Interpretive Gameplay Studies

Josh Antolovic is a designer, artist, and researcher whose work centers on video games and natural systems.

Interpretive Gameplay Studies

2024-25

Building and piloting a game user research method that attaches an agenda to an existing sandbox game, so players can shape a game concept before the concept exists. Piloted with two participants over four sessions in Minecraft, and presented at the Hayes Graduate Research Forum.

Role: Solo (method design, recruitment, facilitation, analysis, and writing)

Themes: games user research, co-design, method development, qualitative analysis, environment-centered game design

At a glance:

A two-day protocol: day one is normal play, day two is the same world with an agenda attached, bracketed both days by informal conversation rather than a structured interview
Piloted with two participants across four sessions, one hour of play each, all four recorded to screen and webcam
The method's premise: playtesting arrives too late to inform an idea, so bring players in before there is an idea to test
The finding that justified the study: a participant spending her second hour putting materials back into the world, a behavior I had never seen in a Minecraft session
Presented at the Hayes Graduate Research Forum in February 2025
Primarily method development and qualitative analysis, not a game build

The Question

Games user research is young enough that it still borrows most of its methods from other fields. Methods built for game environments specifically, rather than adapted into them, seem more likely to support the making of new games. That is the gap this project works in.

My narrower problem was one of sequence. Playtesting puts a prototype in front of prospective users and watches what they do with it, which tests usability and whether a design intention survived contact with a player. It is good at that. But it happens after the idea is settled. In a traditional pipeline, of idea, pre-production, production, post-production, players do not appear until there is something to hand them, which means their in-game behavior can refine a concept but never inspire one. Their input arrives too late to change what the game is.

Co-design invites non-designers in for their expertise, on the assumption that everyone is creative, especially about the experiences they care about. Interpretive Gameplay is my attempt at a co-design method for games: instead of building a prototype and waiting, use a game that already exists, attach an agenda to it, and read what players do with it as design material.

Why Minecraft?

When choosing a game to run this in, I think along a spectrum from structured to unstructured. Structured games are sometimes called hallway simulators: predefined paths, limited customization, a shape the designer has already decided. Unstructured games carry many features and no predetermined route through them, often with procedurally generated worlds, so each player is somewhere nobody else has been.

Minecraft sits at the far unstructured end. A sandbox game can be defined as one containing many features with unstructured creativity rather than the maintenance of strictly designed goals, and that definition is what makes it useful here: I can apply structure to an unstructured game by supplying an agenda, then adjust the agenda to fit whatever I am researching. Its core affordances are simple enough to bend, compose and decompose, and enough people already know how to play it that I am not spending the first half hour teaching controls.

The method should also help earlier than a prototype can. In building a sandbox game, there is a stretch where you do not yet have enough features for sandbox-like play to be possible, so there is nothing meaningful to test. Interpretive Gameplay is meant for exactly that stretch.

The Protocol

Each participant played on two separate days, back to back, for about ninety minutes per session: fifteen minutes of conversation, an hour of play, fifteen minutes of debriefing. Both days ran on the same world, so day two opened onto whatever day one had left behind.

Day one is normal mode. The agenda is play as you would at home. Before play we trade Minecraft stories, which is doing real work: it is how I dismantle the feeling of a lab. When did you first play, which worlds do you remember, what frustrates you. During play I ask participants to speak aloud, and when they go quiet I ask what their goal is right now, or why they went that way. Afterward we walk back through their timeline of decisions together.

Day two is agenda mode, structured identically, themed throughout. The conversation beforehand is about the environment rather than about Minecraft: how conscious of it are you, what issues do you actually care about. Then the agenda, written down as well as spoken: play in a way that cares for the environment, however you interpret that is up to you.

Setting Up

Recruitment was a poster and then a conversation, sent to friends to see who was interested. Two separate sessions, one hour of play each, fifteen to thirty minutes of discussion, gratitude offered in cash or baked goods. Both people who said yes were experienced players.

I set up a desk at ACCAD with Minecraft installed, a webcam and mic, a second monitor, and OBS Studio capturing screen and face together. Both participants played in peaceful mode, which I chose deliberately: I wanted to see whether removing the threat of monsters would keep players out of Minecraft's default opening loop, the scramble for wood and coal to build shelter before dark. Coal for light, wood for tools. That loop is so reliable that I wanted it off the table.

Lab setup

Study recruitment poster

Participant 1

She spawned in a tree and immediately cut it down. From there her instinct was resources, exploring, getting sidetracked, and remarking how unique the world was. She built a house by the water and started a small chicken farm, which she had told me beforehand was her favorite thing to do in Minecraft, partly for the eggs to throw at friends. She died once by falling down a hole, which reminded her she needed a bed to move her spawn point.

