Ensure meaningful choice — no dominant strategies, no dead options, no runaway leaders or hopeless laggards. Use this skill when: (1) evaluating whether all player options are viable, (2) identifying dominant strategies that eliminate meaningful choice, (3) designing asymmetric balance (different factions, characters, or roles that feel fair despite being different), (4) building Monte Carlo or statistical models to test probabilistic systems, (5) creating balance levers — the knobs you can turn to adjust without redesigning, (6) diagnosing why a game feels unfair even when it technically isn't, (7) building matchup matrices for competitive games. Medium-agnostic — works for digital, tabletop, and hybrid games.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
Ensure meaningful choice — no dominant strategies, no dead options, no runaway leaders or hopeless laggards. Use this skill when: (1) evaluating whether all player options are viable, (2) identifying dominant strategies that eliminate meaningful choice, (3) designing asymmetric balance (different factions, characters, or roles that feel fair despite being different), (4) building Monte Carlo or statistical models to test probabilistic systems, (5) creating balance levers — the knobs you can turn to adjust without redesigning, (6) diagnosing why a game feels unfair even when it technically isn't, (7) building matchup matrices for competitive games. Medium-agnostic — works for digital, tabletop, and hybrid games.
Game Balance Analyst
Balance is not "everything is equal." Balance is "every choice is interesting." A perfectly balanced game where all options are identical is not balanced — it's boring. The goal is meaningful asymmetry where every option has a reason to exist.
Balance Philosophy
What Balance Actually Means
Balance means: Every player-facing option is a viable choice in some context. No option is strictly better than all alternatives. The player's decision matters.
Balance does NOT mean:
Everything deals the same damage
All factions have the same units
Win rates are exactly 50/50
Every strategy is equally good in every situation
The balance target: Players should lose because they were outplayed, not because they picked the wrong character/faction/build at the select screen.
Types of Balance
Symmetric balance — All players start with identical options.
Asymmetric balance — Players have different options that are intended to feel equally viable.
StarCraft factions, fighting game characters, role-based co-op
Balance challenge: Different capabilities must produce comparable outcomes
Harder to balance, more interesting to play
Dynamic balance — Balance state changes over the course of play.
Early-game vs. late-game power, meta-game evolution, seasonal patches
Balance challenge: A strategy dominant at one phase must be vulnerable at another
Requires balancing across time, not just at a snapshot
Contextual balance — Options are balanced within specific contexts.
Rock-paper-scissors relationship between unit types
Terrain-dependent strategies
Balance challenge: No context should be overwhelmingly common
The ideal — balance through interesting trade-offs
Core Deliverables
1. Balance Framework Definition
Before balancing anything, define what balanced means for this specific game:
BALANCE TYPE: [Symmetric / Asymmetric / Dynamic / Contextual / Hybrid]
BALANCE METRIC: [What are we measuring? Win rate? Option pick rate? Fun?]
BALANCE TOLERANCE: [How far from "ideal" is acceptable?]
(e.g., "All factions should win 45-55% of the time")
INTENTIONAL IMBALANCE:[What is deliberately NOT balanced, and why?]
(e.g., "Late-game options are intentionally stronger to
reward investment")
2. Dominant Strategy Analysis
A dominant strategy is an option that is strictly better than alternatives — it eliminates meaningful choice.
Detection methods:
Option comparison matrix:
OPTION SCENARIO A SCENARIO B SCENARIO C OVERALL
──────────── ──────────── ──────────── ──────────── ────────────
Warrior Strong Strong Medium Dominant?
Mage Weak Strong Strong Contextual
Rogue Medium Weak Strong Contextual
Healer Weak Weak Weak Dead option
If any option is "Strong" in every scenario → likely dominant.
If any option is "Weak" in every scenario → likely dead.
Efficiency analysis:
For each option, calculate value per cost:
Healthy: circular counter relationships. Unhealthy: linear "just be better" hierarchy.
3. Option Viability Matrix
Every player-facing choice should be worth considering. Audit all options:
OPTION NICHE PICK RATE TARGET ACTUAL STATUS
────────────── ──────────────────── ────────────────── ──────── ──────────
[Option name] [When is this best?] [Expected %] [Data %] [Healthy/
Over/Under]
Viability criteria:
The option has at least one scenario where it's the best choice
The option has a distinct identity (not just a worse version of another option)
The option requires a different playstyle (not just different numbers)
A player who picks this option doesn't feel punished
The option has counter-play (it can be beaten)
Dead option symptoms:
Pick rate below 5% in competitive play
No scenario where it's optimal
Community consensus that it's "the worst option"
Only picked by players who don't know better
Overpowered option symptoms:
Pick rate above 40% in competitive play (for games with 4+ options)
Optimal in most scenarios regardless of context
Community meta coalesces around this option
Counter-play exists but is too narrow or too costly
4. Statistical Modeling
For games with randomness, intuition isn't enough. Model the probabilities.
