The Game Was Never Free: How the Industry Learned to Sell You the Same Thing Twice

Fair warning: this one runs long. But this isn't a story of a few years of spectacular fraud and bad decisions — this is decades of slow, deliberate erosion. Each age built on the last. The length is earned. If you're just here for the Valve loot box situation, scroll to The Reckoning. If you want to understand why it matters — and why it was probably inevitable — start here.

The gaming industry has a user-hostile problem. Loot boxes engineered around gambling psychology. Season passes charging you for content that doesn't exist yet. Live service games that vanish — along with your library — the moment a server goes dark. Games that ship broken on purpose because the patch is already written. Valve literally arguing in a New York courtroom that loot boxes are just like a surprise in a cereal box and we should all relax and enjoy the mystery.

It's a lot. And if you've been gaming for a while, you know something shifted — you can feel it even if the exact moment is hard to pin down. And if you're newer to gaming, this is just... how it is. Microtransactions, battle passes, always-online requirements, day-one patches — that's the water you've been swimming in since you picked up a controller.

But it wasn't always like this. And to understand how we got here, we need to hop in the DeLorean, get this thing up to 88, and go back to where it started — because the story of how gaming became hostile to the people who love it most is actually a pretty fascinating one.

I don't mean philosophically. I mean commercially. Legally. Practically.

What did you get when you bought a game?

For most of gaming's history, the answer was simple: you got the game. The whole thing. Whatever the developers built, whatever they could fit onto the physical media — cartridge, floppy disk, CD, DVD — that was it. That was the product. You took it home, you put it in your console or your drive, and you played it.

No updates. No patches. No servers to connect to. No account to log into. If the power went out, you lost your progress. If the game was broken, it stayed broken — forever, on every copy that shipped. The disc was the final word.

That constraint wasn't just technical. It was a form of consumer protection nobody had to legislate into existence. The physics of the medium created accountability. You had one shot to ship a finished product because there was no mechanism to fix it later. If it sucked, it sucked on shelves forever. Reviews would be brutal. Word of mouth would kill you. The market had teeth.

This is the era that younger players never really experienced as the baseline, and it's worth sitting with for a second — because everything that comes later makes more sense when you understand what the original implicit contract actually was.

You handed over your money. You got a complete object. The transaction was done.

N64 cartridges. PS1 discs. SNES carts. Even early PS2. These were finished artifacts. Mario 64 on that cartridge is the same Mario 64 it was on launch day in 1996. The game didn't evolve, didn't require a server handshake, didn't need a day-one patch to be playable. It just worked. Indefinitely. On hardware that may or may not still be running decades later.

The video game crash of 1983 nearly killed the North American market entirely. Atari had an essentially open licensing model — anyone could make a game, flood the market, and collect the money. The result was a tidal wave of garbage. Consumers got burned enough times that they stopped buying. The industry cratered. In E.T.'s case, quite literally — of the 5 million E.T. cartridges manufactured, only 1.5 million sold. Atari buried around 700,000 cartridges of unsold inventory in a New Mexico landfill, a story so absurd it was dismissed as urban legend for decades until someone actually went and dug them up in 2014. They were there.

Nintendo was watching all of this when they launched the NES in North America. They made a deliberate decision: we are not doing that. The result was the Nintendo Seal of Quality — which sounds like branding but was backed by a physical lockout chip, the 10NES, that literally prevented unlicensed cartridges from running. Publishers had to go through Nintendo, use Nintendo's manufacturing, and were limited on how many titles they could release per year. The seal meant something because the mechanism behind it had actual teeth — though the lockout chip was eventually reverse engineered by Tengen, who released unlicensed cartridges and argued in court that Nintendo's licensing practices were anticompetitive. They weren't entirely wrong. Quality control and market control aren't mutually exclusive, and Nintendo was running both simultaneously.

Here's the thing though — Nintendo's business interests and the consumer's interests were accidentally aligned by the constraints of the medium. Nintendo profited from quality control, so they enforced it. That's worth remembering.

Worth a quick note: Sony only got into gaming at all because Nintendo burned them. The two companies had a partnership to develop a CD-ROM add-on for the SNES — Sony built the hardware, Nintendo announced a Philips deal instead at CES 1991, blindsiding Sony on stage. Ken Kutaragi pushed Sony to build their own console in response. The PlayStation exists because Nintendo betrayed a partner. The console wars have roots.

CD-ROMs were cheap to press compared to cartridges. Lower manufacturing cost meant lower barriers to entry, which meant Sony was considerably more permissive with licensing. The PS1 library is enormous and wildly uneven as a result. You got genuinely weird, experimental games that Nintendo never would have approved — and you also got a lot of shovelware and rushed ports. Sony's attitude was closer to "get us a game, here's the dev kit." The medium was the only real constraint, and CDs didn't impose the same cost discipline that cartridges did.

