Part 8. PC and Online Games — How Free-to-Play Was Born

Core question: How do you give something away and still make money—and who first solved that paradox?

The Paradox Begins

“How can we make money if we give it away?”

The question sounds native to the twenty-first-century digital industry, but the PC-software industry first implemented the paradox as a business model in the late 1980s.

The answer was simple: let people experience it for free, then sell to those who want more.

The key is a migration of habit-space. Experience comes before the payment decision; experience produces willingness to pay, and that willingness forms the space. “I tried it and liked it” becomes evidence that “it is worth paying for.”

Arcades worked similarly. People watched, joined a line, tried a game, and inserted more coins. The difference was that the free arcade experience was watching; in PC games, it was direct play.

The Invention of Shareware — The Prototype of F2P

Around 1982, Andrew Fluegelman distributed his communications program PC-Talk with permission to copy it freely before formal purchase: try it and pay if you like it. He trademarked the term “Freeware.” Around the same time, Bob Wallace distributed the PC-Write word processor similarly and called it “Shareware,” which became the industry term.

Apogee Software brought the model aggressively into games.

The Apogee Model: Distribute the first third—episode one—for free, then sell episodes two and three. The Commander Keen series (1990), made by John Carmack's id Software team, established the approach.

The free episode was advertising with no marketing cost. Players copied floppy disks for friends, and word of mouth became distribution—the most efficient early-1990s PC-game channel.

id Software distributed Wolfenstein 3D in 1992 and Doom in 1993 the same way. Doom was estimated to be installed on about ten million PCs by 1994; only some users bought the paid version.

Shareware created willingness to pay through prior experience. It was the prototype of Free-to-Play. The name became “demo” and delivery moved to downloads, but the structure endured.

The Invention of the PC Café — Selling the Device Experience to People Without Devices

In mid-1990s Korea, PCs were expensive. StarCraft (1998) centered on network multiplayer, and many people without a home PC or internet still wanted to play.

PC cafés filled the gap. Customers paid by the hour to rent a PC and connection. The game was free; they paid for space.

Arcades charged the game itself per coin. PC cafés charged time for a space and device, separating payment from content.

Their number grew from about three thousand in 1998 to roughly twenty-two thousand in 2001 alongside Korean broadband. StarCraft sold an estimated 4.5 million packages in Korea, while millions also played without owning it.

The café made gameplay resemble a public good. For developers, network effects could exceed package revenue: more players meant easier matchmaking and greater game value.

This led to partnerships in which companies supplied software to cafés, and cafés provided access to customers paying hourly, often with no extra game fee. Free or inexpensive café access widened the user base.

Monthly Subscriptions — Paying to “Rent a World”

Ultima Online (1997) was the first large commercial MMORPG success in the West. Korea's Nexon had already commercialized the graphical MMORPG The Kingdom of the Winds in 1996; Guinness recognizes it as the longest-running commercial graphical MMORPG.

Ultima's structure was simple: buy the package, then pay monthly. Without the fee, the player could not enter. It was rent for a world.

This differed fundamentally from ownership. Characters, items, and memories existed only while the service lived; players rented rather than owned.

People accepted it. Ultima gained roughly one hundred thousand paying subscribers in six months. EverQuest (1999) and Dark Age of Camelot expanded the model.

The world had to feel large and alive, shared with others. Community justified payment: develop a character, join friends, leave a name in the world.

Korea's Lineage (1998, NCsoft) emphasized PvP war, castle sieges, and clan politics. Users paid not merely to enter but to participate in its power structure. It became a top-earning early-2000s Korean internet game, grew with PC cafés, and expanded into Lineage II (2003) and mobile Lineage M (2017).

Blizzard's World of Warcraft (2004) marked the peak: roughly eleven to 11.5 million global subscribers by late 2008 at $14.99 a month.

Failure: Star Wars Galaxies NGE

Sony Online Entertainment launched the MMORPG Star Wars Galaxies in 2003.

In November 2005, its New Game Enhancements replaced thirty-two specialist professions with nine iconic classes and made combat action-based. After World of Warcraft gathered millions, Sony reportedly sought a more mainstream design.

Forums and communities erupted. Sony did not publish precise subscriber figures, but reduced server activity, waves of departure posts, and increased account sales on eBay were observed.