Expected value calculations:
ACTION: Attack with Flameblade
Hit chance: 70%
On hit: 2d6+3 damage (average 10)
On miss: 0 damage
Expected value: 0.7 * 10 = 7.0 expected damage per attack
vs.
ACTION: Attack with Iceshard
Hit chance: 90%
On hit: 1d6+1 damage (average 4.5)
On miss: half damage = 2.25
Expected value: (0.9 * 4.5) + (0.1 * 2.25) = 4.275 expected damage per attack
Flameblade has higher expected damage but higher variance. Iceshard is consistent. Both are viable if the game rewards different risk profiles.
Monte Carlo simulation approach:
For complex systems where analytical solutions aren't feasible:
Define the variables and their ranges
Run N simulated games/encounters (N = 1000+ for statistical significance)
Record outcomes per strategy/option
Analyze distribution, not just averages
Look for outlier cases — the 1% outcomes that break the experience
Variance analysis:
High variance creates exciting moments but frustrating consistency.
Low variance creates reliable gameplay but can feel flat.
SYSTEM CURRENT VARIANCE TARGET VARIANCE RATIONALE
────────────────── ────────────────── ──────────────── ──────────────────
Combat damage High (d12 range) Medium Too swingy for
tactical game
Loot drops Low (fixed tables) High Discovery needs
surprise
Movement range None (fixed) None Predictability
enables planning
5. Balance Lever Identification
Balance levers are the knobs you can turn to adjust balance without redesigning systems. Identify them early.
Quantitative levers (numbers to adjust):
Damage/healing values
Costs (resource, time, opportunity)
Probabilities (hit chance, drop rate, crit rate)
Durations (cooldowns, effect timers)
Ranges (attack range, movement range, detection range)
Capacities (inventory size, hand limit, unit cap)
Structural levers (mechanics to modify):
Adding/removing counter-play options
Changing resource conversion rates
Adjusting prerequisite chains
Modifying timing windows
Environmental levers (context to change):
Map/level design that favors different strategies
Encounter composition that requires different approaches
LEVER: [What can be adjusted]
AFFECTS: [What systems/options does this change?]
DIRECTION: [Does increasing this buff or nerf?]
SENSITIVITY: [How much does a small change matter? Low/Medium/High]
SIDE EFFECTS: [What else changes when this is adjusted?]
CURRENT VALUE: [What is it set to now?]
VALID RANGE: [What are the reasonable min/max?]
List all player-facing options — Everything the player can choose
Run dominant strategy analysis — Are any options clearly best?
Build option viability matrix — Does every option have a niche?
Model statistical systems — Calculate expected values, run simulations
Identify balance levers — What knobs can we turn?
Build matchup matrices (if applicable) — How do options fare against each other?
Cross-reference with rules (Skill 8) — Are the formalized rules producing the intended balance?
Cross-reference with economy (Skill 10) — Are economic values balanced?
Flag to playtest (Skill 16) — Balance hypotheses need live testing
Responding to a balance complaint
Clarify the complaint — "too strong," "feels unfair," "this option is useless"
Gather data — Is this perception or measurement? Both matter, but differently.
Check the option viability matrix — Is this option actually underperforming?
Check matchup context — Is it underpowered generally or in a specific matchup?
Check the player's skill level — Is this a balance issue or a knowledge issue?
If real: identify the appropriate balance lever
Model the proposed change — What does adjusting this lever do to all affected systems?
Check for cascading effects — Does the fix create a new problem?
Propose the minimum change that addresses the issue
Ongoing balance maintenance (live games)
Track pick rates and win rates per option
Compare actual data to balance targets
Identify emerging dominant strategies
Distinguish between meta-game shifts and genuine imbalance
Plan balance changes in batches, not one at a time
Communicate changes and reasoning to players
Log all changes in Design Iteration Tracker (Skill 17)
Balance Anti-Patterns
The Nerf Spiral — Repeatedly nerfing the strongest option instead of buffing weak ones. Eventually everything feels bad.
The Spreadsheet Trap — Numbers are perfectly balanced but the feel is wrong. A weapon that does 10 DPS on paper but feels terrible to use is not balanced.
The False Symmetry — Making asymmetric options "balanced" by making them functionally identical. If the Warrior and Mage play the same, why have both?
The Complexity Tax — A hard-to-execute option is balanced on paper but its difficulty makes it unviable for most players. Balance must account for execution difficulty.
The Meta Ignore — Balancing individual options without considering the meta-game. Option A is balanced against Option B, but when combined with Option C, it's broken.
The Patch Treadmill — Constant balance changes that prevent players from ever learning the game deeply. Stability has value.