But even on the PS1 — even with that lower bar — the disc was still the disc. You still got what you paid for on the day you bought it. The transaction was still complete.

That's the foundation. Keep it in mind, because every age that follows is the story of that foundation being quietly dismantled, one layer at a time.

Before we get to where things went wrong, it's worth spending a minute on when they were actually... fine. Because there was a period where paying for additional game content was a completely reasonable transaction, and I think it's important to name what made it work — because that's exactly what got dismantled later.

The expansion pack era was primarily a PC gaming phenomenon. Consoles were mostly still living in the cartridge/disc contract world. But on PC, if a game was successful enough, studios would make more of it. And then sell you that more. Separately. After the original shipped.

StarCraft. Then Brood War. Diablo II. Then Lord of Destruction. The Sims with its parade of expansions that were practically games in themselves. Age of Empires II and The Conquerors. Half-Life and its entire extended universe of expansions. These were substantial additions — near-game-sized content drops, often sold on physical media, that added campaigns, units, mechanics, story.

And here's the thing that made it work, the thing that's almost quaint to say out loud now: the base game was never touched.

Here's a test that cuts through all the noise about what "extra content" actually means. I call it the dusty PC test. Find an old machine. No internet connection. Dig out the original disc. Install it. Play it.

Does it work? Is the experience complete?

For StarCraft — yes. Full campaign across three races, skirmish mode, LAN multiplayer if you've got a second machine. Brood War is sitting right there on the shelf next to it and you don't need a single byte of it to have a complete, satisfying experience. Brood War is genuinely additional. It adds a new chapter to a story that was already finished. It doesn't reach back into the original box and remove anything.

Run that test on a modern game and we'll get to some uncomfortable results shortly. But in this age? The test passes every time.

No pressure baked into the base experience. No artificial gaps where content should be. No feeling that what you bought was incomplete. The expansion was genuinely optional because the thing it was expanding was genuinely whole.

The Sims is actually the most interesting case study here because it pushed the model further than almost anyone else and still mostly kept the integrity intact. By the end of The Sims 1's lifecycle there were seven expansion packs. Seven. That's a lot of additional purchases. But you could install the base game on that dusty PC right now and have a complete Sims experience. The expansions added neighborhoods, careers, vacations, pets — genuinely new systems — but they didn't hollow out the core to make room.

Everything that comes after the expansion pack era is the story of the industry figuring out that extraction is more profitable than addition. That if you architect the base experience around gaps, players will pay to fill them. That if you make the complete experience the premium option rather than the default, you've flipped the entire revenue model.

But we're not there yet. In this era, the deal was honest. You knew what you were buying. You knew what you already owned. And your dusty PC, twenty years later, would prove it.

That era didn't last. It couldn't — not once the industry figured out what the internet actually made possible.

For most of gaming history, physics was on the consumer's side. The medium had limits. Developers had one shot. Ship it, done. Whatever made it onto that disc or cartridge was the final word on what the game was.

Then broadband happened. But before we get to Xbox Live and Steam, it's worth noting that online console gaming didn't start there. The Sega Channel was selling subscription access to games via cable adapter back in 1994 — a monthly fee to download Genesis games into RAM. When you turned the console off, the game was gone. No disc, no cartridge, nothing owned. Pure access rental, thirty years before Game Pass. Japan had satellite-broadcast game services even earlier. And the Sega Dreamcast — launched in Japan in 1998, North America in 1999 — shipped with a built-in modem and real online gaming. Phantasy Star Online, NFL 2K1, actual multiplayer infrastructure on a console years before Xbox Live existed.

Xbox Live launched in 2002 and changed the calculus. Not the first connected console experience — but the first to unify it. One service, one friends list, voice chat across every game, Ethernet standard in the hardware. And critically — a $50/year subscription fee. The first service that made paying to access games you'd already bought feel normal.

Steam arrived in 2003. And the physics changed — quietly at first, with consequences nobody fully appreciated until it was too late to put back in the box.

Suddenly developers could push updates after launch. Bug fixes. Balance patches. Content additions. For a while — honestly, a good while — this was just straightforwardly good. Games that shipped with problems could be fixed. Multiplayer balance that was off could be corrected. The community could flag issues and studios could respond. The living document seemed like an upgrade over the locked artifact.

And it was. For a minute.

The accountability that the physical constraint had accidentally built in quietly dissolved. If you could patch it later, you didn't have to finish it now. "We'll fix it post-launch" became an answer to problems that previously had no answer — which meant it became an answer to problems that previously had to be solved before shipping. The pressure to be done was gone. And pressure, it turns out, was doing a lot of work — especially when time is money.

This is where the "going gold" process — the point where the disc gets finalized and sent to manufacturing, weeks before street date — stopped being the hard deadline it used to be. That gap between gold master and launch date used to mean: polish what you have, fix what you can, ship what's done. Now it means something else entirely. It means: ship the disc, have the day-one patch ready.

