GodMode
Game Design + Statistical Analysis
Tools: Unity + C++, Photoshop + Adobe Illustrator
August 2018 - September 2018
Analytics-driven Game Design / Team Project
Overview
GAME CONCEPT
Help Terra on her pilgrimage to the top of the Tower of Cruci by building and destroying objects in her path.
God Mode is a puzzle platforming game where players must navigate each level by placing and removing platforms so that the protagonist, Tera, can reach the goal.
Be careful! You have limited resources, so you have to be sure of when and where to place or delete objects.
METHODS
Design Ideation, User Testing, Statistical analysis
Team
David Arcia, Melissa Avila, Dylan Magsarili, Jonathan Oliveira, Jerry Qu, Samuel Williams
Role
I served as a designer in this team. Some of my responsibilities included designing assets for characters and in game objects, as well as level design.
Project Context
Designing games through analytics
This was a semester-long team project for a Game Design & Analytics course. My team and I were tasked with designing, building, and releasing a video game online through popular game portals. We then utilize anonymous usage statistics from players and analyze this data to improve our game over multiple iterations. Technical aspects of this course include game mechanics, telemetry, A/B testing, and statistical analysis
The rubric of this project does not specify any genre of game but necessitates facets of gameplay that can be used to collect data from players. As a result, we came up with the concept for Godmode: a platforming puzzle game where players must solve puzzles by building and destroying objects in their paths.
1. Design Ideation
Designing games through analytics
My team and I were tasked with designing, building, and releasing a video game online through popular game portals. We utilized anonymous player statistics on these platforms to inform the design process of the game.
Brainstorming
During this stage of development we brainstormed the initial ideas for the game. Individually, group members came up with 2-3 game concept ideas. Thereafter, a group session was organized to discuss these concepts. Ultimately many of these initial concepts share similarities in aspects of genre and gameplay. Using those overlapping elements, we came up with the initial concept of the game:
Puzzle platformer game where you build the environment around you with limited resources then play what you made to reach the exit
Enemies exist and must be blocker because the player cannot attack.
Similar to games like Poly Bridge and Ultimate Chicken Horse
Cons
Hard to balance
With interesting interactions it could be a good game
Based on how building placement interacts with environment
Challenging to make a game with more than one solution
More casual game
2. Game Concept
Vision
Platforming is hard, but not when the player has a divine friend!
The player helps Terra on her pilgrimage to the top of the Tower of Cruci to meet her godly benefactor by building and destroying objects in her path. Stop time itself to place and remove platforms while Terra valiantly jumps on them.
God Mode is an experience that mixes both a physics puzzle game with a platformer in real time. The player can use the power of both god and woman to achieve their goals. Wits and dexterity will be tested as the player protects the adventurous Terra by utilizing the powers of a god.
Player Starts
Player places block to get to high platform
Player uses resources
Player advances, collects key and opens goal to next level
Objective
The primary player objective is protecting Terra as she scales the Tower of Cruci, on her journey to achieve divinity and inner peace at the top of the tower.
Player must wisely use the edit mode (“god mode”) to place and remove a limited amount of resources on Terra’s path to help her maneuver past obstacles and reach the door to the next floor.
Some floors may require one or more key(s) to exit, adding challenge and complexity as players have to be careful not to obstruct their path.
Actions
Challenges
Undestroyable blocks
One of the resource types that exist in the game are undestroyable blocks. Unlike regular blocks, in edit mode undestroyable blocks cannot be removed from Terra’s environment once they are placed (or if they existed in the environment at the beginning of the level).
Large Gaps
Large gaps are essentially death traps (which, when jumped into, lead to death) which will separate Terra from her goal in a way that cannot be solved by her jumping alone.
Falling Blocks
Blocks will fall if they have no supporting structures underneath them. The player, enemies, the exit, and the key to the exit door can all be destroyed by falling blocks. Falling blocks can be both helpful and dangerous to the player. Since falling blocks can kill enemies or create temporary paths over gaps, the player may use them to clear paths or traverse gaps. Falling blocks may be placed at the start of the level as a resource.
