Cumulative P&L — all live paper trades (virtual $50/position)
Performance by strategy arm
Trade blotter — real observed fills, entry lag measured
Strategy league — every named strategy, its verdict, and the evidence (honest fills: stops fill at breaching close)
Strategy universe — every backtested variant (t-stat vs mean return)
Moonshot tail hunt — CLOSED DEAD (deep-panel re-test)
P(graduate closes ≥5x) base 3.65% on
graduates; best first-hour feature lifts it only to 6.7% — need ~25% to profit after costs. Every tail-entry rule tested at t = −4.8 … −7.6.
Verdict pipeline
High-cap "exchange tier" grid — killed by the honesty test (optimistic t shown; all die under pessimistic fills)
Base rates — any random pump.fun launch (measured on 28,846 tokens)
Measured anomalies
Research timeline
Built from on-chain / public API data only · no assumptions · paper trading, no real capital at risk