Matt Charles who was a producer for Borderlands 2 at Gearbox Software, was invited by the SGDA UTD to come to give a postmortem talk about the game. The fun thing is that Matt graduated from UTD in 2008 and talks about his experience during the development of Borderlands 2 which was released in Q4 2012. Matt will cover information about the methodology and processes of shipping a game.
Table of contents
- What does a Producer do?
- Why make Borderlands 2?
- High-Level Goals
- Mission Statement
- Assemble a Team!
- Critical Review Analysis
- Initial Schedule
- Our Process and Framework
- Confluence
- JIRA
- Perforce
- Production Stages
- Design and Story
- Characters
- Level Development
- Art Development
- Creature Development
- Code development
- Coding Development
- Challenges
- Q&A
What does a Producer do?
- Be responsible for the game shipping; on time, on budget, and up to quality.
- Maintain the mission
- Establish goals
- Anticipate needs
- Remove obstacles
- Track progress
- Be camp counselor
- Ask Questions
Why make Borderlands 2?
- Borderlands 1 did well
- We were excited about the future of the Borderlands universe
- Those are ideal conditions!
High-Level Goals
- A Better Borderlands
- ~2.5 Years development cycle
- Still relevant on current consoles
- RPS (Role Playing Shooter)
Mission Statement
- Iterate if possible; innovate where necessary, but always stay true to the proven RPS formula while building a foundation for the future of the franchise.
- Demonstrate a significant, measurable improvement in both customer and critical acclaim.
- Maintain a high performance team that trusts one another in decision making an execution.
Assemble a Team!
- Don’t underestimate the power of the right people in the right positions.
- Lead structure is important.
- Play to people’s strength, not weaknesses.
Critical Review Analysis
- A massive undertaking!
- Gain insights as a team into our game and our processes.
- Absorb findings, make decisions:
- Vehicles? Improve.
- Arenas? Cut.
Initial Schedule
- Create a high-level schedule!
- Break down high-level schedule and missions into attainable, short-term goals.
Our Process and Framework
- Extremely important for a large team
- Establish a ratification process to establish clear goals and avoid wandering direction.
- The ratification process itself requires consensus-building. so problems are aired and worked out sooner.
Confluence
- It’s one thing to have an idea, it’s another to it documented and peer-reviewed.
- Crucial for time-saving.
- The usability of the software and the readability of the document are extremely important.
JIRA
- Task tracking is essential.
- Requires constant maintenance and upkeep, but is infinitely useful.
- Powerful tools help expose trends and help you understand the state of the project.
Perforce
- Source control is required.
- We nearly lost data history once during Borderlands 2 because of a server caught fire.
Production Stages
- “When did we leave pre-production and enter full production?”
- Have important questions been answered?
- Gradient-based on ‘ready-for-iteration’ status.
- Iteration loops can be traps. Keep the testable surface area low.
Design and Story
- Have a good story, but gameplay wins.
- Example: No non-interactive cut scenes.
- Design is the reason why we have the aforementioned framework in place.
- Proximity matters.
Characters
- ~1 month for design consensus on our cast.
- Rules:
- Easy to learn, challenging to master.
- No ‘wrong’ class.
- All play well solo & co-op.
- Action skill work with core FPS loop.
- Familiar and new promises.
Level Development
- Plan stages of progress for each map
- LD is where everything comes together!
Art Development
- Art Style kept the same – or not ?
- Guiding principles remained consistent.
- Art style guide for new artists.
Creature Development
- Establish the baseline: Skags 2.0
- Stalker (1st invisible enemy)
- Loader
Code development
- Construct
- A visual builder instead of a code base builder from BL1 to speed up development.
Coding Development
- Established engine that passed certification.
- Reuse the Borderlands 1 engine, but update production software.
- Debugging tools
Challenges
- Release date uncertinty
- Consensus-building can be difficult
- Planning for QA
Q&A
- Do you have any personal conflicts during the project?
- Did adding new blood help with tunnel vision?
- How do you determine if something is a good addition to the game? (Mechromancer Class)
- How do you manage the iteration during the alpha and the beta?
- What happens when a department can continue if another department fails to meet its goals?
- How do you deal with disagreements?
- How do you prevent burnout?
- More…
