Workshop 1 — The Core Loop and the Spine of Game Design

Find the thing your player will repeat hundreds of times, and make sure it is still worth repeating. That is the whole job of this hour. The deliverable is a one-page Loop Canvas — handwritten is fine.

What a game is (the working definition)

Before hunting for the loop, know what you are building. Scott Rogers’ definition is short enough to use immediately.

1+ player + rules + a victory condition = a game

A ball bouncing against a wall is a time-waster. Add a rule (catch it with your left hand) and a victory condition (ten times in a row) and you have a game. To make it a video game, pass that output through the screen-and-platform filter.

Note what the definition does not contain: graphics, story, or an engine. If you cannot state your victory condition in one sentence, you do not have a game yet.

First principle

A game is what the player repeats — not what you want them to feel

Raph Koster’s argument in A Theory of Fun is that fun is the sensation the brain produces while it is absorbing a pattern. The moment the pattern is fully mastered, the fun stops and boredom begins.

Your players will see the ending once. They will press the same button several hundred times. If the repeated act is hollow, no amount of art, music or story will rescue it.

Where the loop actually lives

Schell’s elemental tetrad splits every game into four parts.

Element	What it is	Player notices it
Mechanics	Rules, goals, procedures — the loop lives here	Little
Aesthetics	How it looks, sounds and feels	Most
Story	The sequence of events that unfolds	Some
Technology	What makes all of it possible	Least

All four are essential and all four influence each other — but they differ sharply in visibility, and that gap is a trap. Teams spend their first month on the most visible element and the least testable one.

This hour we work only in mechanics, because that is where a core loop can be broken or fixed cheaply.

Anatomy of a core loop

Every loop that holds a player has all four steps, and cycles fast enough that boredom never catches up.

1. Act — the player does something that required a decision, not just a button press to advance.
2. Read — the game answers immediately and legibly, through image, sound, number or shake.
3. Gain — something is banked: an item, a score, a piece of knowledge, or the feeling of nearly winning.
4. Change — the next cycle is not identical, because the player is better or the options have shifted.

Step four is the one people forget, and it decides whether cycle twenty is still worth playing.

Play with the turn queue above — the rhythm of player and enemy alternating is the most concrete version of “one cycle of the loop”.

A loop is not flat — it has a shape over time

Schell plots audience interest as a curve. A session that flatlines is a session players leave.

• It starts with a hook — the opening spike is not the game. It buys you the attention you need to teach the loop.
• It rises in a zig-zag — interest climbs through alternating tension and rest. Constant intensity reads as flat.
• It ends above where it began — if the last cycle is less interesting than the first, the loop has stopped teaching.

Good loops nest inside each other

The short loop keeps hands busy, the medium loop keeps the mind busy, the long loop keeps players coming back.

Loop layer	Vampire Survivors	Stardew Valley	Slay the Spire
~30 seconds — the hands	Weave through the crowd, sweep up experience gems	Water, chop, mine — one tile at a time	Spend this turn’s energy on the right cards
~10 minutes — the decisions	Choose upgrades and combine the right weapon pairs	Spend one day of energy where it pays most	Survive a fight cheaply and pick the next card
~1 hour and beyond — the progress	Outlast the timer, unlock new characters and relics	Plan a whole season of crops and relationships	Build a deck that can actually beat the final act

A game starts with a verb, not a theme

If one verb describes what your player does, you have found the core mechanic.

Game	Verb	Everything else exists to…
Super Mario Bros.	JUMP	give the jump something to mean — enemies, pits, pipes
Portal	PLACE	ask one question in every puzzle: where do the two doors go?
Papers, Please	COMPARE	create tension from checking documents against rules that keep shifting
Our demo game	LISTEN	make sound the only basis for deciding to move or hold — the threat is never seen

The most common condition in the room

A very clear theme attached to a still-fuzzy verb. That is fixable, and fixable quickly: always write what the player does before what the player sees.

Innovation is iteration

People stall in this hour because their verb does not feel “new enough”. Rogers answers that with a single lineage:

Space Panic (1980) climbing ladders and digging holes → Donkey Kong (1981) added jumping and power-ups → Pitfall! (1982) added vine-swinging · Mario Bros. (1983) added skill-based combat → Super Mario Bros. (1985) perfected physics and level design → Mario 64 (1996) translated the mechanics into 3D.

Even the most innovative games have predecessors. The Rogers rule puts it bluntly: if you think something is innovative, it just means you have not been paying attention. Your job this hour is not to invent a verb nobody has used — it is to take a proven verb, put it in a new context, and sharpen it.

The theory of un-fun (subtractive design)

This is the idea that gives your pillars teeth.

You cannot add fun. You can only reveal it by chiselling away the un-fun.

Think of carving marble: the block is the initial idea, the statue is the core fun hiding inside, the rubble is the clunky UI, the unfair difficulty spike and the broken camera, and the chisel is objective playtesting plus the willingness to kill your darlings.

