What this covers
Jeff Kaplan, legendary designer of World of Warcraft and Overwatch, sits with Lex Fridman to trace how great games actually get made. The conversation moves between concrete production war stories and philosophical reflections on creative work—why small teams of self-described "weirdos" consistently outpace bloated expert-laden studios, how business incentives corrode craft, and what separates vision from mere ideas. Kaplan draws on three decades of shipping titles, from early Blizzard days through the post-Overwatch reckoning, and refuses the usual industry mythology about passion and hard work, instead laying bare the structural conditions—team size, scope discipline, founder-protection of the studio—that actually determine whether a game ships or becomes another expensive graveyard.
The substance splinters into several territories. On game design itself, Kaplan unpacks Three Types of Fun (for player, designer, and system), Elements of Fun as discrete levers of intrinsic motivation and extrinsic motivation, and why matchmaking exposes raw human psychology—people claim to want fair competition but neurologically latch onto losses. On production, he argues that Minimal Incubation beats anticipatory hiring, that Small Team Advantage is structural fact not sentiment, and that scope discipline belongs with creative directors, not producers. On strategy, he contrasts Titan's collapse with Overwatch's survival through ruthless refocusing and the Crawl Walk Run philosophy. Kaplan also pivots to culture—why Mike Morhaime, Allen Adham, and Frank Pearce built something legendary, why online communities feel hateful yet arrive at conventions radiating love, and why he ultimately left Blizzard when a CFO weaponized revenue targets against game design. The conversation includes reflection on failure, walking away from doomed projects, and the difference between what you want to do and what you want to be.
Kaplan argues that great games come not from chasing money or expertise but from small teams of passionate 'weirdos' who fall in love with a clear vision, scope ruthlessly, treat collaborators as the best in the industry, and protect creative control from corporate incentives.
- WoW succeeded partly because the team 'didn't know what they were doing,' while the expert-laden Titan failed
- Overwatch was salvaged from Titan's wreckage by ruthless scoping and a six-week pitch, contrasting Titan's 'say yes to everything' culture
- Kaplan left Blizzard when a CFO tied his game to revenue threats, illustrating how business incentives corrode creative craft
Small teams with shared creative ownership produce better games than compartmentalized large organizations.
- Small teams produce better games because everyone is in the room for every decision and has a loud voice; as teams grow, members become compartmentalized cogs, and reduced interaction breeds alienation and vilification of other disciplines ('the artists just don't get it'), which is both false and demeaning.
“The bigger the team, the more you become a cog in the machine. And on a small team... If we're a small team... there's only 10 of us, all 10 of us are in the room for every decision.”
- Small studios are the future of gaming because the compelling, new, innovative ideas come out of them; large studios primarily acquire small studios for new IP and ideas, then absorb them.
“Small studios are the future of gaming. The big studios basically acquire the small studios for new IP and ideas... The really compelling, new, innovative ideas are gonna come out of small studios.”
- Game teams uniquely combine vastly different personality types on a spectrum—emotional, illustrating artists on one end and logical, organized programmers on the other, with sound designers, game designers, and producers in the middle—and the brilliance is aiming these very different talents at a shared vision of the game.
“That's actually the creative tension behind games and what makes game teams so unique, is if we were to line them up on some crazy spectrum on one end, you're gonna have the artists who... they are artistes on that end. And on the other end, you have the most logical, brilliant programmers”
- Hiring people ahead of need ('anticipatory hiring')—e.g., bringing on 70 environmental artists before the world, story, art style, or technical constraints are defined—creates idle hands, busywork, and demoralization; the correct approach is to incubate with the smallest possible team using whatever technology and prove the idea before expanding.
“in one year, we would hire, like, 70 environmental artists from all over the world... we hired way too many people.”
Developers must retain creative control and refuse executives' money-driven demands to protect game quality.
- Game developers are the 'golden goose' and repeatedly hand their value to cutthroat businesspeople; they should retain creative control, refuse to let money-chasing executives dictate terms, and recognize that you don't need a CFO to make a great game—you need artists, engineers, designers, producers, and an audio team.
