'Artifact' Isn't a Game on Steam, It's Steam in a Game

Proxies don't exist in Artifact. That's the point.

If you’ve ever been involved in competitive Magic: The Gathering, there’s a decent chance you’ve encountered what’s known as proxy cards. Playing Magic seriously is an expensive endeavor, and a competition-ready deck can set you back $1,000 or more. So, if you want to play—but not pay—like a pro, you use a proxy card: A homemade simulacrum of a costly, unowned card that, with two players willing to shake on it, stands in for the real thing. They aren’t strictly legal, but they are pervasive enough to be accepted (one online tool will even print an entire deck of them for you). And, in any case, Magic publisher Wizards of the Coast’s provision against proxies is about as impotent as the “don’t remove this tag” message on your pillow. As a result, proxy cards have become part of the experience of competitive Magic. Whether you’re just getting into the scene, or a serious player testing out new permutations of cards, proxies offer something akin to the “real” experience without the financial burden.

But when it comes time to compete in official tournaments, you’ll need genuine cards. Judges disqualify players caught using proxies, an embarrassing and potentially costly penalty for any serious competitor. To avoid that, aspiring players have two options: one is to buy booster packs and hope the cards they seek are contained within. The other, less chancy approach is to seek cards out on third-party exchanges, like TCGplayer and Card Kingdom. If you need a staple of competitive play like, say, Scalding Tarn, you can shell out $71.22 and have a well-worn copy in your mailbox by the end of the week. In essence, would-be competitors face a choice: submit either to the whims of cellophane fate or the predictable (if unforgiving) logic of supply and demand.

Proxy cards and online trading have lowered the practical barrier of entry to competitive Magic, even if the financial one remains daunting. But for Wizards of the Coast, which has published Magic since 1993, they amount to bootlegging and aftermarkets, respectively. The company can, of course, adjust its terms of service and add holographic watermarks, but when your product is, physically speaking, an arrangement of pigment on cardstock, there’s only so much control it affords. As a result, Wizards of the Coast has mostly resigned itself to the status quo. Magic, of course, is an extraordinary game, which accounts for its enduring appeal. Even so, if the real game you’re playing is maximizing returns on investments, it’s hard not to look past millions of happy Magic fans and see a tremendous amount of value being created, and then lost, slipping away like sand through fingers.

“A game without a metagame is like an idealized object in physics. It may be a useful construct, but it doesn’t really exist,” wrote Magic’s legendary designer, Dr. Richard Garfield, in a 2000 essay on his design philosophy. Garfield’s point is that games, no matter their format, bleed. Tempting though it is to mark out where a game “ends” and the rest of the world begins, such boundaries always break down under close scrutiny. This is true of all games, in one way or another, but it’s especially clear with Magic, whose aftermarkets and proxies are inescapably “part” of what the game is in the world. It’s also true, then, of Dr. Garfield’s latest card game, Artifact, which was released on Steam last week and will arrive on mobile sometime next year.

Here’s another truth: Artifact is extraordinary. Dr. Garfield may be a mathematician by training, but a he’s a designer by calling, and Valve is still home to some of the most talented artists and developers in games. To play Artifact is to witness masters at the peak of their craft. Everything about Artifact—from the pleasing rustle of cards as they unfurl across the board and to the ways your held cards respond, ever so gently, to the touch of your cursor—displays unmistakable signs of talented people giving a shit.

As a longtime Dota 2 fanatic, I’ve especially enjoyed the artful translation of Dota 2 mechanics into the conventions of card games, of which its three simultaneous lanes is only the most obvious example. It’s hard not to marvel at the depths of Artifact, which, to me, feels like Hearthstone in three dimensions. To be sure, I am not an especially gifted or consistent player of card games, and yet I find myself compelled to go deeper, pulled in as if by magnetism. Artifact is the rare game whose complexity does not make you feel stupid, so much as it inspires you to be smarter.

