How to Prototype a Board Game in One Weekend

If you want to know how to prototype a board game, the answer is a pack of index cards, a Sharpie, a handful of pieces borrowed from other games, and one weekend you refuse to spend making it pretty. The goal is a first playable: something you can sit down with on Sunday afternoon, play through to the end, and learn from. Not a product. Not even a good game. A thing that works well enough to tell you what is wrong with it.

Most people who want to design a game never play their first prototype, because they never finish it. They design cards in a graphics program, or write a twelve-page rulebook, or wait until the idea is complete. The idea is never complete. The prototype is how you complete it.

What do you need to prototype a board game?

Almost nothing, and the less the better. A short shopping list:

• Index cards, 3x5, at least 200. Blank on one side. These are your cards, your tiles, your board spaces, and your rulebook.
• A Sharpie and a pencil. The Sharpie for anything you are fairly sure about, the pencil for everything else.
• Cubes, meeples, and dice pulled from games you already own. Nobody minds a Catan road standing in for a soldier.
• Coins or poker chips for money, points, or health.
• Sticky notes for rules you make up mid-game and need to remember.
• A big sheet of paper or the back of a poster for a board, if your game needs one. Hex or square grids can be drawn freehand.
• One or two patient friends for Sunday. Family counts.

Do not buy card stock, sleeves, or print-on-demand anything. You will throw away most of what you make this weekend.

Saturday morning: what is the core loop?

Before you write a single card, answer three questions in one sentence each. Write them on an index card and keep it in front of you all weekend.

1. What does a player do on their turn? Not the whole game. One turn. “Draw two cards, play one, move your piece.” If you cannot say it in a sentence, you do not have a game yet, you have a theme.
2. How does a player win? Most points, first to a goal, last one standing, the team beats the game. Pick one.
3. What is the interesting decision? Every turn should have a moment where the player wants two things and can only have one. In Ticket to Ride, it is draw cards or claim a route. In Azul, it is take the tiles you need and leave the ones that hurt you. Name yours.

This is the hardest part of the weekend and should take two hours at most. If you are stuck, borrow. Nearly every great game is a known mechanic with one twist. Agricola is worker placement plus feeding your family. Codenames is word association plus teams. Pick a mechanic you love, add one idea, and move on. For the mechanic side of this, the breakdown in Worker Placement, Explained Through One Game shows how one well-worn system can carry a whole design.

Saturday afternoon: how do you build the first playable?

Now make the pieces. Set a timer for three hours and do not stop to improve anything.

Make the minimum deck. If your game has cards, you need enough for one full play and no more. For a four-player game, forty to sixty cards is usually plenty. Write the name and the effect in Sharpie. No art, no flavor text, no borders.

Use templates, not uniqueness. Do not write sixty different cards. Write eight to twelve card types and make copies. “Gain 2 gold” times six. “Move one extra space” times four. You can add variety later. Right now you need a deck that functions.

Draw the board ugly. If there is a board, draw it in five minutes. Squares, arrows, a start and an end. Label spaces with words, not icons.

Write the rules on one index card. Setup, turn structure, how it ends, how you win. If it does not fit on one card front and back, your game is too complicated for a first prototype. Simplify until it fits.

Assign the borrowed pieces. Red cubes are wood. Blue cubes are water. Meeples are workers. Write the key on a sticky note.

At the end of the three hours, you should have a pile of index cards, a drawn board or none, a bag of borrowed pieces, and a one-card rulebook. That is a first playable. Stop building.

Building something? The newsletter sends one piece like this each week.

Saturday evening: can you play it alone?

Yes, and you should, before anyone else sees it. Play two or three hands yourself, taking every seat. This is not a real playtest. It is a smoke test, and it catches the problems that would waste your friends’ time tomorrow.

You are looking for:

• Dead ends. A player who cannot take any legal action on their turn.
• Runaway math. Someone who gets twenty points on turn three and can never be caught. If this shows up, read about catch-up mechanics in Designing the Comeback before Sunday.
• The obvious move. A card or action that is always correct. If there is one, the decision you named this morning is not a decision.
• Length. If a solo three-hander takes two hours, the real game will take four. Cut the deck in half or lower the winning score.

Fix what you find with the pencil. Cross out numbers. Tear a card in half if it is broken. Then go to bed.

Sunday: how do you run the first playtest?

Invite two or three people. Tell them it is a prototype, it is ugly, and it will probably break. Then follow these rules.

Teach it in five minutes. Goal, turn, exceptions. If the teach takes longer, note that; it is data about the design.

Play to the end. Even if it is clearly broken by turn four. The end state tells you whether the scoring works, whether the game knows when to stop, and whether anyone cared who won.

Say nothing during play. This is the hard part. Do not explain what a card was supposed to do. Do not suggest moves. Do not defend a rule. Sit on your hands and take notes.

Watch, do not ask. What people say after a playtest is polite. What they do during it is true. Who checked their phone? Who leaned in? When did the table go quiet, and was it the good quiet or the bored quiet? Where did someone misread a card, and was the misreading better than the rule? The full method is in How to Playtest Without Lying to Yourself.

Ask three questions at the end. “What was the most interesting turn?” “When were you bored?” “Would you play again?” Take the answers as leads, not verdicts.

Then play it a second time with the biggest fixes penciled in. The second play teaches more than the first, because now you can see which problems were learning and which were design.

What should you change first after the playtest?

You will leave Sunday with a list of fifteen things. Pick in this order.

Priority	Fix	Why first
1	Anything that stopped the game from ending	A game that does not finish cannot be evaluated
2	The dominant strategy	If one path always wins, nothing else matters
3	Dead turns	Turns with no real choice are the fastest way to lose a table
4	Length	Cut until the game ends before interest does
5	Confusing rules	Reword, then simplify, then cut
6	Theme and feel	Last, and only once the loop holds up

Resist the urge to add. Nearly every first prototype is too big. The right second version has fewer cards, fewer rules, and a lower winning score than the first. Add things back only when players ask for them.

What are the common mistakes in a first board game prototype?

• Starting with the story. Forty pages of lore and no clear turn. Theme is the easiest part to add later.
• Making it pretty too early. Index cards exist to make this impossible.
• Too many systems. Resources plus combat plus trading plus a tech tree. Pick one.
• Testing only with yourself. A solo test finds broken math; only other people find broken rules.
• Arguing with the playtest. If three people played it wrong, the rules are wrong.

FAQ

How do you prototype a board game cheaply?

Use index cards for all cards and tiles, a Sharpie for text, and pieces borrowed from games you own. A first playable prototype should cost under ten dollars and take one weekend.

How many cards should a first board game prototype have?

Enough for one full play with your target player count, usually forty to sixty cards built from eight to twelve repeated types. Add variety only after the core loop works.

Should a board game prototype have art?

No. Art slows iteration and makes you reluctant to change cards. Words in marker are enough until the rules have been stable through several playtests.

How long should a first playtest take?

Aim for under an hour, including the teach. If your first prototype runs longer, cut the deck or lower the winning condition before the next session. Most finished games end up shorter than their designers first planned.

Bring your prototype to a table

Every published game started as a stack of index cards in front of a few patient friends. If you have an idea, the only thing between you and a first playable is one weekend. The Ludarium newsletter sends one piece a week on design and playtesting, plus notice of prototype nights when the Indianapolis lounge opens. Subscribe at theludarium.com.

The game is the reason. The people are the point.