Playtesting spaces

A playtest is a session where someone plays your level while you watch whether it works. The Level Design Book defines it exactly that way, and calls it "a foundational skill and process in game design" (The Level Design Book: Playtesting).

For spaces, playtesting shows whether players find the way, read the climbable ledge, notice the landmark and feel the scale you intended. A drawing cannot show this. Playtesting is the only reliable check on wayfinding, affordances and flow. As The Level Design Book's layout chapter puts it, "the player never plays your drawing" (The Level Design Book: Layout), and its wayfinding chapter stresses that much of the psychology behind player guidance stays unverified until you test it (The Level Design Book: Wayfinding).

#Four kinds of playtest

The Level Design Book describes four (The Level Design Book: Playtesting):

Valve built its process around regular testing. The Level Design Book quotes a Valve developer: "Everything we do during the week is focused on that Friday playtest" (The Level Design Book: Playtesting).

Game design textbooks put testing at the same central place. Tracy Fullerton's Game Design Workshop, which its publisher describes as an "exercise-driven, non-technical approach to game design", gives playtesting its own chapter (chapter 9) between prototyping and balance (Fullerton 2024). Level design research is thinner here: Karlsson, Brusk and Engström's field study of level design at two studios hardly discusses playtesting as a practice, though one sound designer's wish sums up its aim: "it should be fun to play the level as soon as the level designers are done" (Karlsson et al. 2023).

#Running a playtest, step by step

#1. Plan

Decide the scope (which part of the level), the focus (wayfinding? combat? readability of the art pass?), what data you'll collect and how you'll take notes (The Level Design Book: Playtesting). Write two or three questions you want answered, such as "Do players find the side room?" or "Do they recognise the rope as climbable?"

#2. Brief

Tell the tester only what the player would know. The Level Design Book's advice on briefings: "But keep it brief" (The Level Design Book: Playtesting). Don't walk them through the level, and don't tell them where the secret is.

#3. Observe

"Most of the time should be spent watching the screen", ideally "from a short distance behind the playtester(s)" (The Level Design Book: Playtesting). The hardest part is staying quiet: "resist the urge to interrupt the playtest to apologize or explain" (The Level Design Book: Playtesting). Step in only if "the playtester is definitely stuck or if the level is clearly just broken" (The Level Design Book: Playtesting). For remote tests, ask testers to record their screen or stream privately (The Level Design Book: Playtesting).

For critique sessions, ask the tester to "try to say what you're thinking constantly" (The Level Design Book: Playtesting). Hearing "I think I'm supposed to go up there?" tells you exactly what your signifiers are saying.

#4. Record

Note what players do as well as what they say:

• Hesitations. Where do they stop, turn the camera around or backtrack? (The blockout page suggests flagging any pause longer than about 3 seconds.)
• Misreads. Where do they try to climb something decorative, or walk past something usable? PlaytestCloud describes testers walking "right past climbable objects" (PlaytestCloud blog: Playtesting and Navigation: Explaining the…).
• Times and deaths. The Level Design Book suggests tracking "basic stats such as play time, deaths, win percentages, etc.", and notes that "just attentive note-taking" is often enough (The Level Design Book: Playtesting).

Larger teams add telemetry, "automatically collected data sent to a stats server", and heatmaps, "top-down visualizations of where players frequently move, earn points, score kills, or die" (The Level Design Book: Playtesting).

#5. Debrief

Afterwards, "make it into a conversation" (The Level Design Book: Playtesting). Ask open questions: "What were you trying to do here?" "What did you think this building was?" Have a few prepared questions, but let the tester talk.

Accept suggestions gracefully. The Level Design Book notes that even when a suggested fix is weak, "the reason for the suggestion" and "the issue it highlights" are "almost always real and valid" (The Level Design Book: Playtesting).

#6. Decide and fix

Then "figure out what the heck happened and what you're going to do about it" (The Level Design Book: Playtesting). Group observations into problems, rank them by how many testers hit them, fix the top few, and test again.

#How many testers?

The Level Design Book doesn't give a number (The Level Design Book: Playtesting). Usability research offers a starting point. Jakob Nielsen argued in 2000 that "after the first study with five participants has found 85% of the usability problems", the best move is to fix them and test again; rather than one big study, "Spend this budget on 3 studies with 5 users each!" (Nielsen Norman Group: Why You Only Need to Test with 5 Users). He notes the limits: finding nearly all problems takes about 15 users, and the rule "only holds for comparable users who will be using the site in fairly similar ways" (Nielsen Norman Group: Why You Only Need to Test with 5 Users).

Games add variety (skill levels, genres, input devices), so treat five as a minimum per player type, not a universal answer.

#What to watch for in a space

#Playtesting the art pass

Playtesting also applies after the blockout stage, because an art pass can break what the blockout got right. The Level Design Book warns that "An art passed level that is 'busy' will be 'illegible' and difficult to read" (The Level Design Book: Environment Art). Re-test after every major art pass with fresh players, using the same questions as the blockout test, and compare. If players found the route in the grey version but not the dressed one, the art is the problem. See set dressing and art passes and level design and environment art.

#Further reading and watching