Is This a Limitation of Games, or of Design?

 

So far, we've analyzed why visual novel emotions are special and how they differ from other media. We've confirmed that the visual novel genre has a structure well-suited for handling emotion.

But most real-world visual novels don't use that potential fully. Many provide an experience closer to "walkthrough optimization" than genuine emotion. Why? AI and I dug into the structural problems of existing visual novels.

The World of Affection+1

Most visual novels work like this: Choose option A, the character's affection rises by 1. Choose option B, affection stays the same or drops. Cross a threshold and you enter the romance route. Fall short and you get a bad or normal ending.

This system is easy to implement. One variable, a few conditionals. Testing is simple. Balancing is intuitive. Low production cost means most visual novels adopt this structure.

But this structure has fundamental problems.

When I asked AI to "analyze the affection system of existing visual novels," the response was striking: "Many visual novels aren't simulating emotion — they're just managing route unlock conditions." An accurate diagnosis. And that single sentence pierces the heart of the problem.

Problem 1: Relationships Reduced to Numbers

In an affection system, a relationship is an integer between 0 and 100. Liking at 100 and trust at 100 become the same thing. But in reality, "liking but not trusting" is perfectly possible. So is "respecting but not loving." And "disliking but unable to leave."

A single number can't hold this complexity. Collapse a relationship onto one axis and its subtlety vanishes. And from a relationship without subtlety, no emotion emerges.

When I pointed this out to AI, it agreed immediately with a real-life example: "Love and trust can both be high, but love can be high while trust is low — that's realistic. A single affection score can't express this difference."

Problem 2: Choices Become Quizzes

In an affection system, the player's behavior becomes answer-hunting. What does this character want to hear? Will this choice raise affection? Emotional judgment becomes strategic calculation.

This directly feeds the "walkthrough" culture of dating sims. Follow a guide, pick only correct answers. You can "clear" every route, but no emotion remains. Because the choices aren't personal judgments — they're information-based optimization.

Good visual novel choices should have no correct answer. Whether "be honest" or "let it slide" is right depends on the situation, the relationship, and the player's values. Only from such choices does real emotion emerge.

When I asked AI to "create examples of choices without correct answers," interesting results came out. AI produced structurally balanced choices well, but was weak on judging "which side carries more emotional weight." The weight of a choice comes from context and experience, not logical balance.

Problem 3: Endings Are Cutoff Lines

Affection 70+ means happy ending, 50+ means normal, under 50 means bad. In this structure, endings are like exam scores. Score high enough and you pass; otherwise, fail.

But relationship outcomes aren't determined by scores. In reality, relationships end for reasons beyond "not enough love." Mismatched timing, different expression styles, wanting different things, care that paradoxically creates distance. Relationships succeed or fail for diverse reasons.

When endings are determined by numerical cutoffs, these varied relationship outcomes become inexpressible. "High affection but failed ending" or "low affection but it works out" — impossible. Yet in reality, those relationships are more common.

In the previous entry I discussed Eternal Sunshine — in that film, the relationship fails not because "there wasn't enough love." They loved each other, and the way they loved wounded each other. Affection score 100, but failure. To express this, a single number won't do.

Problem 4: Characters Become Vending Machines

In an affection system, characters are machines that respond to input. Good choice in, positive reaction out. Bad choice in, negative reaction out. Predictable, consistent, mechanical.

People aren't like that. The same words are received differently depending on context. What felt like gratitude yesterday might feel like pressure today. Depending on atmosphere, timing, and context, the same action carries entirely different meaning.

When characters respond like vending machines, players don't perceive them as people. They perceive them as systems. And systems don't generate emotional attachment.

Problem 5: Results Are Immediately Visible

Another problem with affection systems is that feedback is instantaneous. Choose, and the character's reaction immediately shifts. The relationship instantly improves or worsens.

Real relationships don't work that way. The consequences of today's words might surface next week, or next month. "Why did you do that back then?" comes back months later. This delay creates the weight of relationships.

When affection changes instantly, the weight of choice vanishes. A wrong pick is immediately obvious, so save/load can undo it. Reversible choices don't carry regret. And relationships without regret lack emotional depth.

So This Is a Design Limitation

After this analysis, the answer is clear. When visual novels fail to produce emotion, it's not a limitation of games as a medium. It's a limitation of the Affection+1 design.

The medium has ample potential. The tool of choice, the accumulation of time, the participant's emotion. All these elements provide a structure capable of producing deep feeling. The problem is that most visual novels compress this potential into a single number and a cutoff line.

This happens because of production cost and complexity. Designing emotion for real means more variables, exponentially more combinations, more dialogue branches, harder testing. Most teams can't afford this cost, so they choose simple systems.

But what if AI is available? If AI can help with variable management, branch logic, and dialogue generation? The cost structure changes.

This is an intriguing possibility. Designs that were previously abandoned due to cost can now be attempted with AI's help. AI doesn't understand emotion, of course. But it can help build structures that contain emotion. Defining relationships between variables, managing combinatorial complexity, branching dialogue based on state — AI can do these faster and more accurately than humans.

This concludes the "why" exploration. Why make a visual novel, why do emotions emerge, why do existing systems have limits. Eight entries of digging into questions. The answer is clear: the problem isn't the medium — it's the design. And better design is possible.

So starting next entry, we design the emotion system in earnest. Not a single affection score, but multiple emotion variables operating in layers. Not raising love directly, but love emerging from a combination of trust, intimacy, and fear of loss. Building this with AI.

---

Next: Translating Emotion into Variables — Designing the Three-Layer Emotion Model