The service ended in December 2011. As of 2023, tens of thousands still used fan-run SWGEmu servers restoring the pre-NGE game.

Nexon's Innovation — Why It Chose Free Instead of Subscription

Nexon's free-to-play model began with QuizQuiz (1999). After subscriber loss under a monthly fee, it became free in 2001 and sold avatar decorations—one of the earliest examples. MapleStory (2003) adopted the design from launch: play was free, while a Cash Shop sold appearance items, conveniences, and experience boosters in small purchases.

The monthly standard charged every player equally. F2P made enjoyment free and charged selectively for extras.

The economic implication was crucial. Subscription revenue equals subscribers times fee. F2P revenue concentrates among a minority of heavy spenders, or “whales.” Few pay, but they produce most revenue.

Global MapleStory entered North America as F2P in 2005, helping introduce the cash-shop model to the West.

Free users also enriched the world: more party members, opponents, and spectators. As the audience grew, the display value of spending increased. Free users strengthened paying users' motives.

Western observers initially called “free but designed to make you pay” deceptive. The numbers persuaded them: monthly users far exceeded subscription games, while average spending per payer rose. By the mid-2010s, F2P plus cash shop became the mobile standard.

Variations on Subscription — Subscription-Free MMORPGs and Hybrid Models

Subscription was never the only MMORPG model.

Guild Wars (2005, ArenaNet): Buy once and play indefinitely without a monthly fee. Expansion sales funded it, supplemented by character slots and cosmetics. It sold more than one million copies without subscription during WoW's dominance.

RuneScape (2001, Jagex): A browser MMORPG playable entirely free, with more content for paying members. It remains active in 2023, including the separate Old School RuneScape.

Aion (2008, NCsoft): Launched with a monthly fee, then switched North American and European servers to F2P in 2012, increasing concurrent users. It became an early example of recovering a large MMORPG's base through conversion.

The pattern spread in the 2010s: Star Wars: The Old Republic launched in 2011 and added F2P in 2012; The Lord of the Rings Online launched in 2007 and converted in 2010. WoW became an exceptional subscription holdout.

ItemBay and Real-Money Trading — A Payment Habit the Developers Did Not Create

ItemBay launched in 2003 as a Korean market for online-game items and characters bought with real money.

Users, not companies, created the payment space. Once rare items held real value, a market emerged naturally.

Real-money trading occupied a legal gray zone and violated most terms, yet items from Lineage, Lineage II, and World of Warcraft traded on eBay and Korean platforms.

In the early 2000s, eBay already sold items and accounts. Edward Castronova's 2001 study calculated that EverQuest's virtual Norrath had per-capita GDP—converting virtual productivity at real exchange rates—higher than Russia or Bulgaria.

Companies either suppressed or formalized trading. They banned gold-farming and trading accounts, or absorbed the market. EVE Online operator CCP introduced PLEX in 2009: buy it with real money, sell it to another player in-game, or exchange it for subscription time. Some users now fund subscriptions entirely through play.

Steam's Sale Culture — Reframing Perceived Price

Valve launched Steam in 2003. Early resistance asked, “Why install a client for a game I bought?” Requiring Steam for Half-Life 2 in 2004 rapidly expanded the base.

Its strongest habit was the Steam Sale: summer, winter, Halloween, and other mass events, sometimes up to ninety percent off.

Players learned, “Do not pay full price.” They wishlist games and wait, promoting strategic rather than impulsive purchases. Developers sell a smaller share at list price but gain volume through access.

Steam also opened indie distribution. Greenlight (2012) used community votes for small games and was replaced by Steam Direct in 2017.

A player who bought a $60 game for $6 now sees a $30 game and thinks, “Wait for the sale.” Resistance to list price rises. Developers respond with bonuses for “day-one buyers,” trying to create a separate habit-space for launch purchases.

Paradoxically, wallets open during sales: players buy many $2–$5 games because Steam trained the reference “at this price, it is safe.” Wishlist alerts and flash deals exploit that standard.

GOG and DRM-Free — Reasserting Ownership

While Steam dominated distribution, GOG experimented in another direction.