Sit with that for a second. We have reached a point where "the game was complete when it shipped" is remarkable. Seriously — what's the last AAA title you bought that didn't demand a patch on launch day before you could even start playing? Not counting indies, not counting the odd Nintendo title that still actually fits on a cart. I'll wait.

The disc in the box — if there even is a disc in the box — is increasingly a formality. Some Call of Duty physical releases are essentially a small installer that immediately downloads 50, 80, 100+ gigabytes of actual game. You bought a box and a coaster. The product lives on a server somewhere.

Nintendo. The company that built its entire early identity on the cartridge as a complete, trustworthy artifact. The company that implemented a physical lockout chip to ensure quality. Now ships Switch 2 games in two flavors. Some are actual game cards — Mario Kart World, Donkey Kong Bananza, the first-party titles. And then there are Game-Key Cards: physical cartridges containing zero game data, purely a key to download the game from the eShop. Insert it, connect to internet, download the actual game. The card is a receipt. Nintendo put a little key icon on the box so you know which kind you're buying — which is at least honest, in the same way a pickpocket handing you a note explaining they took your wallet is technically honest.

Nintendo's official explanation for Game-Key Cards was media costs and enabling larger games to ship physically. Which might even be partially true — except Nintendo also deliberately chose a proprietary card format over standard microSD, a decision that created the exact cost constraints they're now citing as justification for selling you an empty card. You don't get to manufacture the problem and then use it to explain the solution.

The cartridge that once was the contract is now available in a version that's explicitly, officially, just a key to a contract stored on Nintendo's servers.

Steam deserves more than a passing mention here because it didn't just arrive — it redefined what PC gaming was and set the table for almost everything that came after.

And it started with people absolutely hating it.

Half-Life 2 launched in 2004 and required Steam to run. A single player game. On your own machine. That you bought. Requiring an internet connection and a third party platform to authenticate. The backlash was real and loud — this was widely considered an outrage. An imposition. Who did Valve think they were, putting software between you and your game?

People installed it anyway. And then installed it for everything else. Because Steam was genuinely good. The interface worked. The seasonal sales became their own cultural events — games you'd been eyeing dropping to prices you couldn't justify ignoring. And those drops can be dramatic. A $70 launch day purchase becomes $17.99 six months later. I watched a brand new AAA title — released maybe six to eight weeks prior — already marked $15-22 off during a holiday sale. Still practically new. The ink barely dry on the reviews. The gap between what day one buyers paid and what patient players paid isn't marginal. It's not even defensible.

And watch what happens when a remaster drops — if the publisher hasn't already quietly delisted the original. Suddenly the $5 version you could buy yesterday doesn't exist anymore. The remaster needs the market to itself. Your only option is the $40-50 version with slightly shinier textures and possibly worse performance than the original that just disappeared.

And I mean that more literally than it might sound. Because "a section" undersells what PC game retail actually was.

EB Games. Software Etc. Babbage's. These stores — all eventually folded into what became GameStop — understood that PC gaming was a primary market, not an afterthought. Walk into an EB Games in its prime and you'd find console games occupying roughly the space you'd expect. Then you'd turn around and find an equally large, sometimes larger, section dedicated entirely to PC titles. Boxes organized by genre. Strategy. RPG. Simulation. First person shooters taking up their own row. Flight sims that came in boxes big enough to use as a small table.

And before that there were dedicated software stores — stores where games shared floor space with productivity suites, creative applications, educational software. The PC game wasn't just a gaming product. It was part of a broader software ecosystem that had enough retail gravity to support entire store concepts built around it.

Steam didn't just kill PC game boxes. It killed an entire retail category so completely that most people under a certain age have no frame of reference for what was lost. The store just... stopped existing. Gradually, then all at once.

But here's what that transition actually meant beneath the surface: it removed the last physical constraint from PC game development.

Console games still had to go gold for disc manufacturing. PC games increasingly didn't. The patch could go live the same day as launch, or before it, or after it — the timeline was entirely fluid.

Steam also normalized something that sounded minor at the time: you don't own your games, you license them. Your Steam library is a list of licenses Valve grants you access to. If Steam goes away — if Valve closes up shop, if your account gets banned, if the servers go dark — that library goes with it.

Early Access made this even more explicit. Steam launched it in 2013 and formalized something the industry had been doing informally — selling you an unfinished game, upfront, on the promise of future completion. Sometimes that works out. Sometimes you're still waiting. The money moved either way.

I'll be clear: I love Steam. The platform has done enormous things for PC gaming, for indie developers, for discovery of games that never would have found an audience through retail channels. The Legion Go S running SteamOS that I'm gaming on right now is a direct product of Valve taking PC gaming seriously as a platform worth investing in. None of that is nothing.