This makes it a bit surprising—and a bit of a tragedy—that the game is being pilloried by its target audience, and, based on its already declining player base, is on a direct path to an embarrassing free-to-play resuscitation. With 10,975 reviews in on Steam, Artifact’s score is“Mixed”, which, on Steam, is a sure sign that something has gone seriously wrong. It doesn’t take a lot of searching to uncover the culprit: How Artifact is monetized.

“This is not a card game, it’s a credit card game,” writes one user in a typical negative review, upset that there’s (a) no in-game currency, the way that most multiplayer games (especially free-to-play) have a currency that you can earn through playing (b) it costs money to compete in a lot of game modes that are designed for advanced and competitive players and (c) that the game more-or-less forces players to buy and sell cards on the Steam Marketplace. Early estimates suggest that getting a full set of 310 cards will set you back around $300 – less than an elite Magic deck, but 15 times as much as Artifact costs. Most of these reviews concede as a matter of course that Artifact is, unquestionably, a great game. (It is.) It’s the metagame that’s the problem.

This poses a dilemma for reviewers. Artifact, clearly, is a flashpoint for ongoing debates about the relationship of mechanics to monetization. Artifact’s woes are not just because of how it separates players from their money, but because of how flagrantly it ties these two arenas together, the ritual of play bound and gagged to the vulgarity of buying and selling. This challenges one of the underlying assumptions of what, exactly, we’re reviewing when we review games. Every game may indeed have a metagame, and yet, by common consensus, we reviewers tend to focus on what’s “in” the game, not around or on top of it. Artifact exposes this as the performance it always has been, a distinction as imaginary as any border is. What if, instead, we didn’t review Artifact as the excellent game so many wish it only were? (Plenty of these exist already). What if we reviewed it for what it really is: a market—one Magic could never forge, and only Valve would dare to.

If you’re interested in understanding Valve, and, how, in turn, this led them to make Artifact the way they did, you need to look at the history and, in particular, its knack for finding novel ways to make money. Valve—which is not publicly traded, maintains one of the smallest workforces of any firm of its prestige, and is thought to have the highest profit per employee of any American company—is not like other game companies, and operates according to a different, but specific, set of principles. That fact makes Valve both fascinating and predictable.

“Very few of Valve’s innovations center around the creation of original, inhouse IP,” explain game theorists Stephanie Boluk and Patrick Lemieux in their 2017 book Metagaming, a long reflection on the many things we can do with video games in addition to playing them. “Instead, they have developed a business model based on colonizing, expropriating, and assimilating metagames into a framework of benevolent capitalism.”

When Gabe Newell quit Microsoft to launch Valve back in 1996, the company was, practically speaking, just another game developer. What distinguished it from its competitors was mostly the quality of its products. The game that made Valve famous, Half Life (1998), might have been a watershed moment for game design, but it was made, marketed, and monetized like any other computer game at its moment. Valve’s innovations lay “inside” video games, not in the spaces around them.

Then Steam hit. Launched in 2004 alongside Half-Life 2, the program that was originally intended to keep Valve’s own games up to date quickly became the most important retailer in computer gaming, capturing some 70% of the PC distribution market within a few years of launch. By 2012, other companies had caught on and began to introduce their own distribution platforms alongside flagship titles, giving us the Origins and Uplays we deal with today.

Valve, once again, was a step ahead. Taking a cut from every purchase on Steam might have created tremendous profits at minimal expense for Valve, but it only did so at the point of sale. To expand its market, as is the guiding principle of capital, the company introduced the Steam Marketplace in 2012, marking Steam’s transition from a distributor to something much more complex. Consumers could (if they desired) participate in a market of virtual goods, buying and selling skins for games like Counter-Strike, Dota 2, and Team Fortress 2. At that moment, Valve stopped being a company that makes games and became one that makes markets, which means that Valve’s philosophy of management had to change as well.

“Our job is to maximize the productivity of users in creating digital goods and services,” Newell said in a frank interview in 2012 at the University of Texas. “The markets will determine what the marginal value add of each of those activities are. The kinds of ways in which people create value and creativity and creating frameworks for that are gonna vary.”

