What Mobstar.cc Still Knows About Making Games

When my brother and I were younger, we used to play a browser game called Mobstar.

There wasn't much to look at. Mobstar was a text-heavy online mafia game. You committed crimes, bought and sold drugs, accumulated money, improved your character, joined other players and tried to climb toward the upper reaches of a fictional criminal underworld. Most of what happened was represented through text, numbers, menus and buttons.

It belonged to a category of games that seemed to be everywhere on the Internet in the early 2000s.

And then, mostly, they disappeared.

So recently I did what anyone possessed by a sufficiently useless piece of childhood curiosity might do. I typed:

mobstar.cc

into my browser.

I expected nothing.

A dead domain, perhaps. A holding page. Maybe an Internet Archive result documenting the remains of a small website that had vanished fifteen years ago.

Instead, Mobstar loaded.

Not a preservation project. Not an emulator.

The game itself.

Even stranger, it is apparently still being maintained. Mobstar's website carries a 2003–2026 copyright notice, and on December 5, 2025, its administrators announced that the game had been reset and was beginning again with new updates.

More than twenty years after my brother and I encountered it, somebody is still keeping the lights on.

And seeing it again made me wonder whether these old games represented something more interesting than an obsolete technology. Maybe there was a design philosophy buried in them that is worth recovering.

The game that was also a website

There is a name for the broader family Mobstar belonged to: the persistent browser-based game, or PBBG.

The definition is almost comically simple. A PBBG is played through a web browser, and its world or your progress persists across multiple sessions.

That simplicity was part of what made these games remarkable. You didn't install anything or need a gaming computer. You didn't launch Steam, download a 90-gigabyte update or reserve your evening.

You visited a URL.

That made browser games unusually easy to slip into the cracks of ordinary life. You could log in, spend a few minutes doing whatever actions had become available, check your money, talk to someone, make a trade, discover that another player had attacked you, and close the window.

Then you could come back later. The world would still be there.

This low barrier to entry was one of the defining advantages of browser gaming. A 2023 history of the medium, Caught in the Game: On the History and Evolution of Web Browser Gaming, by Naif Mehanna and Walter Rudametkin, traces browser games from early text-based experiments through modern hardware-accelerated games and emphasizes how simply visiting an address allowed developers to distribute persistent games across different kinds of computers.

For my brother and me, though, none of that was a design principle.

It was just fun.

Mobstar could occupy five minutes or an hour. There was always another number to increase, another risk to take, another player doing something irritating. And because so much of the game existed as abstraction, our imaginations quietly did part of the rendering.

When graphics were expensive

There was an obvious reason games like Mobstar relied so heavily on text and menus.

They had to.

A small web developer in 2003 could not build Grand Theft Auto in a browser. But that limitation created an interesting economic property.

Systems were cheap.

Suppose a developer wants to add a nightclub to a modern high-budget 3D game. The nightclub might require architecture, textures, lighting, furniture, music, character models, crowd animation, sound design, collision geometry, voice acting and weeks of testing.

Thousands of hours of work might be devoted to making the nightclub believable before the player does anything interesting inside it.

Now imagine the same nightclub in a browser-based game.

It could initially be this:

The Blue Comet

Capacity: 430
Reputation: 72
Security: Moderate
Owner: Rob1987
Entry: $25

That is not particularly impressive.

Until you discover that players can buy the nightclub, hire security, book performers, set drink prices, sell it, borrow money against it, use it to conceal other business, compete with the club across town, organize events, form agreements with other players, or bankrupt themselves trying to expand it.

The nightclub that looks less real can, paradoxically, become more real as a system.

Its visual representation is thin. Its web of possibilities is thick.

The complexity budget

Every game has what we might call a complexity budget.

Developers can spend that budget in different places. One place is representation: make the mountains better, make skin behave more realistically, add ray-traced reflections to puddles, animate individual strands of hair, record another fifty thousand lines of dialogue.

None of these things are frivolous. Modern game worlds can be astonishing works of art.

But complexity can also be spent somewhere else: on economics, politics, ecology, crafting, social systems, player organizations, trade, construction, reputation, law, ownership, or strange interactions nobody anticipated when the game was designed.

That second category creates something especially valuable:

emergence.

Instead of writing every story, the developer creates systems capable of producing stories.

A crop shortage changes prices. Higher prices attract merchants. Merchants become wealthy, wealth creates political influence, and political influence produces resentment. Two player organizations begin fighting. A third organization sells weapons to both sides.

Nobody on the development team ever wrote a quest called The Great Cabbage War.

But five years later, players still remember the Great Cabbage War.

That kind of story belongs partly to the designers and partly to the people inhabiting the world.

We have become extremely good at making worlds look real

This may be one reason some enormous modern games can occasionally produce a strange sensation.

