t1k:designer:base:balance-tools

Field	Value
Module	`base`
Version	`1.25.1`
Effort	`medium`
Tools	—

Keywords: balance, balance tools, game balance, stats, tuning

How to invoke

/t1k:designer:base:balance-tools

Game Balance Tools

Executed by

The calculators are cheap-tier once the formulas are stated; the decisions built on them are not. Computing DPS / EHP / TTK / ItemPower / DropRate from the Quick Formulas, and applying the Balance Validation Checklist thresholds to the results, is t1k-metrics-reporter work. The code-based stat audit applies a stated rule to N values and reports one verdict per value, which is t1k-list-classifier work (routing SSOT: rules/orchestration-rules.md § “Task-Type → Agent Routing”):

Step	Agent
Compute DPS / EHP / TTK / ItemPower / DropRate per unit, item, or encounter from the stated formula	`t1k-metrics-reporter`
Apply the Checklist thresholds (TTK in range, no value >2× next tier, power curve 1.1–1.2×, pity drop-rate math) to computed numbers and report which rows fail	`t1k-metrics-reporter`
Stat audit: flag each grepped stat constant >2× the median	`t1k-list-classifier` — one verdict per constant, rubric stated

Stays reasoning-tier: difficulty-spike detection and its remediation, whether a flagged outlier is a bug or an intentional power fantasy, what to retune and to what value, simulation win-rate variance judgment, and every Tuning Knobs range decision — those read the numbers and change the design.

When This Skill Triggers

Tuning stat values (ATK, DEF, HP, speed, cooldowns)
Validating difficulty across an encounter sequence
Auditing item power budgets for outliers
Checking economy balance (gold in vs gold out per session)
Detecting difficulty spikes (TTK jumps >50% between encounters)
Simulating loot distribution across N encounters
Writing automated balance tests

Core Toolkit

Quick Formulas (Inline Reference)

DPS       = (ATK - DEF × reduction%) × AttackSpeed
EHP       = HP / (1 - DamageReduction%)
TTK       = TargetEHP / AggressorDPS
ItemPower = BaseBudget × RarityMult × SlotSize
DropRate  = BaseRate% × LuckMultiplier × AreaModifier

Balance Validation Checklist

Before finalizing any balance pass:

TTK within target range for all unit matchups (3-5 hits for roguelike)
No stat value is an outlier (>2× next highest rarity tier item)
Economy: gold earned per run ≥ upgrade cost of 1 item tier
Encounter win rate 55-75% at recommended level (simulate or playtest)
Power curve: player power 1.1-1.2× ahead of encounter curve at each point
Pity timers trigger within expected window (verify drop rate math)
No difficulty spike >50% TTK increase between consecutive encounters

→ See references/balance-spreadsheet.md for spreadsheet layout and all calculator formulas

Validation Methods

Stat Audit (Code-Based)

To find outliers without a spreadsheet:

# Grep all stat constant definitions
grep -rn "const float\|const int" Packages/ Assets/Demos/ --include="*.cs" | grep -i "atk\|def\|hp\|speed\|damage"

Sort results; flag any value >2× median as a suspect outlier.

Difficulty Spike Detection

Plot TTK per encounter index. A spike is when: TTK[i+1] / TTK[i] > 1.5

Flag and insert a rest/low-difficulty encounter before the spike.

→ See references/validation-methods.md for simulation testing, power curve graphing, automated balance tests, and empirical player-data methods (death heatmaps, completion drop-off thresholds, structured friend testing)

Common Tuning Knobs

Knob	Low Value Effect	High Value Effect	Recommended Range
ATK/DEF ratio	Tanky, grindy	Bursty, one-shot risk	1.5-3×
HP per level	Flat, predictable	Dramatic power jumps	10-15% per level
AttackSpeed	Strategic, positional	Spammy, micro-intensive	0.5-2.0 attacks/s
CooldownDuration	Spammy abilities	High-stakes decisions	5-15s for impactful skills
LootPerEncounter	Slow build, tension	Fast power, trivialized	0.5-1.5 items avg
MovementSpeed	Strategic, positional	Chaotic, hard to read	3-6 units/s

→ See references/tuning-knobs.md for full parameter table with interaction effects and caps

Gotchas

#	Issue	Fix
1	Flat defense scales to 100% reduction	Cap at 75% or use percent-based with diminishing returns
2	Infinite stat stacking breaks balance	Define soft caps (diminishing returns) and hard caps
3	Loot rate feels bad despite correct math	Add pity timer — guaranteed drop after N consecutive misses
4	Simulation win rate misleads	Check variance too — consistent 60% is better than 50%/70% alternating

Cross-References

rpg-game-design — Formulas and design theory behind the tools
game-design-document — Where to record balance decisions in GDDs
engine-specific code skills (auto-activated via registry) — for actual runtime values

Reference Files

File	Coverage
`references/balance-spreadsheet.md`	Full calculator formulas: DPS, EHP, item power, economy, drop rate
`references/validation-methods.md`	Simulation testing, stat audit, power curve graphing, automated tests, empirical player-data methods (death heatmaps, completion drop-off, friend testing)
`references/tuning-knobs.md`	Parameter effects, interaction effects, recommended ranges, caps