Solo vs. Party

tim-cain system-design party-control solo-character rpg-design game-balance stealth reputation testing

Abstract

Problem: Should an RPG give the player control over a single character or an entire party? What are the design trade-offs?

Approach: Tim Cain draws on his experience shipping both styles — solo control (Arcanum, Fallout) and full party control (Temple of Elemental Evil) — to compare the two approaches across player identity, game scaling, combat design, stealth, reputation, testability, and multiplayer.

Findings: Solo control creates a tighter player-character bond, simplifies scaling/testing, and naturally supports multiplayer, real-time combat, and first-person cameras. Party control offers class versatility, richer tactical combat, and easier boss balancing — but introduces thorny design problems around stealth, reputation, unwinnable party compositions, and exponential testing requirements.

Key insight: Solo control front-loads simplicity by making the player the character; party control front-loads player freedom but pushes enormous design and testing costs onto the developer. Choose based on taste, but accept the consequences.

Source: https://www.youtube.com/watch?v=-cG8OfcFkew

1. Why Tim Prefers Solo Control

The core appeal is identity. When you control one character, you are the hero. Every action — opening a door, tripping a trap, talking to an NPC — is yours. There's no blaming "dumb Bob" for stepping on a pressure plate. This creates a tight bond between player and character that Tim considers the most important quality an RPG can have.

1.1. Scaling and Testability

With a single character, the designer knows exactly what they're balancing against. You can ask: is every class viable? Can every skill combination finish the game? Are all boss fights doable? Companions, when present, only make things easier — they're additive, not a variable that explodes your test matrix.

1.2. Death Is Simple

Game over = your character dies. Maybe unconsciousness is survivable if a companion heals you, but the fail state is crystal clear. No ambiguity about "does the game end when three of five party members drop?"

1.3. Multiplayer, Camera, and Pacing

Solo control keeps all options open: real-time or turn-based, first-person or third-person, single-player or multiplayer. Each of these becomes significantly harder once one player controls multiple characters.

2. The Case for Party Control

Tim acknowledges genuine strengths of the party approach:

2.1. Class Versatility in One Playthrough

You don't have to replay the game to experience a wizard, a thief, and a bard. The full breadth of the class system is visible in a single run, which is inherently fun and rewarding.

2.2. Tactical Depth

Party combat — especially turn-based — unlocks rich tactical options: tanks up front, ranged DPS in back, crowd-control specialists who'd be hopeless solo but shine in a group. Out of combat, you choose who checks the door, who talks, who searches for traps.

2.3. Boss Balance Is More Forgiving

Since a total party kill is required for a game over, bosses can be tuned a bit tougher. If three of five characters go down, the remaining two can still clutch a win. Restricted items (class-locked swords, skill-gated armor) also find homes more easily across a diverse party.

3. The Hidden Costs of Party Control

3.1. The Stealth Problem

If the whole party stealths, you're only as stealthy as your clumsiest member. Solutions — splitting the party, sending the thief ahead alone — open a nightmare of design questions: How far can party members separate? Different screens? Different map instances? Each answer spawns more work.

Alternatively, you pick one character's stealth skill to represent the whole party, but then you need UI to communicate which character matters and why.

3.2. The Reputation Problem

When one party member murders an NPC, the entire party's reputation tanks. This feels strange when your party contains both a paladin and an assassin. Tim's solution in Temple of Elemental Evil was alignment partitioning — simply forbidding certain combinations — but it's a blunt instrument.

3.3. Exponential Testing

This is the big one. If testing single-character builds is already combinatorially hard, a five-member party raises it to the fifth power. Unwinnable party compositions almost certainly exist, and finding them all is brutal.

Tim shares that while developing Temple of Elemental Evil, he deliberately tested the hardest composition he could think of: an all-halfling, all-bard party. He finished the game with it, but it was dramatically harder than a balanced party. The takeaway: you have to test degenerate combinations, and that work scales mercilessly.

4. Tim's Bottom Line

Choose the approach your taste drives you toward, but own the consequences. Solo control is limiting in player options but comes with a mountain of design problems already solved. Party control gives players more freedom but dumps significant design, UI, and testing work onto the developer.

Tim's personal preference is solo — ultimately because it answers the question "who is the player?" with the clearest possible answer: that character, right there, the one they made.

5. References