Steam stopped being a digital Gamestop and became something more akin to Uber, a massive platform to distribute labor (or, more accurately, it became both). Like “regular” markets, the Steam Marketplace offered familiar financial opportunities—supply and demand, arbitrage, and high-volume algorithmic trading—to enterprising users. Unlike regular markets, however, Valve set the terms of every exchange and skimmed a bit off the top of every transaction. In effect, Valve could keep extracting value for as long as players were willing to buy and sell virtual goods on their platform. If Steam 1.0 connected studios to players for economic exchange, Steam 2.0 did so for players to players. Once again, Valve had devised a way to sit back and make money while pushing the labor (and risk) everywhere else.

If profit is your ultimate goal, then this logic is nothing short of genius, which is why it’s the guiding principle behind everything Valve does. It’s not enough to say that Valve wants to make money (duh); they only really want to do so in a particular way. In a sense, gaming’s most mysterious company is predictably unique and uniquely predictable. Artifact is only the latest example of a long-running logic, and we shouldn’t be surprised by a company acting like what it has always been. But Artifact isn’t just about Steam—or, really, Steam isn’t just about Artifact. Valve’s platform happens to be the perfect tool for monetizing Magic’s metagames, too.

When you purchase Artifact ($19.99) and launch it for the first time, you play two tutorial games: one with a Red/Green Brawler deck, and a Blue/Black Control deck. Upon completion, Artifact grants you each starter deck so that you can get playing right away. For obvious reasons, cards in the starter decks are the cheapest cards to buy in Artifact. Were I to put one of my two Oglodi Vandals up for sale for a $0.03 (the minimum price for which a card will sell), I would receive $0.01 in return. Those other two cents would be profit for Valve.

Those decks are supplemented by 10 card packs (worth $1.99 each) of 12 cards, one of which is guaranteed to be “Rare.” Alas, my Rare cards haven’t turned out to be worth much more than my common ones, save my prize card, Horn of the Alpha, which is worth (as of now) a respectable $5.09. I’ve also gotten decently lucky on Heroes, given that they determine what color of spells you can play, and thus the breadth of possible decks you can build. I might not have any of the winningest Heroes—Tinker, Drow Ranger, and Axe—but at least I’m not stuck holding an Outworld Devourer, which currently boasts an abysmal 33% win rate.

If I wanted to get competitive, though, I’d need different heroes—probably Axe and Legion Commander, by far the most popular heroes in a recent tournament. As in Dota 2, heroes are ultimately what the game revolves around, not simply because they’re far more powerful than the NPC “creeps" that randomly populate each lane, but also because they determine what color of spell you may play in those lanes. As a result, having a strong lineup of heroes is a (the?) central component of a successful deck. If that’s my goal, I’ve got three options for getting them into my collection. Like Magic, I can open packs at $1.99 each until either Axe or Legion Commander show up. Alternatively, I can trade for them on the Steam Marketplace, where demand for top cards like Axe (currently $21.71, but sold for as high as $38.73 last week) has pushed the cost of individual cards above the price of Artifact itself.

My third option, nominally the most meritocratic one and where most of the ire over Artifact comes from, is to compete. It’s all a bit arcane—enough to require an elaborate FAQ with apologia about its complexity—so bear with me. Artifact divvies up all PvP competition into three modes: “Social Play” (challenging your friends and user-generated tournaments), “Casual Play” (bot matches and unranked matchmaking), and “Expert Play” (which I’ll come back to, since it’s really the meat of the matter). Both “Casual Play” and “Expert Play” contain “Gauntlets,” Artifact’s more malleable take on the Hearthstone Arena system.

You can do as much Casual and Social Play as you’d like without paying a thing, and I suspect many players will. But if you want to win some cards for your collection on the back of your Artifact prowess, you’ll have to do Expert Play: “a series of gauntlets that offer prizes and greater risk vs. reward.” The risk, here, is event tickets and the reward is card packs. Because Artifact doesn’t have an in-game currency, competing in Expert Play requires spending event tickets (you get five when you buy the game, and they’re available in bundles of five for $4.95 thereafter). It’s also the only way to win new card packs.

