Jo-Ha-Kyu and the Art of Game Design

Sam Liberty

This article deals with designing pacing into games. We’re certainly not the first to think about this topic, nor the first to compare games to stories with a beginning, middle, and end. As writing and lit major with a theater background (and a game writer / game designer) the idea of pacing and storytelling has always been a focus of mine, and something I strive to do well in games I design.

You’re probably already familiar with the action diagram of a typical Western story or play, but here it is anyway for reference’s sake.


As you can see, it begins with exposition, the action builds to a climactic moment, then there’s a cool-down where loose ends are settled and the action simmers down, until everything is finally resolved. This is a tried and true method of storytelling, and can certainly be likened to gameplay, especially player vs. player gameplay. The opening moves are the exposition (if you want to be persnickety, you can count learning the game, tutorials, start screens or setting up the board here, too). As players stake out their positions and develop their strategies, the feel is one of rising action. At some point, the game is decided. This is the point in the game where one player makes a move that solidifies him as the winner. If you watch sports or play a lot of board games, you’ve probably seen this moment countless times. “Oh, there’s the game.” From that point on, there’s falling action. The game is concluded (the last Provinces are purchased, the final train cars are placed, or the turn clock runs down), then all that’s left is resolution: counting up scores.

This story structure is so ingrained in the Western psyche that it seems natural to apply it to all of life’s little processes. But it’s not exactly the right fit for gameplay, in my opinion. Note that everything after the Climax is low on impact and meaning, since the game has already been decided. If I could design the perfect game, it would follow a very different diagram. The game would live as long as it could in the space to the left of the Climax. Rising action. This is where strategies are created and executed. It’s the true test of a player’s mettle, and determines the outcome of the end-game. This is what we think of as the early-to-mid game. The end-game becomes the terminus of play, and not just for all practical purposes, but literally the end, holding off the Climax to the last possible moment. The falling action, if there is any, lasts seconds. Resolution becomes almost unnecessary, since the Climax was so decisive, but counting up score is fine.

In other parts of the world, this story structure is actually the norm. The Japanese call it Jo-Ha-Kyu — literally “beginning, break, rapid.”

The philosophy of Jo-Ha-Kyu is central to many Japanese arts, from kendo to the iconic tea ceremony. But it’s probably best known as the dramatic structure of traditional Japanese theater forms such as kabuki and noh. I first learned of the concept from Yoshi Oida in his book The Invisible Actor. In Japanese theater, the play begins slowly. The plot moves slowly, events happen gradually, and sometimes the actors themselves even move slowly, drawing out every action and every movement into a glacial pantomime.

As tension builds, actions become quicker, more deliberate. The pace begins to gradually quicken with each plot development, chaining together into a final climactic moment. At this point, the story is over, but for a brief denouement that only even comes if necessary.

I think about this concept a great deal when I’m designing and playtesting games. Is the game exciting all throughout? When, if ever, is the winner apparent? Do we ever feel like players are just “going through the motions” to complete the game, even though the results are a forgone conclusion? How many players have a stake in the climax? And do players feel like they have enough time to actually develop their strategies, build their engines, and execute them?


The best games engage players from start to finish, filling each moment with tense decisions. Here’s Yoshi Oida on the topic of Jo-Ha-Kyu. Feel free to substitute the word “players” for “audience,” and “designer” for “director.”

Jo, Ha, Kyu isn’t just an esoteric theatrical concept, but a rhythm that the audience senses in their flesh and bones. If the actor or director is not aware of this fact, you could end up with a production where there is a contradiction between the inner rhythms of the audience and the production. In this case, the audience cannot relax and allow themselves to be carried into the performance.

When a game ends, but it doesn’t feel satisfying, I often say to myself “this game had no Jo-Ha-Kyu.” This isn’t the gravest sin a game can commit, but when a game does have Jo-Ha-Kyu, it’s magical. Video game designers have been hip to this structure since the 80s. Think about your favorite video games. Chances are they feature elements such as boss battles, skill-building, and a slow ratcheting up of difficulty if not stakes. The first five minutes of the game feel nothing like the last five minutes of the game, and once you beat the final boss, you’re not forced to mop up a bunch of goombas on your way out.