But it would be dishonest not to connect the dots. Steam's success in killing physical PC media is a direct ancestor of the incomplete-at-launch culture, the day-one patch normalization, and the DLC ecosystem that followed. It built the infrastructure that made all of it possible. The road to horse armor is partly paved with Steam install screens.

Launched December 2020 in a state bad enough that Sony pulled it from the PlayStation Store — something that almost never happens. CD Projekt RED had to offer refunds through their own channels. On last-gen consoles, PS4 and Xbox One, it was effectively unplayable. The gap between the game that was marketed and the game that shipped was embarrassing.

And then they fixed it. Patch by patch, update by update. The Phantom Liberty expansion landed well. The Edgerunners anime on Netflix brought a whole second wave of players in who'd never touched it. The discourse did a complete 180. It now sits comfortably as one of the better RPGs of the generation.

I played the original campaign on my old Steam Deck — on Linux, on handheld hardware that didn't exist when the game launched — and had a great time. That Deck has since been retired, replaced by a Legion Go S that runs the same library without juggling Windows underneath it. A game that was broken on PS4 hardware at launch runs beautifully on hardware that hadn't been invented yet. That's a genuinely strange sentence.

But here's the problem with the redemption arc: the industry learned exactly the wrong lesson from it.

No Man's Sky ran the same playbook before Cyberpunk did. Launched 2016, promises significantly outpaced delivery, Hello Games went nearly silent and just kept updating it for years. Now it's genuinely beloved, one of the great turnaround stories in gaming. Both studios deserve real credit for following through — most don't — but the existence of those redemption arcs handed every boardroom in the industry a permission slip.

Ship broken. Survive the discourse cycle. Patch it into shape. Profit.

The players who bought day one paid full price to be beta testers. The players who waited got a better, cheaper, more complete game. The incentives are completely backwards, and the industry knows it, and it doesn't matter because enough people buy on day one anyway to make the math work.

Once a game is a living document — always being updated, never quite finished, existing in relationship with a server somewhere — it stops being a product and starts being a service. And services have different pricing logic than products. You don't buy a service once and own it. You maintain a relationship with it. You subscribe to it. You purchase access to parts of it.

The disc was the contract. The internet dissolved the contract. And into that void, someone was always going to pour a new business model.

Several of them, actually.

April 3, 2006. Probably not a date you had circled on your calendar. Most people don't remember it at all. But it's the day everything changed.

Bethesda releases a small downloadable add-on for The Elder Scrolls IV: Oblivion on Xbox 360. The Horse Armor Pack. Two cosmetic armor sets for your in-game horse. No gameplay change. No story content. No new areas. Your horse looks slightly different. That's it.

Price: 200 Microsoft Points. Which worked out to $2.50.

The internet, collectively, lost its mind.

The mockery was immediate and total. Gaming forums lit up. Critics called it a cynical cash grab. Players called it an insult. Horse armor became shorthand for everything wrong with the games industry's direction — a punchline so durable it's still being made nearly twenty years later. It was widely, loudly, extensively condemned as a preview of a dark future nobody wanted.

And then people bought it. A lot of people. Enough people that every executive who saw the sales data understood something had fundamentally changed.

Not the mockery. The purchase data. The mockery was noise. The money was signal. And the signal said: players will pay for cosmetic content even when they're angry about it. Even when they're complaining loudly in one browser tab, they'll open the storefront in another. The gap between what people said and what people did was enormous — and the industry spent the next two decades systematically exploiting that gap.

But before we get to what came after, there's something worth unpacking about the mechanism that made Horse Armor possible. Because Bethesda didn't just decide to charge $2.50 for horse armor. They charged 200 Microsoft Points. And that distinction matters more than it sounds.

The Xbox 360 launched with its own proprietary point system for digital purchases. You didn't spend dollars. You spent Points. And the conversion rate was deliberately awkward — 80 Points to the dollar, meaning nothing priced cleanly in real money. A 200 Point item wasn't $2.50 in your head the way it would be if the price tag said $2.50. It was 200 of a number. The psychological distance from actual money was the entire point.

You'd buy Points in chunks — 500, 1600, 4000 — and you'd almost always have some left over. Leftover Points sitting in your account weren't money anymore. They were a balance in a system. Spending them felt different than spending cash. Less real. Less considered. More like using up tokens you'd already paid for.

This was not accidental. Every subsequent in-game currency system — V-Bucks, Apex Coins, Overwatch Coins, CoD Points, the entire ecosystem of fake money inside real games — descends directly from Microsoft Points. The lesson learned in 2005 has been applied thousands of times since: put real money through a conversion layer, and people spend more of it. The friction of thinking "this costs four dollars" disappears when it costs "320 coins" instead.

Microsoft eventually killed Points in 2013 and switched to real currency. By then the template had been copied by everyone. The abstraction layer had become an industry standard that didn't need a platform-level currency to exist — it just moved inside the games themselves.

