game as well.

Interactive Tutorial

Workflow revision

Story Building

To make our game more unique, I thought of creating a coherent story for it that could guide the development of the game as well as set a tone for the overall style of the game. After several discussions, we decided on the story of a group of candy monsters being caught by candies when taking a trip to the candy planet, and the other candy monsters are starting the trip to rescue them by fighting against the candies.

Character & Scene Design

I played around with different visual styles, and found this light, shining, clean and candy-like visual style work best on the TV screen and in creating the aesthetics that best adapt to our users. I also started to learn animation building from novice, and by adjusting frames one by one, I created all the animation for our game.

Visual Iterations

Final Visual Design & Elements

Animation Design

After 1 month of intensive teamwork, our game was finally released on June 2014. Here's a desktop demo video of the game (I don’t have the access to its TV version right now because I am in US. But I would be happy to talk more about it):


When reading the data of this game, we found that even though most users launch our game on a daily basis, they won't stay long(average playing time was 10 min). From some in-person user testing, we got the feedback that the game could be boring after a certain period of time because it's all repetitive work. So we thought about different ways to make it more fun.

After discussing with our developer, we decided to start the second one first, because the IRO is higher than the first one from an implementation perspective.

Sprint 3 Scene Design & Interaction Specs