Why Wildheart Sanctuary went from eight characters to three

The first build of Wildheart Sanctuary had eight characters. The first release will have three. The other five have not been cancelled: they will be added after launch, one at a time, free of charge.

This devlog explains that decision. It is the largest one made on the game so far, and it changed almost every system in it.

The short version

• The first build had eight characters and one affection meter. Every character was thin.
• I cut the first release to three characters and gave each one roughly eight times the writing.
• Depth is what makes the game good. Variety is something that can be added once the game is out.
• The five characters that were moved will be released as free episodes. They were part of the original plan, so they will not be sold as DLC.

What the first build looked like

The first playable build, which I call version 0.0, had the outline of the whole game. There were eight monster girls to bring to the sanctuary. There were expeditions, upgrades and a gallery. And each character had a single affection meter that went up when you spent time with her.

On paper that is a complete game. In play it was not, and the reason was arithmetic.

The build had about 250 lines of character writing in total. Spread over eight characters, that is about 30 lines each. Thirty lines is an introduction, a handful of reactions and one or two scenes. You could meet everyone in the sanctuary in an afternoon and have nothing left to learn about any of them.

The single meter made it worse. With one number per character, every interaction did the same thing: it moved the number up. There was no difference between getting to know someone and being attracted to her, no reason to choose one activity over another, and no way for a character to respond to you differently on different days.

Eight characters looked like a lot of game. It was a wide game, one layer deep.

Width and depth

A game with characters in it can grow in two directions.

Width is how many characters there are. It is the number on the store page, and it is the easiest thing to promise.

Depth is how much there is to do with each one: how many scenes, how many ways to spend time together, how much she remembers, how often she surprises you.

Both cost writing, art and testing, and a small studio cannot afford both at once. The question is which to spend on first.

I chose depth, for one reason above the others. With more depth the game has more quality. A character with ten scenes, her own activities, her own quest and things she only tells you once she trusts you is a character. A character with thirty lines is a portrait with a name.

The second reason is about order. Variety can be added after launch: a new character arrives as a self-contained episode and takes nothing away from the ones already there. Depth does not work that way. Going back to deepen eight shipped characters means rewriting all eight while players are already partway through them.

The third reason is time. Three characters written properly can be finished, tested and released much sooner than eight. A smaller game that exists is worth more than a larger one that does not.

What changed in the game

Cutting to three was not only a matter of deleting five characters. The freed-up effort went into systems that only make sense when there is a lot to say about each character.

Two bonds instead of one

The single meter became two. Trust is emotional closeness. Desire is physical attraction. They grow from different activities and they open different scenes, so two players can know the same character in different ways. Some scenes need both.

Favors, and the right to say no

Each character has favors you can ask for. If the bond is not there yet, she says no. The refusal costs nothing, and it tells you what she would need first. That mechanic only works when a character has enough stages for “not yet” to mean something.

A day with a shape

Days are now split into morning, afternoon, evening and night, and each character keeps a schedule. You find her in the garden, the kitchen or the training grounds depending on the hour. Each one also wakes in a different mood, which changes what she says and what she wants to do.

Stories that belong to one character

Each character has a personal quest in three chapters, told through expeditions, and a journal that fills in as she tells you about herself. The three also have stories with each other: the sanctuary has 23 events, and some of them are about the friendships between the residents.

None of this would have been affordable across eight characters. All of it was affordable across three.

Checking that depth is paced well

More content per character creates a new problem: it has to arrive at a steady rate. A deep character whose scenes all unlock in the first hour is no better than a shallow one.

The game has a balance simulation for this. It plays the game automatically, 40 times over 80 in-game days, and reports when things happen. These are the current medians for version 0.1.

Two of the four are on target. Two are not. The first adult scene arrives a day late, and there is a stretch in the middle of the game where a player waits six days for a new story scene. That gap is the main thing being worked on before release. Conversations, events and quests fill some of it, since the simulation counts story scenes only, but six days is too long.

I am publishing the misses because the point of measuring is to find them.

The five that were moved

Five characters were taken out of the first release: a mermaid, a salamander, a kitsune, a minotaur and a succubus. They are designed and partly written. The regions they come from are already on the expedition map.

They will be added after launch as episodes, one character at a time. Each will arrive with the same depth as the first three: her own scenes, activities, quest and bonds.

No dates are set. I would rather announce each one when it is ready than publish a schedule and miss it.

Why they will be free

The remaining characters will be released free of charge.

This was a deliberate decision. The game was first planned with eight characters. Anyone who follows it should not feel that characters were cut from the game in order to be sold back as DLC. That practice is common, and players are right to dislike it.

Narrowing the first release was a decision about quality and time. It was not a decision about what to charge for. Releasing the rest for free is how that stays true.

What I would tell another developer

Three things, from making this decision on a real project.

1. Do the arithmetic early. Divide your total writing by your character count. If the number per character is small, the count is too high.
2. Cut width before you cut depth. A player forgives a small cast. A player does not forgive a cast of strangers.
3. Say what happens to what you cut. If it is coming back, say so, and say what it will cost. Silence is read as the worst case.

If you are planning a game of your own, the Journal’s guide to how to make an adult game covers engines, stores and payment rules.

What comes next

Version 0.1 is a working build in testing. The work now is balance, mainly closing the six-day gap, and art. Progress will be reported here in the devlog, and the waitlist is told first when early access opens.