System Design for Competitive Games: Final Review
A game with systems from MOBAs and First Person Shooters
What goes into a turret?
Austo-Biography
- Turrets differentiate between friends and enemies
- Totally modular - change everything with public variables
- Design decisions based on team fights and team spacing
- Don't sweat the small stuff
- Born in Los Angeles April 13, 1990
- Grew up in and Around New Orleans
- Played the Alto and Baritone Saxophone throughout high school
- I always wanted to make games, but I wanted to understand games and they why of what we do what we do
- Studied Fine Arts and Music Composition at Loyola University New Orleans
- Moved to San Francisco to study game design.
- Takes Light saber combat and stage fighting choreography classes
- Really, really, really good-looking
Thesis Proposal
Included a team from outside the school
Thesis Actual
- One Multiplayer level
- Turret Systems
- Game Mode
- 5 characters with varying stats and abilities
- Thesis Journal
- 2 Game Design Documents
- Flowcharts and process, including troubleshooting
- Deliverable List
- FPS-MOBA mashup Executable, including
- Detailed Level Map
- 8 Characters
- Items system
- Pre-production Art and High-Level Design
- Concept art
- Character Sketches
- Level Design
- System graphs
- UI layout
- Item ideas
- Process work
- Screenshots through progression
- Blog, showing the process and updated with art, screenshots, and video
- Postmortem paper, explaining the process and what was achieved and learned through the thesis.
Don't focus on the size of these lists. Focus on the design, implementation, and problem-solving
Designing a game mode
Problems that I worked through
Epic Characters in an Epic engine
- Rules of a MOBA, speed of CS:GO
- Football-like strategy
- Limiting time away from the action to level up and buy items
- One size fits all - lack of character models means they're actually all the same character, with different stats and abilities.
- Modularity is key - how easy is it to make changes as the game emerges?
- One character, ten players - making sure every one has different stats and they are replicated properly.
- Feedback is king, not content.
Reasons for my system
- Shorter rounds
- Leveling between rounds keeps players on a more even playing field
- Waiting for enough players is easier than matchmaking systems - allowed faster implementation
Problems and work-arounds
Digging deeper - How Multiplayer works
- Remote Procedure Calls (RPCs) vs Replication
- Game Mode vs. Game State
- Client - to - Server: Running up a waterfall of information
- What is important to everyone / your team / just you
- Networking, Replication, Client & Server Dialogue
- Design transcends engine - Engine changes and starting from scratch
- Design vs. Implementation
- Anything can be fun, but it needs feedback to tell the player its fun / satisfying
- Beyond my scope: Finding my pillars: Game Mode, Turrets, and Characters
- Completely scrapped the store, items, and leveling.
- Multiplayer level design between genres.
- There's always a bug
- Get character controller
- Firing a gun
- UI - one for all, all for one
- Break the player's expectations
Engine troubles
(Don't) Break the MOBA - Level Design
Process in problem-solving
- Unity 4 was lacks network infrastructure
- Bogged down by programming back-end networking and package management
- Became zoned in on cool programming tricks
- Lost track of the true purpose of thesis: to make a game.
- Get stuck
- Research Problem
- Do something else and think
- Come up with 2-3 different solutions to solve the problem
- Rank them based on fastest to implement and most likely to succeed
- Repeat until solved
Game Design Document
Thank gawd
Unreal 4
- Players of MOBAs expect certain things
- Are FPS games much different?
- Short, Medium Long game
- COD - Short-to-Medium
- De_Dust2 (CS:GO) - All 3
- MOBAs - Terrain-ignoring long game & more condensed range variation
- Blueprint for easy scripting and debugging
- Native client-server infrastructure
- Less Programming, more Designing
AND THEN....
- Two versions
- 1st - ELI5
- 2nd - Technical
MOBA Competitive Map
CS: GO Competitive Map
My Map
- All maps held to the standard of De_Dust2
- Greater overlap of ranges
- 2 general tactics that require teamwork when they fall apart/come against the opposing team
- http://steamed.kotaku.com/why-counter-strikes-best-map-is-also-a-big-problem-1695588704
- Singular, because they're all the same
- 3 lanes, 11 turrets per team
- "Jungle" between lanes is a small maze
- Hard rules on player class - lane relationship
- Little deviation on strategy
What's in a (genre) name?
Level Analysis
Lessons from jumping in the deep end
- Working alone is lonely. And it means you have way more work to do. You would be surprised all the little things that break your schedule.
- Balancing - Money, EXP, abilities, damage, health, cooldown times, speed, jump height, firing speed. Spreadsheets and graphs have become my friend.
- Blueprint and C# skills have been honed.
- Level design skills for competition and multiplayer.
- It's not pretty, but that is art's job. I made it fun.
- And it makes my resume look pretty good, too!
FPS
- Player input and camera location
- Fast-paced gameplay
- Immediate feedback to the player
MOBAs
- Persistant characters
- Characters with iconic and varying abilities
- Characters with specific roles or gameplay style
- Necessity of team structure, tactics, and player roles
Project by Austin C. Heck