Why You Can’t Attack That NPC: Game Design Explained

The Immersion Problem: Why You Can’t Always Hurt NPCs

You’re immersed in your character. The NPC is being pushy, annoying, or deserves a comeuppance. Your gut reaction is to throw a punch—or in this case, target something more creative. You reach for the attack action and… nothing. The NPC remains untouchable. Your character can’t do what you’d naturally do in that moment.

This is one of gaming’s most frustrating invisible walls. It breaks immersion right when you’re deepest in the world. But it exists for reasons that go way deeper than simple technical limitations.

Why Games Lock Down NPCs

The core issue is story protection. Most games—especially narrative-driven ones—are built around a predetermined sequence of events. If a quest-giver is supposed to stay alive, the game has to make them unkillable. If an important conversation is supposed to happen, NPCs can’t be disrupted by combat or creative actions.

Opening up full NPC vulnerability means the design team has to account for thousands of branching outcomes. What happens if you kill the quest-giver before they deliver the quest? The game breaks. Players get stuck. Stories collapse.

Developers could technically build contingencies for every possible player action—alternate dialogue if the NPC is wounded, different quest-givers, storylines that adjust on the fly. That’s labor-intensive and expensive. Most games instead choose to protect key NPCs entirely.

Different Games, Different Rules

Tabletop RPGs handle this completely differently. A good GM improvises. You want to kick the NPC? Roll for it. The NPC reacts. The story adapts. There’s no script to break.

Video games exist on a spectrum. Story-heavy linear games lock down NPCs tightly. Open-world games like Skyrim or GTA give you more freedom, but even then, essential NPCs are often protected to keep the main story intact. Games designed around player chaos from the start—sandbox games, immersive sims, some indie games—handle this better because the entire design anticipates unpredictability.

Text-based games and MUDs? Players can do whatever they want because the world adapts. AI responses scale up to handle it.

What About Games That Do Let You?

Some games have pushed back on NPC invulnerability. Bethesda games let you kill almost anyone. The Sims lets you trap NPCs, delete them, torment them. Dwarf Fortress expects utter chaos and rewards it.

These games work because they’re either sandbox-first (no rigid plot to break), consequence-light, or they accept soft-failure states. They also tend to have smaller, tighter stories or completely emergent narratives.

The trade-off is real: you get player freedom at the cost of a less tightly controlled narrative experience.

The Evolution: More Agency, More Complexity

Newer games are experimenting with this boundary. Some use story flags: if you kill someone important, the game genuinely acknowledges it and adjusts. Divinity: Original Sin 2 lets you kill anyone, including your allies and all quest-givers—the game has fallback options. Baldur’s Gate 3 went even further, with reactivity built into the skeleton of the game design.

This requires massive design overhead. You’re essentially building multiple game paths, multiple dialogue sets, multiple resolution options. It’s why these games take so long to develop and cost so much.

The Practical Reality

For most games, the invisible wall around NPCs is a practical design choice, not an artistic one. Developers are balancing player agency against story coherence, budget against scope, and immersion against maintainability.

It’s frustrating because it’s a visible seam in the game world. But it’s also why most games can ship with a complete story intact.

If you want maximum freedom, look for games that build around it from the ground up. Otherwise, you’re hitting the tradeoff that’s defined game design for decades: more player freedom requires exponentially more design and testing.