“World of Warcraft, when we made it, there was no CFO at Blizzard. You don't need a CFO to make World of Warcraft. You need artists, engineers, designers, producers, and an audio team.”
- Blizzard became legendary because it was founded and run by three gamer-programmers (Mike Morhaime, Allen Adham, Frank Pearce) who had made games before running the company, understood what developers went through, shielded them from business nonsense, maintained a high quality bar, and kept the company ~95% developers in early days.
“It was run by three gamers. They were, all three of them, programmers. They made the games before they just ran the company, so they knew what each of us as developers beneath them were going through, and they protected us. They shielded us from all of the nonsense”
- Scope should be controlled by the tech director, art director, and game director—not handed to producers or executives as a production problem—because creative directors know which parts require big investment and which can be cheated (e.g., spending hours not weeks on each of 200 crates).
“The mistake I see on a lot of game teams is scope becomes a production problem... The scoping, your best-case scenario is when your tech director, art director, and game director are doing the scoping.”
Focused, iterative strategy—starting with core proof-of-concept—outperforms attempting everything at once.
- Titan failed because it tried to 'run'—to be the next World of Warcraft immediately—when WoW itself was built on Warcraft I, II, and III establishing a world people cared about; Overwatch deliberately adopted a 'crawl, walk, run' strategy, shipping a focused PvP shooter first just to establish that the universe was worth caring about.
“We had identified the reason Titan failed is we just tried to run; we tried to come up with the next World of Warcraft. But if you think about World of Warcraft, it had Warcraft I, II and III to build upon”
- Overwatch's design emerged from designer Geoff Goodman's wish to make 50 classes each with one or two interesting things rather than six classes with 100 abilities; Kaplan combined this with existing Titan hero art (the Jumper becoming Tracer) to create a hero shooter that plays to the team's strengths and could realistically ship in two years.
“I wish instead of making, like, six classes, I wish we could make 50 classes. And I wish instead of having, like, you know, 100 abilities on the classes, the 50 classes all just had, like, one or two things that was really interesting about them.”
- Ideas are infinite and easy on creative teams; what matters is vision—the ability to shepherd an idea into existence through inspiration plus a technological plan, design plan, art style, and pragmatic production reality. Titan failed because it had ideas without vision and lacked a path to reality.
“ideas are easy. Ideas, you know, I can have 10 in 10 seconds... What I call vision is the ability to not only take a great idea, but shepherd it into existence, and you're doing that through inspiration first and foremost.”
Pursuing perfectionism is a mistake; imperfection and rough early access reflect the true nature of creation.
- When you conceive a game and pitch it, that is the best the game will ever be; every day toward shipping, the reality diverges further from the imagined ideal as you cut and say no, so developers ship a 'greatly sacrificed thing' and become disenfranchised with the concept—living in a constant state of disappointment.
“When you come up with the idea and you start pitching it to people, that's the best the game is ever gonna be, and then you work on it... Every day you get close to ship, the imagination of the ideal game gets farther and farther from the reality”
- Drawing from the Japanese craft of Kintsugi (repairing broken pottery with golden lacquer to make scars beautiful), the philosophy is that nothing is ever perfect and the pursuit of perfection is a mistake; making games is a constant pursuit of imperfection, and releasing rough early-access versions lets players feel part of the creation as it evolves.
“there's also a philosophy in Kintsugi that nothing's ever perfect, and the pursuit of perfection is actually a mistake, and that there's beauty in imperfection.”
- Despite cultural conditioning to never give up, sometimes walking away from a dream is necessary and healthy; persisting against overwhelming evidence (170 rejections in a year) can destroy a person, and cutting off the past completely (throwing away all manuscripts) can be required for another door to open.
“A group of editors at literary magazines across the United States was telling me it was time to give up as a writer, like I wasn't cut out for it. And I stopped.”
- Young people should focus on what they want to do, not what they want to be; society pressures people toward role-fantasies (astronaut, writer, firefighter), but identifying what genuinely brings you fulfillment moment-to-moment when no one is watching reveals what you will actually become.
