Monthly chess opening intelligence

Track where opening theory and practical play diverge.

OpenCast analyzes monthly Lichess opening data, forecasts win-rate movement, and highlights where human results outperform or lag behind engine expectation.

216 openings tracked
41 months of data
62 high-confidence forecasts
Last updated: 2026-06
↑ 34 rising↓ 37 falling→ 427 stable+5pp outperformance: 8-5pp underperformance: 39

Largest absolute engine delta

A44

Benoni Defense: Old Benoni

Engine: 57.7% → Human: 46.7%

-11.02 pp vs engine

Largest absolute forecast slope

D23

Queen's Gambit Accepted

OLS trend slope

-0.4964 pp/month

Fastest rising by 3-month MA

A95

Dutch Defense: Stonewall Variation

Latest 3M moving-average acceleration

+3.00 pp (last 3M)

Win Rate Forecasts

Win Rate Trends

The strongest signals are best read alongside uncertainty bands to separate stable trends from short-term noise.

Engine Delta

Engine vs Human Gap

These outliers flag openings where practical play diverges most from theoretical evaluation.

ECO Family Win Rates

ECO Family Patterns

Category-level heatmap patterns continue to show non-uniform performance over time, with month-to-month variation indicating shifting practical preferences across ECO families. Use 50% as parity baseline: persistent deviations indicate family-level practical tilt.