Participant 1 building a chicken coop

Day two she responded about being semi-mindful of the environment rather than an activist about it. Reusable bags, attention to utilities when she is the one paying, and time on her grandmother's farm, which had no wifi. She was worried about land scarcity in Singapore, and small forests being turned into development.

Then she did the thing that made this method worth continuing. She went to her chest, pulled out the flowers, seeds, and saplings she had banked the day before, and started putting them back into the ground. She got selective about what she gathered and went back to fill in gaps she had left in the terrain. She moved her chicken farm underground and expanded it, and tried to spend everything from day one rather than accumulate more.

Participant 1 placing flowers around their house

I had never seen that in a Minecraft session. This is where interpretation runs both directions: she interprets the agenda and acts on it, and I interpret her atypical behavior as a game loop. A game where the player's job is putting material back into the world is a game I would not have arrived at on my own. Our debrief turned into design questions neither of us could answer, including whether chickens need light, and what a chicken utopia would look like.

Participant 1 placing blocks back into the environment

Participant 2

His first move on generating the world was to turn a full circle, spot the nearest tree, and sprint at it. Straight into the loop, peaceful mode or not. I had to remind him mid-session that nothing was going to attack him. He called peaceful mode baby mode. He decided he wanted a cozy cottage in a nice spot, found a small cliffside next to water, and spent a while unhappy with how ugly the surrounding dirt looked. By the end of day one he had already decided his own goal for day two: relaxed building.

Participant 2 starting a new world

He came in with more environmental history than participant 1, a Greta Thunberg climate march, a first game built around environmental ethics, and a game in development with environmental themes. On hearing the agenda, his first act was to pull saplings out of his inventory.

Most of his hour went into the cottage. He had been working from an online reference build, and partway through he abandoned it, having decided he trusted his own choices. Big windows, because he likes light. Stone flooring torn out and replaced with birch. Then a zen garden out of the flowers he had collected the day before, positioning the house for the best view of the water. What he was describing was not conservation. It was mental wellness and creativity, gardening as an exercise for the mind, care for an environment that turned out to mean the one immediately around him.

Participant 2 organizing flowers in their garden

Afterward he made the observation I keep returning to, which is that Minecraft's infinite world actively discourages sustainability. There is always more somewhere else. He also said the environmental constraint had made him more creative rather than less, which is the same thing participant 1's session suggested from a different direction.

Analysis

The analysis was layered rather than rigorous. I took informal notes live during the sessions. I generated transcripts of the recordings through OneDrive. I rewatched all four videos and took a second set of notes against the transcripts. I used Notion AI to tighten my own writing where the notes were too fragmentary to reread. Then I fed those notes to Claude to help work through the analysis and find what the two participants had in common.

Participant 2’s video transcript

Notion AI used writing organization

Claude anaylzing notes and transcript

What came of it:

Both were experienced enough that they went into resource-gathering mode immediately rather than looking at where they were.
Both put their house next to water for the view.
Both spent day one gathering and day two placing.
Both stayed inside their normal patterns of play while bending those patterns to meet the agenda, and both spent effort justifying their own resource use, out loud, unprompted.
Where they diverged was what care meant: one directed it outward at her chickens and the ground, the other inward at his own head.

Reviewing the Pilot

Giving players real freedom under one simple direction made them creative rather than lost. They designed their own play experience, and both of them produced mechanics I could take directly into my own work: putting material back, curating a space rather than extracting from it, cropping a loop down to its quiet part. The method does what I built it to do, which is generate ideas rather than evaluate them.

Next Ideas

The method is on hold rather than finished. I set it down to work on my thesis, which went a different direction. I expect to utilize it again and revise it for whatever a new project needs, in addition to documenting the revision as part of the method development..

Parked ideas:

An agenda change to create an ecosystem, plus one of two pre-play additions I have not chosen between: a short primer on natural ecosystems before play, or having participants sketch what they already know about ecosystems first. The second is more interesting to me because it produces an artifact I can compare against what they build.

Multiplayer, on a server, with participants collaborating toward one agenda.

Asking participants to draw a map of their world after each session, on both days, so I get their mental model alongside the recording.

For experienced players, dropping to a single longer session instead of two, since day two mostly spent day one's materials. Or keeping two days but generating a fresh world for the second, to see whether a new world breaks the gathering-then-placing pattern.

Having a participant continue a world that a previous participant already played. Their inherited terrain would be full of someone else's decisions, and possible ruin, which makes the previous player's play into level design.

Constraining the test game itself as a variable. Not just what the agenda asks for, but what the game is allowed to offer. What happens if they cannot go underground? What if the world is limited in size?

@joshantolovic

Index Next →