GOG (Good Old Games, 2008, CD Projekt): It began by selling games without DRM. Buyers could download and keep files without a client. Steam required client login; GOG offered “true ownership.”

Its initial customers opposed DRM philosophically or collected classic games. GOG made 1980s–90s PC games work on modern systems, then expanded into new releases.

Its share is far below Steam's, but its existence proves a separate habit-space for paying for ownership. Some users pay a premium for DRM-free games.

Parent CD Projekt also develops The Witcher and Cyberpunk 2077 and offers its own and third-party titles DRM-free.

Early Access — Prepaying for the Unfinished

Minecraft began paid sales in June 2009 at about €9.95 while still an early one-person project by Markus “Notch” Persson. Formal release came in 2011.

Buyers accepted bugs and sparse content for two reasons: a lower pre-completion price and participation in development. Their feedback could change the game.

Minecraft sold millions before release, with revenue far exceeding development cost.

Steam formalized the model as Early Access in 2013, allowing works in progress to receive preorders.

It created a new habit-space: buying the unfinished—not finished ownership, but purchasing participation in development. The identity resembled a backer more than a subscriber.

Successes include Palworld (January 2024, roughly fifteen million Steam copies), Subnautica (2014–18), and Dungeon of the Endless. Yet abandoned projects and poor “full releases” damaged trust in Early Access.

An Unexpected Connection — Twitch and the Economics of Watching Games

Twitch separated from Justin.tv in 2011 as a game-streaming service before Amazon acquired it in 2014. Watching rather than playing extended arcade spectator culture worldwide through the internet.

Its payment structure includes:

• Channel subscriptions: $4.99/$9.99/$24.99 monthly, shared by streamer and Twitch, with emotes, badges, and ad removal.
• Bits: Twitch currency sent with cheers in chat. 100 Bits cost $1.40; the streamer receives $1.
• Donations: Direct transfer through services such as Streamlabs, often read aloud with the message.

It matches Korea's AfreecaTV Star Balloons introduced in 2007, and both transplant the tip culture of offline performance, such as pansori patronage.

Canceling a Twitch subscription does not remove content: subscribers and non-subscribers can both watch. The real benefits—emotes and badges—say in chat, “I support this streamer.” Payment expresses identity rather than access.

The user pays to support a relationship and belong to a community, not to own content.

In the 2020s, AfreecaTV faced CHZZK and YouTube Live, reshaping Korea's streaming market through 2023–24.

Habbo Hotel — A Furniture Store Selling Social Space

Finnish company Sulake launched Habbo Hotel in 2000: neither MMORPG nor conventional game, but a social virtual space.

Users decorated pixel characters and rooms, chatted, visited others, and invited friends.

The revenue model was simple: sell furniture. Real money bought virtual room furnishings, with rare and special items costing more.

Unlike power items, the purpose was aesthetic completion: paying to beautify “my space.” Interior-consumption habits landed in a digital world.

Around 2010, Habbo had roughly ten million monthly active users worldwide and was especially popular among teenagers. Furni traded by scarcity, and informal trading scams were common.

Users declined after 2012 as smartphone social media spread. The service continues at a smaller scale.

Twenty years before “metaverse” became fashionable, Habbo already combined virtual space, social interaction, and item purchases—a structure returning in Zepeto, Roblox, and Fortnite Creative.

GameSpy Shuts Down — The Risk of Infrastructure Dependence

GameSpy supplied server-matching infrastructure for many PC multiplayer games from the late 1990s through the 2000s.

IGN acquired it in 2004, sold it to Glu Mobile in 2012, and the service closed on May 31, 2014.

Dozens of dependent games lost online multiplayer, including PC Halo, Battlefield 1942, Crysis, and Star Wars Battlefront II (2005).

Fans restored some through unofficial servers; developers patched others to new infrastructure; some lost multiplayer permanently.

The event exposed that the value of a purchased game could depend on external services. Players thought they owned a game but held only a right to access a service. As always-online requirements spread after 2000, this became a structural issue.

This Chapter's Hints — Checkpoints for New-Content Planners

Checkpoint 1: When offering “free entry,” where will you create the path to payment?

Shareware sold episodes to those wanting more; PC cafés charged rent for space; Nexon built a cash shop. At the boundary between free and paid, there must be something valuable—not simply something removed, but something desired.

