Project Governance Scope And Time To Kill Failure
Sources: 1 • Confidence: Medium • Updated: 2026-03-14 12:28
Key takeaways
- Kaplan says he treated the two-year schedule constraint as the binding realism driver and effectively ignored the World of Warcraft-like revenue requirement as unrealistic.
- Kintsugiyama is developing a video game titled The Legend of California.
- Kaplan says current AI integration in game development is mostly unreliable and overconfident, and he considers using creators’ work for AI without permission immoral.
- Kaplan says his hindsight view is that Overwatch should have emphasized individual contribution more and that the medal system encouraged blame and toxicity.
- Kaplan says he became Blizzard-connected through his EverQuest guild when the guild leader turned out to be Blizzard designer Rob Pardo, who invited him to send Half-Life levels and visit Irvine.
Sections
Project Governance Scope And Time To Kill Failure
- Kaplan says he treated the two-year schedule constraint as the binding realism driver and effectively ignored the World of Warcraft-like revenue requirement as unrealistic.
- Kaplan says the post-Titan team explored new game concepts via timeboxed ideation, committing fully to each of three pitches for two weeks with an explicit rule to discard each pitch at the end of its window.
- Kaplan says the Overwatch pitch framed development as 'crawl, walk, run' to avoid Titan’s failure mode of attempting to 'run' immediately and positioned the first game as universe-establishing rather than a full PvE experience.
- Kaplan says Titan’s core concept was a future-Earth MMO with secret-agent FPS-style missions at night and day-time simulation elements influenced by The Sims and Animal Crossing.
- Kaplan says a major technical goal for Titan was a one-server one-world architecture instead of World of Warcraft-style sharded realms.
- Kaplan attributes Titan’s failure to leadership (including himself) and says the project failed across engineering, design, and art cohesion rather than a single discipline.
New Independent Studio And Product Definition
- Kintsugiyama is developing a video game titled The Legend of California.
- The Legend of California is described as an open-world online multiplayer action game set on a mythical island version of California during an 1800s Gold Rush-inspired period and is not intended to be historically accurate.
- Jeff Kaplan left Blizzard in 2021.
- After leaving Blizzard, Jeff Kaplan built a new studio with Tim Ford called Kintsugiyama and raised funding while optimizing for creative control rather than maximizing the amount raised.
- The Legend of California’s world is handcrafted to preserve recognizable California geography while also using voxel-based generation and world seeds to relocate points of interest and alter difficulty tiers across servers.
- Kaplan says Rust inspires his current game primarily through the mechanic of a resetting world, and he is considering a roughly monthly reset cadence without Rust-style full-loss theft mechanics.
Ai Tooling Pragmatism And Constraints In Creative And Software Workflows
- Kaplan says current AI integration in game development is mostly unreliable and overconfident, and he considers using creators’ work for AI without permission immoral.
- Lex Fridman says he runs local LLMs on multiple GPU machines and simultaneously uses GPU-accelerated FFmpeg encoding.
- Lex Fridman argues that production software development will require humans in the loop for a long time and that code review remains necessary to reduce breakage risk.
- Lex Fridman describes Finn as an AI customer service agent and claims it achieves a 65% average resolution rate and is trusted by over 6,000 customer service leaders.
- Lex Fridman claims BetterHelp matches users with a licensed professional therapist in under 48 hours.
- Kaplan says AI is most useful to him for mundane automation tasks such as batch resizing and packaging thousands of images.
Franchise Management Tradeoffs Sequels Esports And Live Ops
- Kaplan says his hindsight view is that Overwatch should have emphasized individual contribution more and that the medal system encouraged blame and toxicity.
- Kaplan says work on the concept and pitching of Overwatch 2 began in 2015 before Overwatch 1 shipped in 2016.
- Kaplan says Overwatch 2 was one of his biggest creative leadership mistakes, driven by internal developer desire for PvE and mounting executive pressure tied to previously shown target dates.
- Kaplan says early Overwatch live events such as Summer Games and Halloween were extremely successful with fans, but some developers resisted because they wanted resources shifted to Overwatch 2.
- Kaplan identifies Overwatch League as a major derailment that diverted resources away from live events and Overwatch 2 development and forced the team into reactive support work.
- Kaplan claims Overwatch League was over-marketed to prospective team owners with unrealistic expectations about scale, attracting billionaire investors and increasing external stakeholder pressure.
Organization Scaling And Talent Pipeline Dynamics
- Kaplan says he became Blizzard-connected through his EverQuest guild when the guild leader turned out to be Blizzard designer Rob Pardo, who invited him to send Half-Life levels and visit Irvine.
- Kaplan says he tried to instill an assumption-of-competence framework on Blizzard’s Team 4: treat new hires as best-in-industry and listen carefully to their suggestions rather than defaulting to doubt.
- Kaplan says small game teams enable broader cross-discipline participation and reduce compartmentalization, while larger teams push individuals into specialization.
- Kaplan says as cross-discipline interaction decreases on larger teams, people can alienate and vilify other disciplines, leading to stereotyping such as blaming 'artists' as a group.
- Kaplan says Blizzard posted an associate quest designer role for World of Warcraft that he believed appeared tailored to his creative writing background, and that he went through a rigorous six-month recruiting process because Blizzard rarely hired external designers.
- Kaplan claims Blizzard shifted from being about 95% developers when he joined to roughly a 50-50 split between developers and operations by the time he left.
Unknowns
- What is Kintsugiyama’s current team size, burn rate, and the exact amount and structure of funding raised (equity, control rights, IP ownership terms)?
- What are the concrete gameplay loops, networking model, and persistence rules for The Legend of California (PvP vs PvE mix, progression retention, anti-griefing, server capacity targets)?
- Will the planned alpha/Early Access timeline occur as stated, and what quality bar will be met at each stage (content completeness, stability, performance)?
- What is the actual reset cadence decision for The Legend of California, and what player assets (if any) persist across resets?
- To what extent do Titan-to-Overwatch asset and engine reuse claims match publicly verifiable timelines and artifacts, and what reusable components were technical vs content vs tooling?