A strong level often feels simple when you play it.
You enter. You see something. You try it. You understand it. The game changes the context and add one more layer. You use the idea again and again, you discover different ways of using it, a new thing calls your attention. Then, near the end, the level asks one final question, a final test.
But underneath that natural feeling, there is structure.
One of the clearest structures for this is Kishōtenketsu.
Kishōtenketsu is a four-part structure that comes from East Asian poetry and Japanese narrative tradition, but in game design we read it in a very practical way: a structure for teaching through play.
It is a way to introduce one idea, let the player understand it, surprise them with it, and then give them a satisfying final test.
Ki = Introduction
Shō = Development
Ten = Twist
Ketsu = Conclusion
Think of it less like a story formula and more like a lesson that feels like an adventure.
The important point is that the level is not just a corridor of obstacles. It is a conversation about one playable idea.
This is a close reading, not a formula
I do not use Kishōtenketsu as a rigid recipe.
The danger of any structure is that designers start copying the skeleton instead of understanding the function.
A level does not become good because it has four parts, it's obvious. It becomes good when each part changes the player’s relationship with the central idea.
So the question is not:
Does this level have Introduction, Development, Twist, and Conclusion?
The better question is:
What does the player understand differently after each beat?
That is where Kishōtenketsu becomes useful.
In level design, structure is not there to make the level look organized. Structure is there to change the player’s understanding in a controlled way.
Ki: Introduction
The introduction presents the idea in its simplest readable form.
This is where the level says:
Here is the toy. Touch it. See what it does.
The player should meet the mechanic, object, enemy, hazard, or spatial rule without too much noise around it.
This is not the moment to test mastery.
It is the moment to create the first mental model.
If the idea is a moving platform, the player should understand how it moves.
If the idea is a wind enemy, the player should understand that wind changes position.
If the idea is a lava bubble, the player should understand the rhythm of eruption.
If the idea is a new jump interaction, the player should understand the body of the jump before the level asks for precision.
The introduction should answer:
What is this?
What does it do?
Can I predict it?
Can I try it without paying too much cost?
A good introduction is generous. It gives the player enough safety to observe and enough clarity to act.
It teaches through play, not through explanation first.
Shō: Development
Development repeats the idea, but it adds consequence.
The level says:
Good. Now let’s see more of what this idea can do.
The player has already seen the basic behavior. Now the level asks them to use it with more intention.
The platform is no longer isolated.
The wind now appears near an edge.
The lava bubble crosses the path.
The enemy pattern now intersects with movement.
The jump interaction now requires timing.
The important rule is simple: do not add five new ideas at once.
Develop the one idea you already introduced.
Development is not simply more difficulty. It is the same idea made more specific.
The player should feel:
I know this, but now I must use it properly.
This is where the level starts turning comprehension into skill.
A bad development section only adds clutter.
A good development section adds a reason to care.
Ten: Twist
The twist is where the level changes the relationship.
The level says:
You thought you understood it. Now look at it from another angle.
The object that was a hazard becomes useful.
The enemy that pressured the player becomes a tool.
The safe path becomes the risky path.
The timing pattern is inverted.
The moving platform now moves vertically instead of horizontally.
The mechanic appears in a different spatial orientation.
The player must combine the idea with another idea.
The key is that the twist should not feel random.
It should feel like a meaningful mutation of the thing already learned.
The player should think:
Oh, it can do that too.
That is the important flavor: surprise that expands understanding, not surprise that breaks trust.
A good twist expands the player’s mental model. It does not erase it.
If the twist requires a completely different mental model, it may feel like a new lesson instead of a transformation.
Ketsu: Conclusion
The conclusion lets the player prove understanding.
The level says:
Now show me you understand.
This is the payoff.
It does not need to be the hardest possible version of the idea, but it should feel like the cleanest final statement.
The level has already said:
You have seen the idea.
You have practiced the idea.
You have seen the idea transformed.
Now use it with confidence.
The conclusion can be a final challenge, a clean execution, a reward path, a small mastery test, or even a calm exit that lets the player enjoy what they now know.
The important thing is that the level should close the thought.
Without conclusion, a level can feel like it simply stops.
With conclusion, the level feels like it resolved.
The fifth beat: safe validation reward
I often add one more practical beat after conclusion:
Safe validation reward.
After the player understands the idea, give them a small place to use it again without heavy pressure.
This can be a reward space, a coin path, an optional shortcut, a small playground, a camera moment, or a safe repetition.
Why does this matter?
Because learning needs emotional confirmation.