Checkpoint 2: Who among your users is the “whale”?

Most F2P revenue comes from a few heavy spenders. They pay both to enjoy more and to be seen. More free users raise the display value of spending. The groups need each other.

Checkpoint 3: How will you treat a payment habit-space created by users?

ItemBay was not designed by a developer. Suppression and formalization each have consequences; EVE absorbed trading into official revenue. A spontaneous market can be read as a clue to the formal model.

Checkpoint 4: Which infrastructure does your content depend on?

GameSpy showed that content value can disappear with infrastructure. Third-party services, platform policy, and network effects are foundations. Design for what happens if they change.

PC Gamers' Payment Psychology — Why the Same Tricks Fail

Energy systems, D+0 starter packs, gacha, and rewarded ads—the mobile techniques examined in Part 9—often fail in PC online games because the payment habit-space differs.

1. Price sensitivity: Waiting trained by Steam sales

PC gamers repeatedly see $60 releases fall to $12–$15 within a year. “Wait and it gets cheaper” is embodied knowledge. They resist full price and react less to FOMO offers such as “today only.”

Yet sales open wallets: $2–$5 games are bought in batches because Steam established an acceptable reference. Wishlist alerts and flash deals convert that habit.

2. Ad avoidance: A habit shaped by browser extensions

AdBlock and uBlock Origin usage is much higher on PCs. Gamers accustomed to ad-free browsers find “watch an ad for a reward” unfamiliar. This helps explain why ad-funded PC games are rare, while mobile hyper-casual succeeds. PC attempts meet blockers or the response, “Why are there ads in this game?”

3. Session length: Why energy systems fail

Mobile energy assumes five-to-fifteen-minute fragmented sessions. PC users commonly sit for one to three hours. If energy ends, they can choose another of many games rather than wait and return, making the mobile pattern weak.

4. Community scrutiny: Review bombing and public debate

Steam reviews, Reddit, YouTube criticism, and specialist media analyze and criticize business models in real time.

When users dislike monetization, operations, or a developer statement, review bombing directly harms exposure and sales. Fallout 76 faced the $76 collector-edition canvas-bag controversy in 2018; EA's Battlefront II faced pay-to-win gacha criticism in 2017; multiple Activision titles have been targeted. App-store reviews exist on mobile but lack the same coordinated media infrastructure.

5. DRM resistance: Distrust created by SecuROM and GFWL

PC gamers have deep DRM fatigue. SecuROM and StarForce in the mid-2000s to early 2010s reportedly reduced performance or conflicted with software. Server-authenticated games sometimes stopped working when authentication servers closed.

Microsoft's Games for Windows Live, introduced in 2007, was criticized for instability and poor UI. Its effective shutdown in 2014 created access problems for purchases and saves.

Accumulated experience taught users that purchases last only while platform services do. GOG answered with DRM-free files and the promise, “What you buy is yours,” a powerful reason to pay for a specific group.

6. The paradox of perceived ownership

PC gamers feel games accumulated in a Steam library are “mine,” although account suspension removes the whole library. It is a license that looks like ownership.

This makes them especially sensitive to digital-ownership issues. Strong reaction to Sony's planned Discovery removal in 2023 and PSN linking in 2024 reflected awareness that the same loss could apply to Steam libraries.

Conclusion for PC-online business models: persuasion, not manipulation

Where mobile models lean on triggers and behavioral steering—energy depletion, D+ sequences, rewarded ads—PC models evolved to persuade users that payment is reasonable.

• WoW subscription: this world is worth entering each month.
• Steam sale: this price is reasonable.
• Nexon cash shop: this appearance is worth my money.
• EVE PLEX: this in-game resource has real monetary value.
• DLC: this additional content is worth its price.

Without persuasion, PC gamers do not pay. The model depends less on “buy or be unable to continue” and more on “buy and it becomes better.” That is both its defining quality and its limit.

The next chapter moves to mobile: how smartphones changed game business models, the invention of $0.99, the final success of Angry Birds, the origin of gacha, and the journey of the arcade coin into the palm of the hand.

Kim Dongeun WhtDrgon@MEJE.kr 2026