Here's the other thing worth naming about Horse Armor specifically: it was cosmetic. Your horse looked different. Nothing about the gameplay changed. You weren't buying an advantage. You weren't buying story content carved out of the base game. You were buying the appearance of armor on a digital horse you rode occasionally.

That matters because it was the easiest possible case to make for paid DLC. Purely optional. Purely cosmetic. If you don't buy it, you lose nothing except the armor appearance. Bethesda could, and did, argue that nobody was forced to buy it. The base game was complete without it. This is true.

But it established something that couldn't be unestablished: players would pay for things inside games they'd already bought. The specific thing didn't matter. The precedent did. Once that door opened, every subsequent conversation about paid content wasn't "should we do this" — it was "how much can we charge and for what."

The horse armor was cosmetic. What came next often wasn't.

Horse Armor proved the concept. What followed was an industry systematically figuring out how far the concept could stretch.

The early DLC era had some honest entries. Small cosmetic additions. Bonus missions that genuinely were made after the base game shipped. Map packs for multiplayer games that extended the life of something people were actively playing. Annoying, maybe, but operating within the spirit of the expansion pack logic — here's more stuff, you don't have to buy it.

Then the incentives did what incentives do.

If you could sell additional content after launch, and players would buy it, the next logical question wasn't "what great content can we make?" It was "what content can we hold back?" And that question — once someone asked it in a boardroom and the math checked out — broke everything.

The most egregious version of the new model wasn't content made after launch and sold separately. It was content physically present on the disc you already bought, locked behind an additional payment. Not on a server somewhere. Not developed after release. On the disc. In your hands. Already paid for in the manufacturing cost of the product you purchased.

Mass Effect 3 launched in 2012 with "From Ashes" as day-one downloadable content — a full squad member, Prothean companion Javik, with his own story arc and missions. Players who didn't purchase it played through one of the most story-rich trilogies in gaming history missing a character whose presence had direct narrative relevance. Data miners found him on the disc within days of launch. The content wasn't downloaded when you bought it. It was unlocked. You paid for a key to a room that was already in the house.

Street Fighter x Tekken shipped with twelve additional characters locked on disc. Twelve. Capcom's defense was that keeping the data local reduced download times for players who purchased the DLC. The internet was not particularly moved by this argument.

This is the moment the expansion pack logic inverts completely. The expansion pack said: here's a complete game, we made more stuff after, want it? On-disc DLC says: here's most of a game. We finished the rest alongside it. Pay again to access what's already there.

And the part nobody in a publisher's PR department will ever say out loud: they could have shipped it all. The content was done. The disc was being pressed. The decision to withhold wasn't a development constraint — it was a revenue decision. "From Ashes" wasn't cut from Mass Effect 3 because they ran out of time. It was cut from the base game because someone in a meeting calculated that charging separately for it made more money than including it. That's it. That's the whole explanation. The creative work was finished. The business work was just getting started.

Run the dusty PC test on Mass Effect 3 without the From Ashes code. The disc loads. The game runs. But there's a squadmate-shaped hole in your crew that the game will never acknowledge, because acknowledging it would require admitting it was always supposed to be there.

At least on-disc DLC was finished. The season pass asked you to pay upfront for content that didn't exist yet. A bundle of future DLC, sold at a discount before any of it was made, on the promise of delivery at unspecified future dates. You were pre-funding development. The risk transfer was essentially total — from publisher to consumer.

Sometimes season passes delivered. Sometimes studios shipped half the promised content and went quiet. Sometimes the studio closed before finishing. Your money moved regardless. The legal exposure was minimal because the terms of service — that document nobody reads because reading it would take longer than the game itself — covered every contingency except you getting what you paid for.

It's not just that the content is missing. It's that the entire concept of "the game" becomes unstable. Is your Mass Effect 3 the real Mass Effect 3 if you don't have From Ashes? Is your Street Fighter x Tekken complete without the on-disc characters you didn't unlock? The base game was designed around the assumption of that content existing. The gaps are architectural, not cosmetic.

This is what separates the DLC era from the expansion pack era in the most fundamental way. StarCraft without Brood War is a complete StarCraft. Mass Effect 3 without its day-one DLC is a Mass Effect 3 with pieces missing that were finished before the disc was pressed.

One sold you a complete thing then made more. The other sold you an intentionally incomplete thing and called the missing parts optional.

The next age took that logic and built an entire business model around it.

In September 2017, Epic Games added a free Battle Royale mode to Fortnite — a game that had launched two months earlier as a co-op survival title nobody was particularly excited about. Within months it had hundreds of millions of players. Within a year it was generating billions of dollars annually from cosmetics that had zero impact on gameplay. Building materials and shotguns were free. The llama backpack skin cost real money.

