Landmark is the ambitious sandbox aiming to map out the future of player creativity



Publisher: Sony Online Entertainment Developer: In-house Format: PC Origin: US Release: TBA

When EverQuest Next was announced in August 2013, there was some confusion about why a second game had been announced alongside it – particularly one that looked so much like its MMOG stablemate. EverQuest Next Landmark was pitched as some combination of a block-building adventure game, player content creation tool, and an MMOG in its own right. It was to be an opportunity for players to pitch in on the development of EverQuest Next, but its potential uses stretched far beyond its sister game.

Director of development Dave Georgeson is transparent about the fact that the name was confusing. Sony Online Entertainment chose to append EverQuest Next to Landmark’s title when it became worried that its fans would believe it had delayed development of its next MMOG to work on an unrelated Minecraft-alike. Developing Landmark’s underlying technology was essential for the development of EverQuest Next, and reflecting that in the title was meant to be a way to make the studio’s intentions clear.

Landmark is not simply an EverQuest game, however, as its alpha release and rechristening show. Its voxel-based construction system is vastly more complex than the block-based systems it resembles at a glance. Your tools can be blocks, wedges, or spheres; you can resize, twist and manipulate materials before placing them; and you can use similar shapes and principles to delete matter you’ve placed. After that, smoothing and texture-painting tools can be used to add further detail.

The game’s natural environments and the bumpy surfaces that can be achieved quickly using dirt, wood and stone make the system ideal for creating medieval-looking buildings. Indeed, the first thing you’re likely to make will be a shack or castle tower. Landmark isn’t intended to be a fantasy game, however: SOE wants players to branch out beyond the thematic remit of the EverQuest name.

This Viking warship is assembled out of pieces of prop furniture. It is similar, in many ways, to the sculptures players built using Star Wars Galaxies’ housing system.

A Landmark server is comprised of 50 procedurally generated islands, each roughly two square miles in size. When you begin playing, you’re able to roam the entirety of the world in search of materials, but your building rights will be initially limited to a single patch of land. Despite this restriction, players in the alpha have already managed feats of architectural design that push at the limits of what the designers intended. In some cases, uses have been found for the game’s voxel-shaping tools that exploit their programming to create new effects.

This might mean manipulating a quirk in the way that voxels automatically fill in negative space to create a massless ‘zero voxel’, a moveable absence that can be used to manipulate the shape of nearby voxels to create new patterns. Rather than close these loopholes, SOE is looking for ways to bring them more intuitively into the toolset.

“I work all day, and then I go home and play the game,” Georgeson says, “learning what the players have done that day, what tricks they’ve discovered, so that I can go back to the team and make [Landmark] a better and better whole.”

A few months of closed alpha time (the game is now in closed beta) furnished the Landmark team with more mechanical feedback than it can hope to stay on top of. In addition, SOE has committed to being as open as it can about its processes, and when new features will be delivered. The only exception is EverQuest Next itself, which is being made “behind the curtain” to avoid spoilers. Yet the community still has significant scope to impact the development of both games.

“There’s a hundred thousand of them and there’s a hundred of us,” Georgeson says. “There’s so many minds percolating the game, and that’s what we’re excited about. By opening up, we have all these other minds second-guessing what we’re doing. That’ll always end up resulting in a better product.”

It’s a straightforward rejection of the notion that the best games are made by small teams or auteurs. Landmark’s designers see themselves as much as toolmakers as content creators in their own right – the team’s end goal isn’t to realise a long-held vision, but to create a platform that other people will be able to take and run with. This is an assertive acknowledgement of the powers that are on the rise in PC gaming at the moment, from Minecraft through to the Steam Workshop and SOE’s own Player Studio.

This train’s smooth edges and detailed wheels surprised Landmark’s programmers, who didn’t believe it would be possible. The only prefab part is a forge, used to create the smoke from the chimney.

“When we recruited for this team, we recruited essentially egoless people,” says Georgeson. “Creative, brilliant people who weren’t wrapped up in the notion that it had to be their idea. What we want is a great entertainment experience. How we get there doesn’t matter at all.”

The current version of Landmark is a convincing proof of concept, demonstrating that SOE’s ambitious voxel-shaping tools not only work, but can already be used to create remarkable things. The next step is to implement the game’s survival features – combat and more substantial resource gathering, which will provide players with a reason to explore beyond discovering what their neighbours have been building.

NPCs within Landmark will operate according to an ‘analogue’ AI system that SOE is developing in conjunction with London-based AI consultancy Storybricks. Rather than scripting the actions of monsters or NPCs based on the presence of the player, the Storybricks-inspired system works by ‘tagging’ environmental objects as parts of a symbolic taxonomy that the AI understands. Areas might be safe or dangerous, rich or poor, and so on, and the behaviour of computer-controlled creatures will be moderated accordingly by a set of built-in preferences and motivation. In EverQuest Next, these systems will be designed by SOE. In Landmark, meanwhile, players will – within the confines of their claims – be able to experiment with the AI to suit their own purposes. This might mean establishing a town with merchants, or a dungeon with monsters to defeat.

Players will eventually be able to package up all or parts of their builds and sell them to each another for real money via Player Studio. The AI system expands the implications of this idea: in addition to trading castle pieces and sculptures, players will be able to trade game design ideas, too. It’s a wildly ambitious goal, aiming to fall somewhere between Minecraft, LittleBigPlanet and Second Life. Yet SOE’s plan to pull this off is practical and grounded in a creative resource – players –  that has the potential to yield results many times in excess of what a developer can achieve on its own. Landmark not only deserves to be thought of as it its own game, but may end up being much more important than the MMOG that it presages.