Archaeogaming

Hello Games launched No Man’s Sky for North American Playstation 4 on August 9, 2016, gradually rolling out the game to the rest of the world and onto PC (via Steam) shortly thereafter. NMS contains an entire, literal universe for players to explore. That universe—all the planets, flora, fauna, structures, and artifacts—is completely procedurally generated, meaning that human-created algorithms and art combine to create a digital built environment that no one—not even the developers—have seen before. Based on the possibility of conducting archaeological investigation in a synthetic universe, I began the No Man’s Sky Archaeological Survey (NMSAS), and invited others to join me in a methodical exploration of new worlds and (we hoped) new civilizations. As an archaeologist, I was curious to see if we could apply existing archaeological methods to digital spaces. Can we do a real archaeological survey in a game? What would we find, and could we create typologies? Would there be seriation? Stratigraphy? Would there be an excavation mechanic (a rule in the game that would allow for digging)? No Man’s Sky seemed to be the best candidate at the time for looking for answers to these questions, in anticipation to other open worlds (or open universes) that might contain more robust algorithms that might create entirely new cultures from the ground up. Finding examples of what I call “machine-created culture” (MCC) has been my own personal windmill.

This is not the full report, but is instead a modest accounting of the first year of work in NMSAS as I presented at the EAA conference in Maastricht on September 2, 2017. I will finish the full report this year, submitting it to peer-reviewed journals for possible Open Access publication. You can download my slide deck.

Method

I spent a few months ahead of the game’s release thinking about how to best approach a survey of an unknown planet with potentially unknown cultures on it, and decided on a top-down approach guided by the published methods of the recent Eastern Korinthia Archaeological Survey (EKAS) and the Pyla-Koutsopetria  Archaeological Project (PKAP), and discussions with the leaders of those projects. We would start big, and gradually adjust to a more fine grain based on what we found.

The first step was to choose a system in the nearly infinite galactic map into which we could warp. There are four types of stars, each with variations in the nature of the worlds orbiting them. We were curious if the types of systems affected what we would find in an archaeological context. Following the warp, we would select a planet and then conduct a series of orbital surveys, much like what “space archaeologist” Sarah Parcak does with reviewing satellite imagery of culturally sensitive areas on Earth. Following the orbital survey, we would drop our spacecraft down into the landscape to conduct a series of low-altitude transects to identify and count features, both natural and not. Then we would conduct a series of field-walking surveys in sites or over swaths of territory to look for settlement patterns, construction, waste, etc. We would also identify sites and features, marking them for a second team to return for thorough documentation and possible excavation. For each level of the survey, we would take screen grabs and video capture to tie to custom software context sheets coded by the FAIMS project at Macquarie University (see below). At least that was the intent.

In-Game Problems at Launch (with game mechanics)

The archaeological sites we find are not necessarily the ones we want, and we must do our best to document the methods and results of our investigations. The game that we (and others) had hoped for had not been realized, and it was nearly impossible to do anything archaeologically meaningful in the initial versions of NMS:

• There was no Cartesian coordinate system (or any other kind of coordinate system) to allow us to pin a thing or place to a map.
• There was no way to determine which way is north.
• There was no way to dig/excavate or set markers or lay out a grid.
• There was no way to communicate with or see other players even if they were on the same planet.
• High-altitude orbits yielded nothing of note on the surface below, as what was found on the surface was not reflected from what we saw from orbit.
• Every two weeks, all discoveries were wiped from the game, preventing any kind of backtracking or revisiting of sites of interest.

What did We Find?

Despite the significant obstacles to NMSAS’s goals, several team members persisted in those first few months, collecting mostly screenshots of what they found, which was enough to start creating typologies, although both relative and absolute chronologies remain meaningless in the game. We did find multiple examples of procedurally generated buildings, monoliths, and ruins of various types, that could be lumped into a handful of general types, and then further organized by subtle differences in features. We frequently found waste in the form of “loot” crates, which contained natural resources, battery packs for our spacesuits, materials for ship repair, and also artifacts relating to the three different races found throughout the universe. These artifacts could be collected and sold for relatively large sums of money. In later versions of NMS, these could be repatriated to non-human species (or left at monoliths) for rewards of reputation or the location of a planetary portal. Despite the wealth of worlds to explore, there was a lack of diversity in what was discovered, and I found myself in the trap of getting bored by the “coarseware” of these worlds, distracted by the occasional shiny object. I was a 19th-century archaeologist trapped inside a 21st-century game.

