STS2 Run Analyzer

A Python toolkit for analyzing your Slay the Spire 2 save data — win rates, card performance, relic tracking, and interactive dashboards.

Save File Location

Save files live at:

C:\Program Files (x86)\Steam\userdata\<steam_id>\2868840\remote\profile1\saves\history\

Each run is stored as a timestamped .run JSON file (e.g. 1772733327.run).

Setup

git clone https://github.com/YOUR_USERNAME/sts2-analysis.git
cd sts2-analysis
pip install -r requirements.txt

Quick Start

# Print a stats summary + generate an HTML dashboard
python scripts/analyze.py analyze \
  --saves-dir "C:\Program Files (x86)\Steam\userdata\YOUR_STEAM_ID\2868840\remote\profile1\saves\history" \
  --output dashboard.html

# Inspect raw keys of a single save file (useful for exploring new fields)
python scripts/analyze.py inspect \
  --saves-dir "C:\...\history"

Or open the Jupyter notebook for interactive exploration:

jupyter notebook notebooks/exploration.ipynb

Data Available Per Run

Field	Description
win	Whether the run was won
ascension	Ascension level
game_mode	standard, daily, etc.
seed	Run seed string
run_time	Total playtime in seconds
start_time	Unix timestamp
killed_by_encounter	What ended the run
acts	Act IDs traversed
players[].character	Character class
players[].deck	Final deck with per-card floor acquired
players[].relics	Relics with floor acquired
map_point_history	Full per-floor timeline

Per-floor data (map_point_history)

Each floor records: HP, gold, damage taken, card choices (offered + picked), cards gained, encounter ID, turns taken.

Project Structure

sts2-analysis/
├── sts2_analysis/
│   ├── parser/        # Load .run JSON files
│   ├── models/        # Run, CardEntry, RelicEntry, FloorStats dataclasses
│   ├── analysis/      # run_stats, deck_analysis, relic_tracker
│   └── viz/           # Plotly dashboard generation
├── notebooks/
│   └── exploration.ipynb
├── scripts/
│   └── analyze.py     # CLI entry point
└── data/
    └── sample_saves/  # (gitignored) put .run files here for testing

Sample Stats (445 runs, 196 hours)

Metric	Value
Win rate	29.7%
Avg run time	26.4 min
Avg floors reached	30.5
Most played character	Regent
Best win-rate character	Silent (42.9%)
Top win-rate card	Backflip (64.5%, 31 appearances)
Top win-rate relic	Tungsten Rod (76.5%, 17 appearances)

Contributing

The save schema is versioned (schema_version field, currently v8). If you discover new fields after a patch, add them to sts2_analysis/models/run.py and open a PR.