“focus on what you wanna do, not what you wanna be. The pressure that society kind of puts on us is, you know, "Oh, do you wanna be an astronaut? Do you wanna be a firefighter?”
- Good matchmaking is one of the hardest design/engineering problems because players claim they want fair, even matches but actually want to feel slightly better than the opponent and win; since it's zero-sum there must be winners and losers, and human psychology fixates on losing streaks while ignoring prior winning streaks.
“the reality of what they want is they want a match where they're slightly better than the other guy. Like, they want it to feel like it was close but then win. And you can't architect that.”
- Because top studios can recruit the best talent in the world, when a new hire (e.g., the best prop artist in the industry) makes a suggestion, leaders should assume competence and listen rather than starting from doubt and defending their own ideas; this reframing turns ego-driven dismissal into productive collaboration.
“That is the best prop artist in the industry. That's who's gonna show up on our doorstep, so when they show up here, we should treat them like the best prop artist in the industry instead of starting from a place of doubt and cynicism.”
- A periodically resetting world (as in Rust, and planned for Legend of California) is a powerful mechanic because it lets the high highs and low lows coexist, prevents over-attachment, and equalizes players so newcomers and veterans can have a meaningful shared experience starting 'naked on the beach' together—unlike WoW where a level-80 and level-1 cannot meaningfully play together.
“if we wanna play World of Warcraft, and I'm level 80 and you're level one, there's no meaningful experience we can have together—but in Rust, we just wait for a reset and we're both naked on the beach”
- A hindsight regret: Overwatch put all its eggs in team win/loss and downplayed individual contribution (no scoreboard, a flawed medal system that both teams earned and players weaponized), but players are fundamentally selfish in how they derive joy, so a hero shooter built today should emphasize individual contribution more.
“if I was making a hero shooter from scratch today, I would make it less team-focused. And we put all of our eggs in you noticing if the team won or lost.”
- There are three types of fun in game design: fun for the player, fun for the designer, and fun for the computer; good multiplayer designers must ensure the experience is fun for the player rather than indulging the designer's own amusement (the 'ant farm' temptation).
“there's a really great famous Sid Meier quote where he says there's three types of fun. Fun for the player, fun for the designer, and fun for the computer.”
- Integrating current AI into game development is mostly a hot mess because AI is overconfident in what it delivers and only right about one in ten times (e.g., giving confidently wrong answers to UI coding questions); AI is useful for tedious mundane tasks but cannot replicate the irreplaceable human spirit of great artists and storytellers, and using creative work without permission is immoral theft.
“The problem with AI right now is it's overconfident in what it tries to deliver... it, like, overconfidently gives me the wrong answer. And it's, like, right one in 10 times. So its hit rate has to be a lot better.”
- Crunch (extreme overtime) is wrong and immoral when it is forced, unnecessary, or mandated by executives who themselves work 40-hour weeks; but voluntary intense work driven by personal passion for one's life's work can be fulfilling and formative, and the speaker never blamed his self-chosen hours on anyone but himself.
“where crunch is problematic and people are good to be vocal about being opposed to it, is when it's forced and unnecessary... I don't think executives who take off and work 40-hour weeks should be telling anybody to stay late.”
- Text-based games like Zork derive their lasting power from imagination—the world exists in your head—which is why the later graphical 'return to Zork' shattered the experience, the same reason a book is often better than its movie adaptation.
“they made a return to Zork, but it now had graphics. And somehow, that just shattered everything, because the Zork you knew in your head didn't exist anymore.”
Multiplayer design emerges from laying groundwork and watching players create unpredictable emergent experiences.
- Designing a multiplayer game is like being a movie director who sets up all the actors, sets, props, and scripts ahead of time, then must leave the room and let the actors create their own experience—you lay out the chessboard and watch what players create, which is the unique joy and challenge of multiplayer design.
“imagine if you were gonna be a movie director... your goal was to get a certain movie made. But we're gonna ask you, the director, to just... You're gonna leave the room. You can set it all up ahead of time, and then you're not allowed to be there”
- The most magical stories from video games are emergent player stories (e.g., giving a potion to a stranger who turns out to be Blizzard's founder, or accidentally shooting an NPC in Red Dead and triggering chaos), which are more interesting than most authored game writing—though exceptional narrative games like Red Dead Redemption 2 combine both.
