After six years of development, Garry Newman's game creation platform s&box, described as a "spiritual successor to Garry's Mod," is set to launch next month. But before the studio prepares to release its biggest project since 2004, Newman has been reflecting on the trajectory of his original creation. The contrast is stark.
When Valve first floated the idea of selling Garry's Mod on Steam, Newman said no because he wondered: "who would pay for that?" Yet when pressed on what he thought might happen commercially, his projections were remarkably conservative. Newman has revealed that he originally expected the physics sandbox to generate roughly $30,000 as a best-case scenario.
The actual outcome bears almost no resemblance to that estimate. By December 2020, the game had generated revenues of US$119,836,074, and by November 2024, Garry's Mod had sold more than 25.5 million copies. The game has been named in the Guinness Book of Records: Gamer Edition 2025 as the best-selling PC exclusive of all time.
The gap between expectation and reality reveals something important about how digital products succeed. Newman's original mod, released in 2004, was never designed as a commercial venture. He started developing games as a hobby while working as a PHP programmer for a dating website. When Valve approached him about turning it into a commercial release, the engineer was doubtful that anyone would actually purchase what amounted to a toolkit for experimenting with another company's game engine.
What transformed Garry's Mod from a curiosity into a sustained money-maker was neither technical innovation nor corporate marketing. The game's success has been fuelled by a robust community and popular game modes, with the level of versatility in the game making it one of the best sandbox games ever, and waves of popularity on YouTube and Twitch among big creators playing an important role in keeping it relevant for 18 years. Iconic modes like Trouble in Terrorist Town and Prop Hunt were created by players, not developers. The community essentially built the game's longevity.
There is a counterargument worth taking seriously: not all community-driven products achieve this kind of success. Minecraft's approach to modding differs markedly. So does Roblox's. The fact that Garry's Mod sustained itself for two decades on Steam suggests that Newman and Facepunch made deliberate choices to maintain and support it. The game received constant updates, and the studio created infrastructure for players to contribute without losing ownership or economic opportunity.
Now Newman faces a similar but inverted challenge with s&box. When asked if s&box might match the success of Garry's Mod, Newman said "As long as it can pay its own way, that's what I want," adding that "when you enjoy what you're doing and you enjoy the product that you're making, it naturally gets that kind of snowball effect where it brings people in," as they found with Rust and Garry's Mod. His expectations remain deliberately restrained.
S&box is built using a heavily modified Source 2 engine and will be officially released in April 2026. Facepunch released s&box's source code under the MIT License in November 2025, though the underlying Source 2 code remains proprietary. This open-source approach mirrors what made Garry's Mod successful, though on a fundamentally different technical foundation.
The story of Garry's Mod teaches a pragmatic lesson about software economics. Large upfront investment does not guarantee returns. Sustainable engagement from a passionate user base often beats marketing spend. And the willingness to let users create and reshape the product often outperforms top-down design. Newman's initial pessimism about the market was not cynicism; it was a reasonable assessment based on limited data. That the outcome proved him wrong should matter less than the principle he seems to have learned: make something players enjoy, support them generously, and the commercial viability tends to follow.
Whether s&box replicates that formula remains to be seen. Newman is not banking on it. But his humility about s&box's prospects, standing in the shadow of a product that exceeded his expectations by orders of magnitude, suggests he has learned what actually drives value in interactive software. Community, time, and genuine engagement matter more than confident projections.