A Veteran of Ultima Online and Star Wars Galaxies Says the 'Theme Park Assembly Line' MMO Has Structurally Collapsed

Picture a factory floor where every station produces the same thing: a dungeon, a raid, a story chapter, repeat until players churn. For roughly two decades, that was the dominant philosophy behind big-budget MMO development, and it worked well enough to mint franchises. Then the factory got expensive enough that the math stopped being math and started being a eulogy. According to PC Gamer's coverage of remarks from Raph Koster, the designer behind both Ultima Online and Star Wars Galaxies, the industry has now, in his words, 'hit the wall.'

What the Assembly Line Actually Was To understand

why the model broke, you have to understand what it was. The 'theme park' MMO treats players like guests on a guided ride: you follow a scripted narrative path, consume handcrafted content at a steady pace, and the studio's job is to run the conveyor belt fast enough that you never reach the end and cancel your subscription. This is fundamentally different from the 'sandbox' philosophy that shaped Ultima Online, where players generated much of the world's drama themselves. Koster helped build both models, which gives his critique unusual credibility. He is not an outsider lobbing takes; he is the person who designed the systems that the industry later industrialized and then ran into the ground. The critique, as reported by PC Gamer, is structural rather than aesthetic. The problem is not that theme park content is bad. The problem is that producing enough of it to keep a subscriber base satisfied requires a production scale that has become economically irrational as development costs spike across the industry. This is not a take born from nostalgia. It is a practitioner reading a cost curve and calling the number out loud.

The Cost Curve That Broke the Conveyor Belt Raising the

cost ceiling on games has been a documented trend for years. Koster examined the historical trajectory of game development budgets in a 2018 analysis on his own site, noting sustained upward pressure across successive hardware generations. The implications for a content-dependent live service are particularly brutal: when a single dungeon that took a small team weeks in 2004 now requires months of work from a much larger and more expensive team, the throughput math inverts. You are spending more to produce each unit of content while the player's appetite for that content has not proportionally increased. This pressure is not isolated to MMOs. Kotaku's Ethan Gach reported on leaked internal documents from Insomniac, with Aftermath subsequently covering the findings, revealing that Spider-Man 2 cost roughly $300 million to make, nearly triple the cost of the first game in the series, and landed $30 million over budget. That single data point illustrates what 'dev costs spike' actually means in practice: a sequel to a proven franchise, with an established engine and an existing player base, still nearly tripled its predecessor's budget. For a live-service MMO that has to keep producing content indefinitely, apply that cost multiplier to every content drop forever and see where you land.

Why SWG Is

the Right Case Study Star Wars Galaxies is famous in MMO history for two things: being genuinely weird and ambitious at launch, and then being gutted by the New Game Experience update in 2005, which replaced its complex simulation systems with a more conventional, theme-park-style design. As Koster documented in his postmortem series on his own site, the original SWG team that formed in Austin was a small group, refugees from Origin Systems, the studio behind Ultima Online. That small, scrappy team built a living-world simulation. The NGE replaced it with exactly the assembly line model Koster is now calling out as unviable. The irony is instructive for anyone studying game design or production strategy. The industry looked at SWG's complexity and concluded that conventional content delivery was safer and more scalable. That conclusion drove two decades of big-budget MMO development. Koster's current argument, per PC Gamer, is that the industry optimized its way into a corner: the 'safer' model turned out to require a production infrastructure that costs more to sustain than the subscriber math can justify. The sandbox he helped build was chaotic and hard to support, but it offloaded creative labor to the players. The assembly line internalized all of that labor, then watched costs compound.

What This Means

for the Next Generation of Designers Koster is not just filing a post-mortem on a dead model. He is actively building Stars Reach, a new MMO, which signals that he believes there is a workable alternative, even if the theme park conveyor belt is finished. For learners in game design, production planning, or live-service strategy, his argument is essentially a master class in how external cost curves force internal design rethinks. The lesson is not 'theme parks bad, sandboxes good.' The lesson is that any production model has a cost structure baked into its design philosophy, and when external costs rise fast enough, the model's assumptions break before anyone notices. Omdia's research into why game development has become so expensive points to the same structural forces from the supply side: team sizes, tooling complexity, and platform requirements have all scaled in ways that compress margins across the board. What Koster adds is the demand-side corollary: when your content model requires infinite throughput to retain subscribers, rising costs per unit of content are an existential problem, not a budget line item to negotiate. For anyone designing a live-service game today, that is the constraint that has to be designed around from day one, not patched in after launch. Watch what Stars Reach actually ships. If Koster's structural critique is correct, the design choices he makes to avoid the assembly line trap will be more instructive than any postmortem. That is a design education you can follow in real time.