Every major publisher watched those numbers and had the same thought simultaneously. As a friend of ours put it with rather elegant simplicity upon seeing something they wanted: "I need that."

What followed was one of the most expensive, sustained, and spectacular series of failures in entertainment industry history.

Fortnite was free. Completely free. The barrier to entry was zero. It arrived at exactly the right moment when battle royale was culturally exploding. It ran on everything including low-spec PCs and phones. It had the network effect of everyone already being there — your friends were playing it, which meant you played it, which meant your friends stayed. The cosmetic monetization worked because the player base was already massive, which meant people wanted to look good in front of other people.

You cannot manufacture those conditions. You cannot manufacture network effects by launching a competing product. You cannot buy the cultural moment that Fortnite landed in. But that didn't stop anyone from trying.

Anthem. BioWare's live service looter shooter, one of the most anticipated games of 2019, launched fundamentally broken and never recovered. EA eventually cancelled the planned overhaul and quietly let it die. A studio that made Mass Effect and Dragon Age spent years on a game that's now a cautionary tale.

Babylon's Fall. PlatinumGames — the studio behind Bayonetta and Nier: Automata — launched this with Square Enix in February 2022. Servers shut down in February 2023. Less than one year. The peak concurrent player count on Steam was 1,084. Not a typo.

Marvel's Avengers. Crystal Dynamics spent years building a live service game around one of the most valuable IP portfolios on earth. It launched undercooked, shed players rapidly, and the servers went dark in September 2023. If you bought cosmetics, they're gone. If you spent in-game currency, it's gone.

And then there's Concord.

Concord deserves its own paragraph because it represents something almost perfectly emblematic. Sony's Firewalk Studios spent reportedly around $400 million and eight years building a hero shooter to compete in a market dominated by Overwatch, Valorant, and Apex Legends — all of which were either free or had massive established player bases. Concord launched in August 2024 at $40. Peak concurrent players on Steam: approximately 700. Sony pulled it from sale two weeks after launch. Two weeks. Refunds were issued. The game no longer exists.

Suicide Squad: Kill the Justice League. Rocksteady — the studio that made the Batman Arkham games, some of the most beloved superhero titles ever made — spent years building a live service shooter set in that universe. It launched in early 2024 to immediate backlash and rapidly declining player counts.

Redfall. Arkane Studios — who made Dishonored and Prey, games celebrated precisely for their depth and craftsmanship — released a live service co-op shooter that felt like it was made by people who'd been told to make a live service co-op shooter and weren't entirely sure why. Servers shut down in 2024. The planned offline mode that Microsoft promised took months to actually arrive.

This is not a complete list. It's not close to a complete list. And it illustrates a problem beyond just the live service model itself — nearly every studio on that list had a proven identity, a genre they were genuinely great at, an audience that trusted them for specific reasons. Arkane made immersive sims. Rocksteady made single player superhero games. BioWare made story-driven RPGs. The publishers who tasked them with live service shooters weren't just chasing Fortnite — they were asking craftsmen to build something with tools they'd never used, in a style they'd never worked in, for an audience that hadn't asked for it.

Shockingly, that didn't generate blockbuster hits.

And when the losses came due, it wasn't just the failed projects that paid the price. Studios that had nothing to do with the live service gambles got shuttered anyway — collateral damage on the balance sheet, headcount reductions to satisfy shareholders who'd already gotten the bad news. Tango Gameworks. Arkane Austin. Alpha Dog. Roundhouse. Studios closed not because they failed, but because someone else did and the budget had to balance somewhere. The live service gold rush didn't just waste money. It burned studios that had nothing to do with it, taking years of institutional knowledge and creative identity with them.

Some of that's just gone now. Some of the people who made those studios worth anything — genuine artistic talent, designers who understood things about games that can't be taught from a document — left the industry entirely afterwards. Not laid off to be rehired somewhere else. Done. Gone to other fields, other lives. The games they would have made don't exist and never will.

You can reopen a studio. You can't recover what walked out the door.

For the live service games that did find audiences, the season pass evolved into something more psychologically sophisticated: the battle pass. Pay per season — usually around $10 every two to three months — for a progression track of cosmetics you unlock by playing. Fortnite popularized it. Now it's everywhere.

The battle pass is clever in ways worth naming. The theoretical pitch is that if you play enough you earn back enough premium currency for next season's pass — a perpetual motion machine of content. In practice the math is designed to come up just short, requiring a small top-up. The content is time-limited, creating genuine FOMO. Miss this season's items and they may never return. Your friends have the skin. You want the skin. The social pressure is the product.

When a live service game dies — and most of them do, eventually — it takes everything with it. Not just the game. Every purchase you made inside it. Every cosmetic. Every in-game currency balance. Every hour you spent unlocking progression. The transaction was never really a purchase in any meaningful sense. It was a rental that didn't tell you it was a rental.

