CS:GO, Dota Underlords, Half-Life:Alyx - these are the games Valve is most keen on to discuss during our visit, but they're not the only games being worked on across those nine floors. There's also Dota 2, the former heavyweight champion of Steam, recently knocked off its perch by the rise of CS:GO - but when you own both competitors and the platform they're duking it out on, you're probably not too bothered. Of greater concern is Artifact, a digital card game spinning out of Dota 2 that launched in November 2018 and, within a few months, had all but disappeared.
On paper, the game had a lot going for it. There was the Dota connection; the popularity of similar games such as Hearthstone and Gwent; the involvement of legendary card game designer, Mr Magic: The Gathering himself, Richard Garfield. And yet, less than a year-and-a-half from launch, the game's concurrent player count peaks at around 160. No, we're not missing a zero. For all the talk we've heard during our visit about numbers not mattering, Artifact has been a disaster for a studio that's used to pumping out hits - and Valve hasn't been shy to admit it.
"Artifact was an interesting failure in its first go-round," Valve CEO and president Gabe Newell tells us. "We were surprised. We thought that it was a really strong product." The question now, as Newell puts it, is: "What the hell. Where did we go wrong?" That's what the studio is currently trying to find out. Last March, after a few months of updates that failed to turn things around, Valve announced it was taking Artifact back behind closed doors. It hasn't been updated since while Valve, as Newell puts it, "does some soul-searching" and works on a revamped version.
"We ran an experiment, we got a negative result, and now we need to see if we've learned anything from that, so let's try again," he says. "And that's what [the Artifact team] have been doing and that's what they're getting ready to release. Based on the reaction to it, what was wrong with the product? How did we get there? Let's fix those things and take another run at it."
The problem is, with a lot of variables at play, it's hard to isolate the exact reasons Artifact failed. Perhaps it was the game's reputation for being overly complex — you don't play just one game of cards in Artifact but three, simultaneously, across Dota-style lanes. Or it could have been down to the business model: unlike the aforementioned Hearthstone and Gwent, it's not free-to-play, but players still had to buy extra cards, whether in random booster packs, through ticketed events or individually on the Steam Marketplace. Valve's love of free-market economics meant that single sought-after cards soon rose to prices higher than the game itself. (That's no longer the case - you can now pick up a full set for just over £30.)
"That was the biggest source of arguments: what went wrong?" Newell says. "You have a list of 50 different things, so let's say you change 20 of those things. What are you going to learn? Not much - you could have made both positive and negative changes to the design." Those initial updates were an attempt at more controlled experiments, but it became clear that wasn't going to be enough. "With Artifact, we have to do a larger reboot in order to justify its existence to customers and to markets," Newell says. This second go-around is referred to internally as Artifact 2, he says, though it's not clear yet (even to Valve itself, apparently) whether this will be presented as a full-blown sequel, an expansion, or just a big update after a long gap.
Valve isn't talking about what exactly this reinvention will involve, but given the cross-pollination at the company, a good bet is to look at what its other games are doing. Free-to-play seems likely, given what it's done for the fortunes of CS:GO. Following Gwent and Hearthstone onto mobile is possible, since the Underlords team have shown that Source 2 (the engine that also powers Artifact) can be squeezed onto the smaller screen. For a studio whose development process tends towards the scientific method, this would be a lot of variables to change at once - but it seems like that's the way Valve is leaning. "It's a lot easier to make small experiments than big experiments," Newell says. "But occasionally, you're in a situation where you have no choice, the experiment you're running has to be really big - and then you just hope you're right."
No comments:
Post a Comment