The Experimental Gameplay Workshop: the GDC session that showed Braid, Portal and Journey first

Key takeaways

Most conference sessions are about finished work. Someone ships a game, the game succeeds, and a year later they explain how they did it. The Experimental Gameplay Workshop inverted that. Developers bring things that are not finished, often not even playable in a normal sense, and the room decides whether the idea holds up.

That format has produced an unusual track record. This site sits on the domain that hosted the workshop’s public pages through the 2000s, which is why the archive at the workshop record is still here.

What the session actually is

A single time slot at GDC. Several developers, a few minutes each, a projector and a room that fills past capacity. Presenters show a prototype or a mechanic, explain what they were trying to find out, and sit down. There is no pitch, no publisher, no demo booth.

GDC’s own description of the session is that it highlights games taking approaches to interactivity that have not been tried before. That is a narrower brief than “indie showcase”, and it is the reason the line-ups read strangely in hindsight: some of the games became famous, and some were never released at all.

Game Developers Conference show floor in 2007

The track record

When GDC marked the workshop’s twentieth year in 2022, the games it pointed to as having appeared there included Portal, Journey, Untitled Goose Game, The Witness, Braid, Baba Is You and What the Golf? That is a strong list for a session that has never had a budget, a stage design or a headline sponsor.

The individual stories are more interesting than the list. Braid was shown in the 2006 session alongside flOw and a freestyle mode prototype from the Guitar Hero team, which is covered in detail in the piece on the 2006 line-up. Later sessions produced the same pattern at a different scale, and a single bill more than a decade later contained three games that went on to define their corners of the market.

Outside the Game Developers Conference venue in 2004

Why unfinished work is the point

A finished game answers a question. A prototype asks one, and asking in public is uncomfortable in a way the industry usually avoids. The workshop’s value is that it gives developers a room where showing something broken is the expected behaviour rather than a risk.

Our own reading is that the session survives because it costs nothing to be wrong there. A publisher meeting punishes an unproven mechanic. A conference room full of designers rewards it, because the audience is made of people who have their own unfinished projects and know what they are looking at.

That is also why the format has never needed to grow. A bigger session with more presenters and a longer slot would attract people pitching rather than people asking, and the thing that makes it work would quietly disappear.

Independent Games Festival expo floor at GDC 2013

What happened to the games

Three outcomes recur. A small number become commercial landmarks – the ones GDC lists. A larger number ship as smaller releases that never make a headline but keep their creators working. And a real share never ship at all, because the prototype answered its question and the answer was no.

The archive on this domain is a record of all three. The session line-ups preserved at the workshop archive include games you can buy today and games that exist only as a paragraph in a write-up from the year they were shown.

Why this site keeps the record

experimental-gameplay.org carried the workshop’s site through its early years. When the domain changed hands, the choice was to let the old pages return errors or to keep the record and point the old URLs at it. We kept it, and the workshop pages remain the most linked part of this site.

There is also a thematic reason. This is now a site about a game that develops in public, with unfinished features shipped behind toggles and kept or dropped based on what players report. That is the same bargain the workshop asked of its audience, at a different scale.

GDC’s own anniversary write-up is at gdconf.com. The rest of this series is on the EXG blog.

How a line-up gets chosen

The session runs on submissions. The organisers publish a call for proposals months ahead, developers send in what they have, and a small group selects the handful that will fit into the hour. The calls for 2014, 2018 and 2019 are preserved in the archive on this domain, and they read the same way each year: send us something unusual, it does not have to be finished, it does not have to be a commercial product.

That curation step is what separates the session from an open showcase, and it is the part that cannot be automated. Someone has to decide that a maze game about Windows hotkeys belongs on the same bill as a rhythm game about dread.

The name that causes confusion

Two things share this name. The Experimental Gameplay Workshop is the GDC session. The Experimental Gameplay Project was a student research project at Carnegie Mellon’s Entertainment Technology Center in 2005, which produced around fifty prototypes under a seven-day rule and gave the world Tower of Goo.

They are separate efforts from the same period, and the confusion is permanent because both are usually shortened to “experimental gameplay”. The Carnegie Mellon side is covered in its own piece on the seven-day prototyping rules.

What the format asks of an audience

Watching unfinished work is a skill. A prototype with programmer art and no sound is easy to dismiss, and the whole value of the session depends on a room that does not do that. The audience has to look past execution to the question underneath, which is exactly what a publisher, a storefront and a review score are all structurally bad at.

That is the resource the workshop actually provides. Not exposure, not funding – three hundred people willing to judge an idea on its own terms for four minutes.

Images: Experimental Gameplay Sessions and Game Developers Conference photography via Wikimedia Commons (CC BY / CC BY-SA).