You're Halfway There: Inside VelociCode Playtest Night

by Alamo Tech Collective community, events, programming
VelociCode Playtest Night at ATC

On day 17 of 30. Was the exact point in the game jam where you’ve stared at your own project so long you can’t see it anymore. You know where every button is because you put it there. You know what the mechanic does because you wrote it. Which means you have no idea whether anyone else does.

On Friday, July 17th, about twenty people showed up to the Alamo Tech Collective hackerspace to find out. Nine games, laptops open, controllers passed hand to hand, everybody playing everybody else’s half-finished build.

What VelociCode is

VelociCode is a month-long, learning-first game jam run by ACM San Antonio, a 501(c)(3) nonprofit, in partnership with the Greater Gaming Society of San Antonio. Build a game. Any game, any medium, one month. Solo or teams up to four. All skill levels, beginners actively encouraged.

This year’s jam ran July 1st through July 30th, with the VelociCode Showcase on August 1st. Playtest Night sits deliberately in the middle, late enough that people have something playable, early enough that they can still do something about what they learn.

ATC hosted the space. ACM SA ran the night.

No rubric, no rotation, no script

VelociCode Playtest Night at ATC
The format was open floor. No sign-up sheet, no timed rotations, no feedback form. People wandered, sat down at whatever was open, played for a few minutes, and said what they thought.

That informality was by design, not an oversight. Structured playtests are great once you know what you’re testing for. Halfway through a jam, most builders don’t know yet; they need to find out what’s confusing before they can decide what’s worth measuring.

Why this works better than it should

Here’s what developers kept saying all night: handing the controller to someone else surfaced problems they had stopped being able to see. That’s not a soft observation. It’s the entire basis of small-sample usability testing. Jakob Nielsen and Thomas Landauer modeled how usability problems get discovered as you add testers, and the curve is steep and then flat. Roughly five testers surface about 85% of the usability problems in a given interface, with each additional tester after that turning up progressively less that’s new.

Nine games. Twenty people. Two hours. Every build in that room cleared the five-tester bar in a single evening, for the cost of some folding tables and a plate of dino nuggets.

The practical version: you don’t need a research budget. You need five people who have never seen your menu screen. What they trip over in the first thirty seconds is almost always the thing you’ve been blind to for two weeks.

Some of what surfaced was mechanical: a control scheme that made sense to exactly one person, an objective nobody could state back to you, a menu option that went unfound. That’s the point. Fixing those on day 17 costs an afternoon. Finding them on day 30 costs the submission.

VelociCode Playtest Night at ATC
The other half of the value is triage. Thirteen days is not enough time to act on everything twenty people tell you, and trying to will wreck a jam build faster than any bug. What a night like this actually gives you is a ranking: the problems that three separate strangers hit independently are real and structural, and the ones that came up once are probably taste. You leave with a shortlist instead of a wishlist. Several teams spent the back half of the evening doing exactly that, sitting with their notes, arguing about which two things were worth the remaining time, and letting the rest go.

The superlatives, which were not judging anything

At the end of the night ACM SA handed out five superlatives, and said plainly from the front of the room that they had nothing to do with placement or the final outcome of the jam:

  • Eye Candy! (best use of visual graphics) from ‘What Remains’
  • Who Needs a Designer? (most effective use of primitive shapes to get the point across) from ‘Quetzalcōātl’
  • Got That Dino In ‘Em (best use of theme) from ‘Petrified Peril’
  • Picture’s Worth a 1000 Words (best use of imagery) from ‘Lula_Vuelo’
  • Most Humane Use of a Dino (best use of dinosaur empathy) from ‘Domestic Dino’

Small call, worth noticing. A 2023 systematic review of game jam research covering 25 studies published between 2010 and 2022 found that the cognitive, social, and emotional gains people take away from jams hold up when the event is organized as a safe and inclusive space for learning. Awards that read as an early verdict would have worked against that. Awards that are obviously for funzzies don’t. Saying so out loud took ten seconds and kept the night what it was meant to be.

The room

VelociCode Playtest Night at ATC
DJ Chaku was on music, which regulars will recognize from the AI April Kickoff. Somebody brought dino nuggets, which did not survive the evening. The swag table carried ACM stickers, ATC stickers, and enough dinosaur decals to cover a laptop lid twice over.

Zaq’s safari outfit deserves its own paragraph and is not going to get one.

Thirteen days became one

Submissions closed July 30. The community showcase is Saturday, August 1; finished games, in public, with results that actually count this time. If you played a rough build on July 17 and want to see what thirteen more days did to it, that’s the payoff. Checkout VelociCode’s official website!

If you run a meetup, workshop, or jam in San Antonio and need somewhere to put it: our space is free for community tech events. VelociCode is what that looks like when it works.

👉 Tour the Space

See you at the next one.

Alamo Tech Collective

Building San Antonio’s tech community, one event at a time.

Resources & Further Reading