Unstable blocks
An unstable block is one that may fall in an unpredictable way, potentially impeding Terra’s path or falling on an exit or resource needed to exit the level, thereby destroying it. A block becomes unstable if less than 50% of it is unsupported by blocks underneath it (on a single half).
Mechanics
The main mechanic of God Mode is swapping between control of Terra and her divine friend.
Outside of edit mode, the user plays as Terra and the game plays like a basic platformer. Terra can jump and move around the play area.
Once the player uses edit mode, the game pace slows down and the game becomes more of a physics puzzler. The player can now place and remove various types of blocks to make Terra’s traversal easier.
Some of these blocks include: blocks that do not fall, blocks that cannot be removed, and blocks that stick to other terrain. The player needs to manage their time in both modes in order to succeed.
Game Mockups
Easy Levels
This is what the “playable” mode would look like:
The menu UI in this mode is very simplistic, with options for undoing a previous move that was made in the editor mode, restarting the level, and accessing the menu screen
And what the “editor mode/”god mode” would look like:
Notice in this mode the inclusion of gridlines to aid the player in resource placement and the appearance of the resource area at the bottom of the playable screen.
Medium Level
Hard Level
3. Prototype & Analysis
Friends Release
During this stage of the development process my team and I started to build a prototype of our game based on our game concept.
This was the first prototype of the game which included a rudimentary build with mechanics, basic design elements, characters and behaviors. The analytics for this game was recorded during a playtesting with sessions with peers.
Newgrounds Release
Using the feedback we received during the friends release evaluation, we made various tweaks to this prototype to improve the gameplay experience.
Some Major implementations to this prototype:
Pause menu & Win screens
Selection borders during Edit Mode, thin borders on entry, thicker borders on hover
Color-coded borders to behavior, dynamic based on state
Key destruction animation integration and physics pausing
Door opening animation integration and physics pausing
Tera ground check up by 0.15 so she isn’t hovering
Resetting resource blocks when dragged onto resource bar
Proper crumble animation triggering
Added crumbling blocks to the game
Added dissolving blocks to the game
Added animations for key, door, and crumbling blocks
New Grounds Release Prototype
Analytics
A/B Testing
We tested whether the insertion of a win screen at level completion would increase how engaged users were with out game.
We thought that the extra incentive of getting complimented on completing a level would potentially encourage the users to play more levels. Ultimately, neither version of the game did better than the other with any statistical significance.
Comparison to Friends’ Release
We can see that there is a much higher degree of retention in the early levels in the Newgrounds release than the friends release.
Notably, we retained 75% of players only up to level 4 in the friends release, but we were able to retain 75% of players until level 7 in the Newgrounds release. Furthermore, there is no substantial drop, suggesting that our improvements allowed us to introduce new mechanics to players in a better way.
Since we did not substantially change the content of these levels, we believe this is due to the increased polish we put into the game, including additional animations, fixes to bugs, better tutorialization, and an improved UI.
Burnout trends in Friends’ and Newgrounds releases
level burnout charts indicated that the largest drop-off in player retention occurred between level 11 and 12
4. Final Release
Kongregate Release
This release represents the most polished form of our game in regards to both visual and programmatic elements.
Some Major implementations to this prototype:
Analytics
Player Retention Burndown for all three releases
Overall, we believe player engagement showed improvement from the friends release to the Kongregeate release.
We notice that the is a particularly large increase between the friends release and Kongregate release in the early levels, and a increase in most later levels. There is not nearly as strong of an increase from the Newgrounds release to the Kongregate release. In the Kongregate release, levels 8 to 9 had the largest drop-off in player retention.
That aside, the rest of the curve is relatively smooth, and we were able to keep more players for the first half of our later levels, which we feel is a huge improvement. We were also able to retain slightly more players for our early levels, so we feel that the overall changes we made to the game greatly improved engagement.