Babylon's Fall lasted less than a year. Every player who spent money on cosmetics got roughly eleven months of use before it evaporated. Marvel's Avengers players who bought skins in 2020 got three years. Concord players who bought the game got two weeks.

The physical disc you bought in 1998 still works. Put it in the drive, it runs. The live service game you spent $200 on across two years of battle passes exists nowhere. It's gone as completely as if it never shipped.

At least the horse armor was yours. Absurd, overpriced, cosmetic. But yours. Load Oblivion today and the horse is still wearing it.

Here's the thing about building an industry on mechanics borrowed from gambling psychology: eventually the people who regulate gambling notice.

It took longer than it should have. But it's happening.

The secondary market for in-game items and currency had existed for years before anyone tried to formalize it. World of Warcraft gold farming was an actual industry — operations, predominantly in developing countries, running accounts around the clock farming currency to sell to players who'd rather pay real money than grind. Against the Terms of Service. Technically banned. Functionally unstoppable. Blizzard eventually introduced the WoW Token in 2015 — an official mechanism to buy gold with real money or convert gold to game time — essentially legalizing and taxing something they couldn't eliminate.

But Diablo 3 went further. When it launched in 2012 Blizzard built a Real Money Auction House directly into the game. Not a grey market. Not a third party site. An official in-game marketplace where players could buy and sell actual gear drops for actual dollars, with Blizzard taking a cut of every transaction — 15% plus a cash withdrawal fee.

The logic made a certain cynical sense: the secondary market existed anyway, players were already spending real money on items, Blizzard was capturing none of it. Why not capture it officially?

What it actually did was break the game at a fundamental design level. Diablo's entire appeal is the loot hunt — the dopamine loop of killing monsters, finding gear, getting stronger. The moment you could just buy the gear, the loop collapsed. Why farm when you can purchase? The economy the RMAH created undermined the reason to play the game it was built into.

Blizzard shut the Real Money Auction House down in 2014. The Reaper of Souls expansion removed it entirely and overhauled the loot system. It's one of the rare cases where a publisher tried something genuinely experimental, watched it destroy their own product's core appeal, and actually reversed course.

The lesson should have been: when real money mechanics enter the gameplay loop directly, they corrupt it. The industry heard something slightly different: the execution was wrong, not the concept.

Counter-Strike: Global Offensive had weapon skins — cosmetic finishes for guns that changed nothing about gameplay. Valve sold cases containing random skins, opened with keys purchased for real money. Some skins were common. Some were extraordinarily rare. The rare ones were worth real money on the Steam marketplace. Hundreds of dollars. Sometimes thousands for the rarest drops.

That secondary market created something Valve almost certainly didn't intend to officially sanction but also didn't move quickly to stop: a sprawling grey-market skin gambling ecosystem. Third party sites let players deposit CS:GO skins as currency and gamble them on match outcomes, roulette, coin flips. Real money, real stakes, zero age verification. The skins were the chips.

In 2016 it got worse. Prominent gaming YouTubers and streamers — TmarTm and ProSyndicate among them — were caught promoting skin gambling sites they had ownership stakes in without disclosing that relationship to audiences that skewed heavily young. They were, in plain terms, gambling influencers promoting their own casinos to teenagers while pretending to be regular players getting lucky.

The FTC and various state regulators took notice. Valve eventually sent cease and desist letters to the gambling sites. The ecosystem contracted but never fully disappeared. The underlying loot box mechanic remained completely intact.

In April 2018, Belgium's Gaming Commission ruled that paid loot boxes constitute gambling under Belgian law and should be banned. The Netherlands followed with similar findings. Publishers had a choice: modify their games for those markets or pull them. EA pulled FIFA's Ultimate Team loot boxes from Belgium rather than redesign the system. Which tells you everything about how central that mechanic was to the product — it was easier to exit a market than remove the feature.

Other countries moved slower. The UK Gambling Commission examined loot boxes and initially declined to classify them as gambling — then spent years revisiting that position as the evidence mounted. The US moved at the pace you'd expect from a regulatory environment where the industry being regulated has significant lobbying presence.

June 2019. Kerry Hopkins, EA's VP of legal and government affairs, testified before a Digital, Culture, Media and Sport committee examining loot boxes. Her argument: EA doesn't call them loot boxes. They're surprise mechanics. Like a Kinder Egg. People enjoy surprises. Surprises are fun. Children enjoy Kinder Eggs.

The committee was not particularly charmed by this framing. The internet was actively less charmed. "Surprise mechanics" became an instant shorthand for corporate euphemism so brazen it circled back around to being almost admirable in its audacity. EA was essentially arguing that the problem wasn't the mechanic — it was the vocabulary being used to describe it.

The mechanic remained in the games.

Valve is currently attempting to get a New York loot box lawsuit dismissed — and their argument will sound familiar. People enjoy surprises. CS:GO cases are entertainment products. The randomness is the feature, not a gambling mechanism. The ghost of Kerry Hopkins' Kinder Egg testimony haunts every paragraph of that filing.