Expert Play is broken down into kinds of Gauntlets: Expert Constructed (play with your own cards), Phantom Draft (build a deck from five packs, but you don’t get to keep the cards), and Keeper Draft (same as Phantom Draft, but you get to keep the cards). Expert Constructed and Phantom Draft require one ticket each, and, should you win fives games without losing two, will yield one ticket and two card packs (whew). Keeper Draft, on the other hand, requires buying two tickets and five card packs (meaning it’s ~$12 to play each round of Keeper Draft Gauntlet), but you can win up to three additional packs on a good Gauntlet run. Draft modes aren’t for everyone, of course, and if you want to win cards with your own hard-earned collection, you’re stuck playing Expert Constructed. But it doesn’t take a lot of straining to see the problem: to win (reliably at least) you need great cards, but to get great cards you need to win. And the only way guaranteed to break that feedback loop is to buy what you need on the Steam Marketplace. In essence, there is no way to get new cards in Artifact without paying in one way or another (or two). No wonder Artifact’s critics are calling it pay-to-pay-to-win.

Many have wondered whether or not Artifact’s reliance on the Steam Marketplace is an invitation to price gouging. Yet Valve, in contrast to its typically flippant attitude towards media, has gone to some lengths to get ahead of these assumptions, speaking at length on the topic in an interview with IGN. Asked whether it were possible for a single card to reach $1,000, no less an authority than Dr. Garfield himself responded that this “cannot happen, and in no way we would want to happen. It can’t happen because [of] the arbitrage. As long as we’re publishing cards, there’s a maximum value any particular card can have.”

“We actually do want people to play, and we want you to be able to play with your friends,” added Brandon Reinhart, a programmer at Valve. “So making sure there's a reasonable price ceiling is really valuable for that.”

It’s easy to be cynical, of course, and claim that Dr. Garfield and Reinhart are just feeding us a slurry of convenient lies. But I’d prefer to take them at their word, in part because of the distinctiveness of how Valve profits. It’s entirely possible to buy every common and uncommon card in the game for a couple of dollars. Kotaku’s Ethan Gach reports building a viable Black deck for $15.92, and even a tournament-ready deck will, at present, cost only $30 or so dollars—far, far less expensive than Magic ever was. And as more cards enter the ecosystem, the supply of Rare cards will necessarily rise, and prices will, in turn, drop. And, of course, with multiple “casual” game modes, it is possible to enjoy Artifact without spending a penny beyond your initial investment (as I have). Doing so limits you to a few modes—unranked casual games, bot matches, friend-to-friend challenges, etc.—but, for a great many players, that may be enough. The issue is what happens if you want more.

Like all commodities, Artifact cards have what Karl Marx called “use value” (what you can do with a commodity, e.g. play Axe in your center lane and wallop enemy creeps) and “exchange value” (what you can get for a commodity on the open market). This is true of every virtual good you can find on the Steam Marketplace, from Counter-Strike guns to Dota 2 maps, but the entanglement of these two kinds of value is particularly apparent in Artifact. Like most card games, Artifact—through its open API and Valve’s willingness to accommodate third-party analysis sites—embraces tier lists, power rankings, win rates, and all the other hallmarks of data-driven deck-building. But given that every card is, also, a few short clicks away, in-game power and market value are entangled as never before. (This means, also, that balancing cards can itself be a form of price manipulation, one that goes beyond the self-correcting, “invisible hand” logic of arbitrage that Dr. Garfield points to). Artifact encourages you to know the value of every card you have, at every moment, both as a weapon (i.e. use value) and a tradable commodity (i.e. exchange value). Everyone who buys Artifact becomes, like it or not, an entrepreneur too. Now that’s alchemy.