You enter an exquisitely rendered building. The floor reflects the light correctly. Dust floats in the air. Every chair has scratches on it. Someone spent three weeks modeling the fire extinguisher.

And you cannot sit down.

There is nothing inherently wrong with this. A game is not obligated to simulate everything. But it reveals an interesting imbalance.

Over several decades, games have made extraordinary progress in the fidelity with which they represent worlds.

Perhaps another frontier is the fidelity with which players can affect them.

Consider a tree. Modern technology can render a magnificent tree. Its leaves respond to wind, its bark contains photogrammetric detail, sunlight passes through its canopy, and birds land on its branches.

Now ask another set of questions.

Can I cut it down? Can I plant another one? Can I own the land underneath it? Can I sell its lumber? Can I start a lumber company? Can other players protest my lumber company? Can the town pass a law protecting the forest? Can I run for mayor and repeal the law? Can someone bribe me not to? Can somebody establish a religion that considers this particular tree sacred?

The graphics describe what the tree is.

Systems describe what the tree means.

And perhaps games have considerably more unexplored territory in the second category.

The browser game didn't entirely disappear

It would be wrong to claim that this philosophy vanished.

One of the clearest surviving examples is Torn, another crime-themed text RPG that began in 2004 and remains active more than two decades later. Torn still presents itself as a text-based online RPG built around crime, commerce and interactions among players.

There is also still a small community organized around PBBGs themselves. Developers continue making persistent browser games, discussing them and experimenting with the format.

But culturally, the center of gravity moved elsewhere.

Many ideas that once made browser games distinctive migrated into social games and mobile games: persistent progression, energy systems, daily actions, virtual currencies, clans, leaderboards, and timers that encourage you to leave and return later.

The strange little browser RPG was not simply exterminated.

Parts of its DNA escaped into the larger gaming ecosystem.

But something may have been lost

The interesting thing to revive is probably not the exact presentation.

We do not need to cover websites in tiny Times New Roman text and beige tables again. Modern browsers are incomparably more capable than the browsers Mobstar was designed for. They can support sophisticated graphics, networking and hardware-accelerated 3D experiences that would have seemed absurd in 2003.

So a revival of this philosophy does not require pretending the last twenty years never happened.

A systems-first game could use beautiful illustrations. It could use pixel art, animated maps or sound. It could even have selective 3D environments.

The important constraint would be philosophical rather than technological:

Spend less of the development budget telling players what the world looks like, and more of it deciding what the world allows them to do.

That is especially interesting for independent developers.

A small team cannot outspend Rockstar or Nintendo or Electronic Arts.

But it might be able to out-weird them.

It can build an economy nobody would approve in a fifty-million-dollar production meeting. It can create a political system whose consequences are genuinely unpredictable. It can let players manufacture objects the designers never expected them to value. It can give people ownership over parts of the world. It can create systems complex enough that players begin inventing uses for them.

The limitation becomes an advantage.

Instead of competing on spectacle, compete on possibility.

The strange freedom of abstraction

There is another advantage to all this simplicity.

Text leaves room for imagination.

If a game tells you:

You have reached the abandoned observatory.

your brain builds one.

The developer doesn't need to specify every chair. And because the imagined observatory costs almost nothing to render, the developer can afford to create ten thousand places like it.

This is one of the peculiar freedoms of abstraction.

A game does not necessarily become larger by drawing more of its world. Sometimes it becomes larger by refusing to draw it.

Books learned this trick several thousand years ago.

Games occasionally forget it.

A small website that refused to die

Which brings me back to Mobstar.

I don't know exactly what combination of affection, habit, economics and stubbornness has kept its server alive.

But there is something delightful about discovering that it survived.

The Internet is extraordinarily good at erasing itself. Websites disappear, domains expire, databases get deleted, companies collapse, software dependencies rot, and entire online communities vanish without leaving much more than a screenshot and a broken hyperlink.

Yet here is this little mafia game my brother and I played when we were younger.

Still answering HTTP requests. Still letting people create accounts. Still keeping track of fictional money.

Still occasionally being updated.

Its graphics never became spectacular.

That may be part of why it suddenly feels interesting again.

In 2003, a game like Mobstar was mostly text and numbers because the web couldn't do very much.

In 2026, a developer could choose the same kind of simplicity for precisely the opposite reason.

Because computers can do almost anything.

And when everything is possible, deciding what not to render becomes a design decision.

Perhaps the next interesting frontier in games is not another order of magnitude of graphical fidelity. Perhaps it is worlds where players can do things nobody thought to put in the trailer.

Worlds whose complexity lives below the surface.

Worlds that look small and turn out to be enormous.

Mobstar may be a relic of an older Internet.

But relics sometimes preserve ideas the present has forgotten.

And this particular one is still running.