Deep investigative analysis that challenges assumptions and uncovers hidden patterns in tournament data. Use when exploring non-obvious questions, testing common beliefs, or running targeted statistical investigations that the standard pipeline does not cover.
Question-driven investigative analysis on data/ CSVs. Every analysis starts as
a testable question, is validated empirically, and is explained accessibly.
Does NOT duplicate work already in STATS_REPORT.md or ANALYTICAL_BRIEF.md.
STATS_REPORT.md or ANALYTICAL_BRIEF.md.ANALYTICAL_BRIEF.md and to know what is already covered.STATS_REPORT.mdspirit_of_the_law_analysis.py (or individual investigation functions).SPIRIT_FINDINGS.md (report), assets/spirit/ (charts), data/spirit/ (derived CSVs).| # | Name | Question |
|---|---|---|
| 1 | Snowball Effect | Does winning game 1 predict winning the series? |
| 2 | Draft Order Advantage | Does picking first confer a measurable edge? |
| 3 | Fatigue Factor | Does performance decay across a series? |
| 4 | Comfort Picks vs. Wild Cards | Do players win more on their most-played civs? |
| 5 | Clutch Factor | Who over/underperforms in deciding games? |
| 6 | One-Sided Civ Matchups | Which civ pairings are most imbalanced? |
| 7 | Map Specialists | Which players over/underperform on specific maps? |
| 8 | Upset Probability | Do underdogs win more often than ELO theory predicts? |
| 9 | Tempo Control | Does controlling game length correlate with winning? |
| 10 | Meta Evolution | How do civ picks shift as tournament stakes increase? |
SPIRIT_FINDINGS.md -- structured report, one section per investigation.assets/spirit/ -- one visualization per investigation that warrants a chart.data/spirit/ -- derived CSVs (e.g. civ_matchup_matrix.csv, player_map_affinity.csv)..head(), .shape, .describe() only. Never dump full CSVs.