Playtesting
Watch real people play without helping, collect what they do more than what they say, and turn it into changes.
THE GOLDEN RULE
Don't explain, don't help
Hand testers only what a real player would get: the box, the rules sheet, the first screen. Then stay quiet. When they get stuck, don't jump in. Being stuck is the finding. Every time you feel the urge to explain, write it down instead: that's a place where the game failed to explain itself.
A tester stares at the red cards for twenty seconds, then ignores them all game. Don't say "those are the traps". Write: "red cards unclear, ignored for the whole game".
Check yourself
During a playtest, Sina can't work out how to start his turn and looks at you for help. What do you do?
- Explain the rule quickly so the session doesn't stall
- Stay quiet and note where he got stuck and what he tried
- Play his turn for him to show how it works
- Stop the test; the game clearly isn't ready
Show the answer
Stay quiet and note where he got stuck and what he tried
Right. His confusion is exactly what you came to find. Note where and why. If he's completely stuck for a long time, you can help at the end and mark it as a failure of the rules sheet.
Run a playtest session
- Hand it over with only the real materials
The rules sheet, the cards, the board. No speech about what the game is "really about".
- Ask them to think aloud
"Please say what you're thinking as you play, like 'I think I should go left'." You hear their reasoning, not just their moves.
- Watch faces, pauses and hands
Note where they lean in, where they hesitate, where they get bored and start checking their phone.
- Ask a few open questions at the end
"What was the goal?", "What was the most fun moment? The least?", "Where were you confused?", "Would you play again?"
Record the problem, not their fix
What testers are great at
Finding problems. When a player says "I got lost" or "that part dragged", believe them. They felt it, and other players will too.
What testers are often wrong about
The fix. "Add a map" is one guess at a solution. Write down the problem underneath ("got lost after the second room"), consider their idea, then design your own fix: maybe a landmark, maybe a shorter level.
Check yourself
After a session, a tester says: "You should add a map." What goes in your notes?
- "Add a map", as a task for tomorrow
- Nothing; one person's opinion doesn't count
- The problem underneath: they got lost, and where
- "Tester didn't understand the game"
Show the answer
The problem underneath: they got lost, and where
Yes. The real finding is that they got lost. A map is one possible fix; a clearer path or a landmark might be better. Record the problem, then design the fix yourself.
Open questions, not leading ones
Leading
"Wasn't the boss fight exciting?" "The jump feels good, right?" The question contains the answer you hope for, and polite people give it to you.
Open
"What was the most exciting moment?" "How did jumping feel?" The tester has to pick, and if they don't mention your boss fight, that tells you something too.
Check yourself
Leading question or open question?
- Wasn't the boss fight exciting?
- What was the most exciting moment?
- You liked the new cards, didn't you?
- Where were you confused?
- The ending was clear enough, right?
- What do you think the goal was?
Show the answer
Leading: Wasn't the boss fight exciting?, You liked the new cards, didn't you?, The ending was clear enough, right?
Open: What was the most exciting moment?, Where were you confused?, What do you think the goal was?
PLAYTEST LOG
Date: 12 May
Version: v3
Tester: Leila (plays board games monthly)
What happened:
- Ignored red cards all game
- Leaned in on the last 3 turns
- Asked twice "is it my turn?"One change to try in v4:
Print "TRAP" and a skull on every red card.Date, version, tester, what they did, one change. Five minutes of writing per session, and a month later you can still see why v4 is the way it is.
Check yourself
Following Nielsen's rule of thumb, what's a good plan for testing your paper game?
- One big session with 30 people, then launch
- About five testers, fix what you found, then about five new testers
- The same two friends every week, so they get good at it
- Wait until the game is finished, then test once
Show the answer
About five testers, fix what you found, then about five new testers
Yes. Around five people surface most of the big problems. Fixing them and testing with fresh people finds the problems the first fixes revealed or hid.
Check yourself
It's best to use the same testers every round, because they already know the rules and can give deeper feedback.
Show the answer
False
False. Repeat testers have learned the game, so they no longer trip over what confuses a newcomer. A returning tester can help with long-term strategy, but each round needs fresh people to see the game as a new player does.
Lesson recap
- You can't judge your own game; test early, even a scrappy version.
- Don't explain and don't help. Every urge to explain is a note: the game failed to explain itself there.
- Ask testers to think aloud, and watch faces, pauses and where they get stuck or bored.
- Testers are great at finding problems and often wrong about fixes: record the problem, design your own fix.
- Ask open questions, not leading ones. Test with about five people, fix, then test with new ones.
- Notebook prompt: start a playtest log with date, version, tester, what happened and one change to try.