Ultimately, I don’t have much interest in arguing whether or not Artifact is “worth it”—that answer lies with each player, who is free to participate as much or as little as they choose. What interests me is the underlying structure of Artifact’s market, and what makes it distinctive from its competitors. Artifact’s supporters are right to point out that the game will very likely cost less to play than Hearthstone, which, for very serious players, can cost up to $400 per year. Yet the Hearthstone comparison misses something important about Artifact’s economy: There is no “trading” in Hearthstone (unless you count buying from Blizzard as a kind of trade), whereas Artifact is built on it. Artifact doesn’t have to cost as much per year to you to be just as valuable to Valve, because it’s players—not Valve—who take on the labor of buying and selling. Skimming two cents off of one million common card transactions ($20,000), is, ultimately, just as valuable as selling 1,000 copies of the base game. (As of December 1st, there have been 6,056,282 cards traded, the fees for which have likely netted Valve around $250,000). And because each card can be traded an infinite number of times (digital cards don’t fade) the long tail of Artifact is very long indeed. Such transactions are very nearly pure profit for Valve, and the company has every reason to keep consumer pricing down to continue extracting rents. Valve isn’t lying when it announces, in defiance of Artifact’s rocky reception and crumbling player base, that they’re “in it for the long haul.”

But so what? It goes without saying that no one is forced to participate in Artifact’s financial metagames—Casual and Social Play will always be there, and the churning cycles of game news have largely moved on. And it’s no revelation to suggest that video games have economies, whether of the in-game (e.g. EVE Online’s famously treacherous free market) or out-of-game (e.g. the robust trade of Counter-Strike skins) variety. Even so, it’s commonplace for us to regard the latter kind of video game economy as something that a game has, as if it’s supplemental to what, on some ontological level, a game is. And yet Artifact is, in many ways, exactly the opposite: not a game with an economy, so much as an economy with a game. This means that Artifact’s real significance—what, I think, we ought to be “reviewing”—might not have anything to do with games at all. If we want to see Artifact’s real significance, we need to look elsewhere—or, really, everywhere.

Like most of Valve’s “innovations”, Artifact isn’t any less exploitative than its peers so much as it’s much better at hiding its exploitation, turning play into work and drawing it out over time, across the horizon where intention becomes infrastructure. In that, it has something in common with today’s tech dystopia as it attempts to capture more and more of human life. This is what platforms—from Uber to Facebook to Airbnb to Etsy—do, and Artifact is a game for the age of platforms. In their classic sense, platforms are infrastructures that connect two or more users for social economic exchange, and the ability for algorithms to manage this “labor” at scale with have made them the dominant economic form of our era. This, in essence, is Artifact, and to endorse it is, implicitly, to endorse the entire system it reflects.

Though we have largely welcomed platforms—because they let us speak farther than our voices can carry, because they have empowered us to be our own bosses, because they open others’ homes to us—the so-called techlash has made it increasingly obvious that we made a Faustian bargain that gave big technology companies more control over the world than ever before. Whatever new ways of being social platforms offer, they will always be on someone else’s terms, and for their profit. Convenience, as it turns out, was an invitation to even greater form of domination by capital.

“Whereas peasants and serfs were dispossessed of land and forced to urbanize and take up wage labor, internet users are being dispossessed of their time and attention, which are being enclosed through surveillance, then quantified and monetized,” writes Dr. Daniel Joseph, a political economist (and Waypoint contributor) at the University of Toronto, who wrote his dissertation on Steam. “Platforms are the means for this enclosure, mediating activity for the purposes of controlling all forms of exchange around it.”

Perhaps that all seems far from Artifact—a humble card game, after all, about wizards and shit—but it’s really not. Facebook dreams of capturing all that is social (which is to say everything), and while Artifact has a different target in mind, the underlying principle of enclosing more is the same. Artifact does to Magic what Facebook did to friendships because capturing metagames is not, in the end, all that different from capturing the social. Dr. Garfield was right that “games without metagames … don’t really exist.” Yet where he saw metagames as the essence of what made games matter, Valve saw an opportunity to turn Magic’s metagames—its proxy cards, its aftermarkets—into markets of its own, a transmutation only Steam was equipped to do. That, in the end, is Artifact in its iron heart: a machine for capturing metagames.