“the most magical stories that I've ever heard come out of video games are player stories about, you know, the time I gave Barfa a potion and then I met him in real life. Like, that's better than any video game writing”
- Most games fall off the back of the wave—no one plays them or they last two weeks—so if you are lucky enough to catch the wave (a hit), you should ride it to the end with continuous live events rather than diverting resources to a sequel; abandoning live content for Overwatch 2 was a mistake.
“There's this phrase of catch the wave, ride the wave. Most games fall off the back of the wave. They don't catch the wave... If you're lucky enough to have caught the wave... ride it till the end.”
Polish, responsiveness, and foundational technical breakthroughs matter more than scope or graphic ambition.
- Blizzard's renowned polish stems from a studio-wide culture where no one tolerates bugs, a best-in-industry QA department of passionate gamers tightly integrated with developers, and master-level engineering that architects client-server systems to be hotfixable—able to deploy server patches (e.g., disabling an exploited hero within 30 minutes) without players' clients going down.
“if you really wanna run a game like Overwatch or World of Warcraft successfully, you need master level engineers who have architected the client and server in such a way that you can hotfix the game on a dime. And what a hotfix is, is a server patch that no one's client has to go down for.”
- Audio and lighting are the two elements of a game that no one realizes how much they contribute until they are missing; great composers, sound designers, and lighting artists tap into senses that otherwise go unused, and a game without them feels subtly off and wrong.
“there are two things that no one realizes how much they bring to a game until they're missing, and that's audio and lighting. Because most of the time, we're playing without these things, and it just feels a little off and wrong.”
- John Carmack and id Software achieved foundational breakthroughs in gaming engines (Wolfenstein 3D, Doom, Quake) where the magic was not just graphics but the smoothness, responsiveness, and sharpness of gameplay; the intimacy of the first-person camera transported players in a way that pushed the entire FPS genre forward.
“the level of breakthrough that they achieved with Wolf 3D... it wasn't just how the game looked, it was how it played. The smoothness kept getting better. The responsiveness, the sharpness of the gameplay.”
Game design taps discrete motivational elements players move between fluidly across intrinsic and extrinsic drivers.
- Game designers tap into discrete 'elements of fun'—progression (investment recognition), mastery, creativity, customization—and motivate players via both intrinsic (content/experience) and extrinsic (loot) drivers; players fluidly drift between these motivations, e.g., becoming temporarily loot-motivated to overcome a boss in order to return to being content-motivated.
“The two driving motivators are extrinsic and intrinsic. And all of us, at different times in our lives... we can shift from being intrinsically motivated to being extrinsically motivated.”
- WoW revolutionized MMOs by overloading experience-point rewards into quests, making the path of least resistance to leveling a story-driven journey through the world rather than grinding monsters in one spot; this created 'directed gameplay' that felt optional but really wasn't, enabling solo play within a massively multiplayer world.
“we said, why don't we make the path of least resistance, so in this case, the way to get the best experience the fastest not to be killing creatures in one place, but will overload the experience into the quests themselves, and then that will move you through the world”
Social media toxicity and online negativity destroy creator morale and drive talented people from the industry.
- Social media is designed so maximum hyperbole—usually negative—wins attention, creating a self-reinforcing culture where people only speak up to hate; this fire-hose of criticism is a new dangerous skill requirement for creators and has driven talented people (e.g., Diablo III's Jay Wilson) out of making games for years, robbing the world of their creations.
“social media in general is designed in such a way where the maximum hyperbole works, and that's how you get the most points is by being max hyperbolic. And usually, unfortunately, it's more in the negative direction than the positive direction.”
- The same communities that appear hateful online (forums full of flames about lag and bugs) are, in person, an outpouring of love and shared passion; the BlizzCon experience revealed that online negativity is a distorted signal because passionate people express themselves harshly online to get attention.
“And when I showed up at that show, it... One of the most emotional things in my life. It was nothing but an outpouring of love.”