Jo-Ha-Kyu and Engagement

Up until now, I’ve been speaking very abstractly about “games,” but at EGL, all of our games all have the primary design goal of engagement and real civic action. Our flagship game Community PlanIt succeeds in engaging players and creating action in many ways where other grass-roots, online civic tools tend to come up short. Civic platforms that come in the form of online suggestion boxes or forums fall into disuse and lack engagement because citizens aren’t driven to engage at the same time. They begin, but don’t end. There’s no sense of forward progress. The narrative never changes. They lack Jo-Ha-Kyu.

How does Community PlanIt use Jo-Ha-Kyu to engage its players throughout?

Puzzle games like Tetris capture Jo-Ha-Kyu by ratcheting up difficulty and complexity, and reducing the amount of time available to players. The last seconds of play feel very different than the first seconds of play – and this is what we like to engage players. Community PlanIt does this in a couple ways, even though it is not a puzzle game.

The first way may seem obvious, but it’s the number one thing that Community PlanIt does that other online community planning platforms don’t: Tight time limits. A game of Community PlanIt lasts for three missions, and each mission lasts just one week. Once a mission’s over, it’s over, and a new one begins. For example, if I play Mission 1 in early November, every other player is playing it at the same time. If I join in late November, I start right on Mission 3, so I’m exploring the same content as every other player. This means that I miss out on Mission 1 and 2, but that’s the price we pay for concentrating player attention and action. Plus, it creates a real reason to log on and play. You know that it won’t be there forever. A prominent countdown clock marks the minutes in each mission, so players know they must act fast.

The second way is through escalating stakes. In mission one, challenges are worth 10 coins each. In mission two values doubles to 20, and in mission three it triples to 30. This means that late actions are worth more than early actions, and so the winner of the game is not clear until the final moments. The #1 Cause and the #1 player at the end of Mission 1 are virtually never the winners at the end of the game. Things start out slow, and then explode near the end. Special coin awards that players can earn through forum actions create even more variability and uncertainty about who will win the game.

The last way we create a sense of Jo-Ha-Kyu is through narrative. Each mission has its own narrative, which we build through content. The challenges of a given mission are framed by the introduction to the mission, and the mission video, and the subject matter builds on that of previous missions. We do this to build a through-line that pulls players from the beginning of the game to the end. An individual mission of Community PlanIt contains this in microcosm, too. There are three blocks of challenges, each punctuated by a Crat – the annoying but well-meaning mascots of Community PlanIt who ask trivia questions to the players. Beating a Crat unlocks the next set of challenges. The final test of each mission is the Boss Crat. He asks the hardest trivia question of all, and beating him gives you 5x the coin value of a typical challenge. This puts the highest stake event (where else?) at the very end of the game.

Tips on fostering Jo-Ha-Kyu in games of all sorts

Here are some signs that your game lacks Jo-Ha-Kyu:

  • The game feels like it ends abruptly
  • The last turn (or last few turns) is boring
  • The last turn feels no different than the first turn
  • Players complain that they didn’t have enough time to build
  • Everyone knows who will win the game halfway through play

If your games are suffering from any of the above symptoms, you might want to tweak the pacing. Here are some things we’ve done in the past to enhance the element of Jo-Ha-Kyu:

  • Introduce items or actions that are selected early, but take effect on later turns
  • Use mechanics with rising action and climax (but no falling action) built in, such as a best-of-three showdown
  • Make confrontations decisive, but include catch-up mechanics
  • Play with the end-game condition itself; should this game end sooner? Or later?
  • Give players plenty to do in the early and mid games, but make sure choices made on turn one or two don’t make or break each game
  • Resolve the game after a set time instead of giving players the power to end (and therefore extend) the game
  • If the end-game must be player-triggered, make sure each action a player takes inexorably pushes the game toward completion

Hope you found this useful. In the spirit of Jo-Ha-Ku, this article is now over. Go design games!