There is a trap inside this: fun degrades with exposure. After you have played the same level a hundred times during development, you lose all objectivity — you stop seeing bad mechanics because you have simply got used to them. This is why “is it fun yet?” is the worst question you can ask in early development. The usable questions are “is this loop still teaching?” and “which piece of un-fun have I not chiselled off?”

Two habits worth adopting

• Play bad games — you learn more by deconstructing a broken game and working out how you would fix it than by only playing masterpieces.
• Don’t be precious — never cling to a mechanic just because you created it. If it does not serve the game, throw it out.

The triangle of weirdness

While hunting for a verb, it is easy to make everything strange at once. Rogers points at three corners that need balancing.

Corner	The question
Characters	Who does the player control, and who do they meet?
World	Where does this happen?
Activities	What does the player do?

Only one corner may be profoundly weird. The other two have to ground the player in something recognisable. Make all three weird and there is nothing to hold on to, so they leave. Our demo game is weird in its activity (deciding by ear); the world (an Isan village) and the character (an ordinary person with a lantern) are deliberately ordinary.

Design pillars — how to say no on purpose

A pillar is not a slogan. It is a sentence that lets your team make a decision without re-arguing it.

• A good pillar rejects something. “Silence is the weapon” gives you the nerve to cut the entire soundtrack. “Great graphics” and “fun gameplay” reject nothing, because every game in the world wants both.
• A good pillar does not fit anyone else’s game. Paste your pillar onto the project of the person next to you. If it still sounds reasonable, it is too broad to be useful.
• Three is enough. For the demo game: (1) silence is the weapon, (2) players lose because they hurried, not because they lacked skill, (3) fear comes from what is never shown.

A finished example — the demo game’s canvas

Field	Content
Core verb (one word)	Listen
One cycle takes	About 20–30 seconds: listen, choose a route, move quietly, threat closes in, listen again
What the player decides every cycle	Push on to reach the thing they need, or hold still and stay safe. The stake is distance from the threat.
What changes next cycle	The threat learns the routes you have already used, and the lantern burns lower every turn
Three pillars	1. Silence is the weapon 2. Players lose because they hurried 3. Fear comes from what is never shown
One sentence, in the player’s voice	“Last time I moved too early. This time I know exactly where to wait.”

If you get stuck during this hour

No idea, or an idea that will not fit into any box. Use this instead of sitting quietly.

If you feel…	Apply this fix
Overwhelmed by infinite possibilities	Timebox and narrow the focus to microscopic elements — design one turn, not the whole game
Stagnant and sluggish	Force blood circulation. Walk away from the desk and let the subconscious connect things
Plagued by background anxiety	Complete the mundane distracting task first, then come back
Bogged down in drudge work	Skip to the good stuff briefly to regain momentum, then return to the schedule promptly

Sources of ideas that actually work: break your media diet (look at architecture, biology, history instead of replaying the same genre), analyse the bad, and dig into your real passions — Satoshi Tajiri turned a love of insect collecting into Pokémon.

Test on an outsider

Present the idea to someone with zero vested interest in games. If they cannot say what the player does, the problem is your verb, not them.

Hands-on — write the Loop Canvas for your own game (14 minutes)

One page, handwritten, no more.

1. Write your core verb. One word, and it may not be an adjective.
2. Draw the four steps of your loop and label how many seconds one cycle takes.
3. Answer what the player decides every cycle, and what is different next cycle.
4. Write three pillars. Each one must be capable of rejecting a feature.
5. Finish with one sentence, in your player’s voice, about why they want another round.

Three rules

• One page maximum — if it does not fit, it is not sharp enough yet
• No graphics, audio or story talk in this hour
• If you cannot answer a field, write “don’t know yet” and move on. Do not invent an answer.

This workshop’s worksheet

📄 Loop Canvas — blank template + worked example (PDF, 2 pages)

Page 1 is the template you fill in during this hour. Page 2 is the demo game’s version for comparison. Print one copy of page 1 per person.

Before you present — run these checks

Use this both on your own canvas and while listening to everyone else.

Signs of strength	Warning signs
You can explain the loop in 30 seconds without describing the story	The setting has to be explained before the play makes sense
You can name the decision the player makes every cycle	The verb is broad — explore, survive — and says almost nothing
Cycle twenty is clearly different from cycle one	The player repeats the same act and nothing shifts
Each pillar has already made you delete something	The pillars sound good but would suit anyone’s project

Workshop 1 wrap

1. A game is player + rules + victory condition. If you cannot state the victory condition in one sentence, you do not have a game yet.
2. Fun is the brain being rewarded for learning; a loop that stops teaching stops being fun.
3. The loop lives in mechanics, and mechanics are the cheapest element to fix.
4. You cannot add fun, only chisel away the un-fun — and a pillar that cannot reject a feature is not a pillar.

Next: we turn this one page into a document other people read the same way you wrote it → Workshop 2