Things we Got Out of NMSAS:

Over the first year of play, the NMSAS team saw a 99% attrition rate, and it appeared at first review as if the project had been a failure. We weren’t able to accomplish the goals in the digital universe, and what we found was largely contextually meaningless. We did, however, find some benefits that can now be applied to future projects in digital built environments (i.e., synthetic worlds):

1. Prior to launching NMSAS, Catherine Flick (DeMontfort University) and L. Meghan Dennis (University of York) recommended that the team and project follow a code of ethics on how to conduct ourselves and our work within the confines of an unknown universe among unknown “life forms.” Our agency in the game might affect future investigation. Once the code of ethics was written by the three of us, we posted it publicly online and also emailed it to the other team members.
2. When NMSAS was announced, I was approached by Brian Ballsun-Stanton and Georgia Burnett of the Field Acquired Information Management Systems (FAIMS) project at Macquarie University. They had created an Android app that would allow for data collection in the field via smartphone and tablet. I provided the workflow and data definitions, and FAIMS provided bespoke digital context sheets for the team to use, that would update a central database in real-time. Because NMS lacked any kind of coordinates system or way to measure distance, we were unable to make much use of the software (which was quite good and easy to use). With the release of v1.3 of NMS, we should be able to renew efforts to use the software to document what we find on the survey. The software could also be changed to accommodate data from other synthetic worlds that are explored in a similar fashion to NMSAS.
3. It is likely a mixed blessing that NMS was released without expected functionality as the team was able to test and modify its methods on the ground. Working in a “failed” environment also allowed me to streamline processes and reconsider when to create and launch digital field projects. We could also determine what tools we were lacking in the synthetic universe, and could observe and adopt those that were created as mods in the player community.
4. Stepping outside of NMS I was able to create an experimental Harris-style software matrix to track changes in the game between versions, treating each version as a stratum in order to read game stratigraphy. The methods of creating a software matrix can be applied to any digital built environment, game or not.

Lessons learned:

After one year of playing NMS (and actually trying to conduct an archaeological investigation within that space), there are a handful of lessons to take away about managing a team-based digital project set in a synthetic world:

1. Video game communities for games such as NMS contain thousands of citizen scientists and offer access to thousands and thousands of wiki pages, reddit threads, etc. Archaeologists must take advantage of this, involve the community in large projects, and treat this as a kind of public archaeology. The community will collect much more data than a team of archaeologists would. They will also create more mods more quickly in order to facilitate the navigation and use of a digital built environment.
2. Archaeology is social, and surveys are social events. In the natural world, no one works alone during a field season. It is a team effort. This should also be the case with digital projects. Working alone is no fun, is often thankless, and leads to abandonment of a project. Other reasons for attrition include professional and personal obligations, as well as leaving something boring for something new. No clear end goals or timeline also contribute to poor morale.
3. Learn the landscape (play the game) before launching a project plan/project. No archaeologist ever showed up while a new culture was beginning. Granted with games we work on an accelerated timeline, but we should take time to play a game in order to fully understand its mechanics and what is possible to do within that environment, developing an archaeological project plan around that instead of discovering everything we cannot do in that space.
4. Manage expectations. Don’t over-hype and over-commit. NMS failed in the beginning because too much hype led to unrealistic (and unattainable) expectations by players and the media. Later patches rolled out quietly, and over the past year have made the game quite close to what it was promised to deliver in 2016. Managing team expectations is also key to project success, letting members know what is expected of them, and what the goals are, both in the short- and long-term.
5. Publish your successes/failures (and data, media, etc.) as Open Access so that others can easily discover and use your work to advance their own. Blog/tweet regularly about the project’s progress, and encourage archaeology “ride-alongs” via streaming services such as Twitch. Not doing more of this was a serious failure of NMSAS.

The (Unexpected) Future of NMSAS:

I was ready to end the NMSAS project after delivering a post-mortem at the 2017 EAA meeting in Maastrict, but Hello Games released v1.3 (“Atlas Rising), which contained much of what players (and the team) had been expecting at the game’s initial release:

• Portals exist for fast travel between worlds across the universe by use of a combination of 16 glyphs (which take forever to find).
• We can excavate trenches with the new Terrain Manipulator function for the mining tool.
• Up to 16 people can group up for synchronous play in the same world, which is crucial for field-walking with a team.
• We now can see where north is, and can see units as well as time to calculate distances across the landscape (and across space).

There are also three major archaeological things to do now that v1.3 is out:

1. With the Terrain Manipulator, all players can create their own objects and art, which can be discovered by others. This follows on the Pathfinder update earlier in 2017, which allowed for base-building (and sharing), and the ability to build beacons and communication terminals that could be discovered by other players. NMSAS can now document these player-made artifacts throughout the universe.
2. With the advent of cheap and largely accessible 3D printing, players (and archaeologists) can print what they find. Thanks also to Open Source tools, it is easier than ever to do photogrammetry in-world, and to export to VR, granting access to spaces to people who do not play or own instances of synthetic worlds.
3. The most exciting new project stems from the fact that v1.3 was for some players a catastrophic event leading to a mass migration from one space to another. The “Galactic Hub” was a planet shared by 5,000+ players. They built bases, shared gardens, and created a Utopian society of scientists who used that world to explore the surrounding systems and record what they found on a communal wiki. When v1.3 launched, it changed the nature of all of the planets in the universe, and turned the Galactic Hub into an uninhabitable wasteland of ice. The population is migrating to a new world, leaving behind the “Legacy Hub,” on which are remains of that human society’s material culture. This demands archaeological attention, and I am collecting glyphs as fast as I can so that I can conduct a preliminary survey prior to beginning the Legacy Hub Archaeological Project (L-HAP). I hope you will join me.

Thank you!

—Andrew Reinhard, Archaeogaming