The lawsuit argues otherwise. New York's consumer protection framework is one of the stronger ones in the country. The outcome isn't certain in either direction. But the fact that it exists — that a state AG is pursuing this in court rather than accepting the industry's self-description — represents something real shifting in the legal landscape.

A parallel legal and advocacy battle has been building around what happens when live service games die and take paid content with them. Ross Scott — the person behind the Freeman's Mind series — has been running the Stop Killing Games campaign for years, pushing for legislation requiring publishers to leave games in a playable state when they discontinue live support. The argument is straightforward: if someone paid for a product, that product should continue to function after the company stops supporting it. Shooting your own product dead and taking customer purchases with it should have legal consequences.

The industry's response — rarely stated this plainly but embedded in every architectural decision — is to make that technically as difficult as possible. When authentication lives on corporate servers, when core game logic runs on infrastructure the publisher owns and controls, when DRM requires a handshake with a data center that no longer exists, "leave the game in a playable state" stops being a policy question and becomes an engineering problem the publisher never planned to solve.

The Crew became the case that crystallized this. Ubisoft shut down the servers in March 2024 — not a free-to-play game, not a live service title people had ongoing subscriptions to, but a game people had purchased outright. Single purchase. Own it forever. Except the game required Ubisoft's servers to function at all, and when those servers went dark the product people had paid for simply stopped existing. Ubisoft revoked licenses from player libraries. Not just unplayable — gone, scrubbed from accounts, as if the transaction had never happened.

Which brings up the fine print most people skip past at purchase and only discover matters when something like this happens: you never bought the game. You bought a license to access it.

Technically this was always true — even for physical media. "This software is licensed, not sold" language has lived on the back of game boxes and in manual fine print since the 1980s. The movie industry runs the same way. That FBI warning at the start of every DVD wasn't decoration.

The difference is that physical media made the license practically unenforceable in the ways that matter to consumers. They couldn't remotely revoke a cartridge. They couldn't brick a disc. They couldn't scrub it from your shelf. The First Sale Doctrine meant you could legally resell physical games — which is the entire legal foundation GameStop's business was built on. The license said one thing. The physical object in your hand meant they couldn't actually do most of it.

Digital distribution handed publishers the enforcement mechanism they always claimed on paper but could never use. The license terms didn't change. The ability to act on them did. Now when Ubisoft decides The Crew is done, the license that was always technically revocable actually gets revoked — and there's no disc on your shelf they can't touch.

I've written about this before in the context of Sony — different situation, same underlying architecture of ownership that isn't really ownership. The more digital gaming gets, the more that gap between theoretical and practical ownership closes — in the publisher's favor every single time.

That specific case galvanized the Stop Killing Games campaign more than anything before it. Because The Crew wasn't a game that needed servers to function properly — it was a game that was architected to need servers so that the publisher retained control of the product after sale. The always-online requirement wasn't a technical necessity. It was a business decision dressed as one.

That distinction — between games that genuinely require server infrastructure and games designed to require it for leverage — is exactly what regulators are struggling to define. And exactly what the industry has every incentive to keep murky.

The campaign gathered enough signatures to trigger a European Citizens' Initiative requiring the European Commission to formally respond. The UK is examining similar questions. California has explored legislation around disclosure requirements for games with time-limited servers.

None of this is resolved. Most of it is moving slowly. The industry's legal and lobbying infrastructure is substantially more developed than the regulatory frameworks trying to catch up with it.

But the gap is closing. Slowly, unevenly, with the usual dysfunction of regulatory bodies trying to understand technology they're years behind on — but closing.

The horse armor cost $2.50 in 2006. The legal bills from what it started are considerably higher.

The DeLorean is back in the present. Let's look at what we came home to.

The honest answer is: it's complicated. Some things have genuinely improved. The structural incentives that created the problems haven't changed at all. And the industry has gotten considerably better at making extraction feel like a feature.

Star Wars Battlefront II is the clearest example of player backlash actually working. EA launched it in 2017 with a loot box system so aggressive it gated gameplay-affecting content — not cosmetics, actual hero characters — behind randomized paid drops. The backlash was immediate and nuclear. A single EA community response on Reddit defending the system became the most downvoted comment in Reddit history. Belgium and Netherlands regulators were already circling. And then Disney called — because it was their IP attached to the controversy — and EA pulled the paid loot boxes within days.

That's the lesson though. It took record-breaking public humiliation, active regulatory threat, and the intervention of one of the most powerful media companies on earth to make EA remove a feature from a single game. That's how high the bar is for player backlash to actually change something.

Battle passes largely replaced loot boxes as the dominant extraction model — which is arguably more honest (you know what you're getting) and arguably more insidious (the FOMO mechanics and time-