The player should not only survive the lesson. They should feel that the lesson now belongs to them.
A reward after mastery is not just payment. It is recognition.
The game says:
Yes, you learned it.
That feeling is powerful.
The simple formula
Here is the compact version:
Ki = introduce the idea safely
Shō = develop the idea through variation
Ten = twist the idea into a new meaning
Ketsu = conclude with mastery/payoff
Or, in more practical game design language:
Show it.
Let them play with it.
Turn it sideways.
Make them feel clever.
My favorite way to remember it is this:
Kishōtenketsu is not:
easy -> medium -> hard -> hardest
It is:
understand -> explore -> rethink -> master
That difference matters.
Difficulty can rise, but rising difficulty is not the point.
The real goal is player understanding with emotional rhythm.
Example: a Bullet Bill level
Imagine a level built around Bullet Bills.
Ki: Introduction
A Bullet Bill flies across an open area.
The player learns:
Trajectory, possible curves, distance, velocity, ... The level provide necessary space for this reading, player thinks: I can dodge it.
The space is safe enough to observe. The rule is visible.
Shō: Development
More Bullet Bills appear with different timing.
The player learns spacing and rhythm.
The idea is the same, but now the player must act with more intention.
Ten: Twist
A Bullet Bill breaks a wall.
Now the danger can also be used as a tool.
The player’s relationship changes.
It is no longer only: avoid the Bullet Bill.
It becomes: bait the Bullet Bill, use the Bullet Bill, redirect the threat.
Ketsu: Conclusion
The player must dodge, bait, and redirect Bullet Bills to open the final path. The mental models is updated, explosion becomes a possibility.
Now the enemy is not just an obstacle.
It became a toy, a lesson, a tool, and finally a mastery test.
That is the power of this structure: a simple readable system can create deep play when the level changes the player’s relationship with it beat by beat.
Why this works so well
Kishōtenketsu works because it respects the player’s mind.
It does not lecture.
It does not overload.
It creates a smooth mental path from curiosity to confidence.
A weak level throws content at the player.
A strong Kishōtenketsu level makes the player feel:
I discovered the rule myself.
That is the magic.
The designer is teaching, but the player feels like they are learning naturally.
PLayers want to learn the game! become a good teacher.
The level arranges situations so the player can build the idea from the inside.
Why this matters in prototyping
Kishōtenketsu is not only useful for final levels.
It is also useful while prototyping.
When I prototype a mechanic, I do not only ask whether the mechanic works. I ask whether it can survive a four-beat progression.
Is it simple enough to be clear?
Does it align with a well known mental model?
Can I develop it without adding noise?
Can I twist it without breaking its identity?
It is satisfying to be repeated again and again?
Can I make players learn it by themselves?
If the answer is yes, the mechanic probably has more depth than a single cool moment.
If the answer is no, the mechanic may still be interesting, but it might be a gesture, not a level idea, do not get rid of it, backlog it because in the future it may be useful.
A mechanic that cannot produce development or twist may still be pleasurable, but it may not yet be strong enough to carry a level, then, it becomes a subject to another post.
A close reading checklist
When looking at a level, I would ask:
Introduction
What is the central idea?
Where is it shown in its simplest form?
Can the player observe it before being punished by it?
Development
How does the level repeat the idea with more consequence?
What skill is being built?
What does the player understand now that they did not understand before?
Twist
What relationship changes?
Does the twist grow from the same idea, or does it feel imported?
Does it expand the player’s mental model?
Conclusion
How does the level ask the player to prove understanding?
Does the ending feel like a resolution or just an exit?
Safe reward validation
Does the game reward or confirm the player’s new understanding?
Is there a moment where the player feels: I own this now?
Before closing it
A last tip.
think about TEASING the next thing.
Once the player is comfortable enough with a mechanic to use it without confusion, begin teasing the next one.
Do not make the new mechanic playable yet. Let it pass through the player’s attention: show it in the environment, place it just out of reach, let another character use it, or frame it as something strange the player cannot yet understand.
This kind of teasing is not just decoration. It is part of the learning process. It creates curiosity, prepares the player’s mind, and makes the eventual introduction of the mechanic feel natural rather than abrupt.
teasing is also a big part of learning process.
Closing thought
Kishōtenketsu is not valuable because it gives us four labels.
It is valuable because it reminds us that a level must be taught.
Not as a lecture.
Not as a list of challenges.
Not as a pile of content.
It should be taught with a beginning, a development, a mutation, and a resolution.
When a level works this way, the player does not simply reach the end.
The player improves their understanding while moving through it.