Most trading education stops at “here's the pattern.” Ours doesn't — every concept comes with the honest question underneath it: does this survive six years of data after fees? When we know the answer, we show it.
Every strategy cohort we publish gets replayed nightly on fresh data — only trades entered after publication count, and the numbers publish win or lose. The page nobody else in this space dares to have.
Candle-close countdowns, funding timers and market sessions in Chrome's side panel — or on every new tab. No account, no tracking, zero network requests.
ChatGPT, Gemini, Grok, DeepSeek, Perplexity and Manus each wrote their best strategy. Identical prompt, first answer only, five hard gates on 6 years of data. The honest scoreboard.
Manus wrote us a textbook-clean trend system. Six years of out-of-sample data later: F — and not for the reason you'd guess. The definitive read on AI-generated strategies.
Connect Tessen's grader to Manus as a custom MCP connector — every strategy it writes gets the five-gate, out-of-sample treatment before it touches money.
Grade your first strategy, read the report, build an agent, paper trade it, and go live with approval mode — step by step.
Expectancy, the win-rate trap, R:R, position sizing, overfitting, fees — each concept illustrated with real numbers from our engine, not stock examples.
RSI 30/70, Connors RSI-2, MACD momentum, Bollinger both ways — the classics everyone teaches, run out-of-sample with real costs. Every grade has a permanent verify link.
The 12 patterns everyone trades, what each one is supposed to mean — and the honest note on what pattern-alone signals are worth after fees.
Head and shoulders, double tops, triangles, flags — structure patterns, their trigger levels, and how to test them instead of trusting them.
Safe API-key setup for Binance, Bybit, OKX, Bitget, Coinbase and Hyperliquid — least-privilege permissions, and how our withdrawal-permission check protects you.
Position size, liquidation price, funding cost, and the breakeven win-rate curve that explains why most high win-rate strategies make nothing. No signup.
The full user guide, grading methodology, AgentSpec schema, and the Developer API reference.
Everything here is education, not financial advice. Markets involve substantial risk of loss. When a page cites performance numbers, they come from our grading engine on historical data — with a public verify link wherever one exists — and historical results never guarantee future ones.