Everything below is taught, sold, and repeated across trading social media. We implemented each one mechanically and ran it through the gates on ~1,900 real MNQ sessions. All eight failed. Bring flowers.
| Idea | What happened |
|---|---|
| "Wait for the pullback" (retest entries) | −35% profit. The best breakouts never pull back. |
| Volume confirmation on the breakout candle | Higher win rate, same money — a wash at best |
| Daily volume filter | Collinear with volatility — zero new information |
| Trend filter (only trade above the 20/50-day MA) | Deleted winners, kept losers; win rate unchanged |
| Fixed tight stops (5–15 pts) | Catastrophic: win rate 2–12% |
| Stop under the breakout candle ("structural stop") | Sharpe −2.2. Worse than random. |
| Move to breakeven quickly / trail tight | Noise tags breakeven, ejects future winners |
| VWAP-tag entries with tight stops | Fires constantly, dies constantly |
Half of them die by the same sword: stop scale. A 5-minute bar of NASDAQ noise is 20–40 points. Any stop inside that band — under the candle, a fixed 30 points, a fast breakeven — gets executed by randomness, not by being wrong. If your edge is built on a ~100-point structure, your stop must live at that scale. "Tight stop = small risk" is the most expensive sentence in retail trading; what it actually equals is high probability of a random loss.
The other half die by cutting the right tail. Pullback entries, early profit-taking, trailing stops — each converts a few big winners into many small outcomes. Feels safer. Costs everything, because in a ~50%-win strategy the entire profit lives in the tail you just amputated.
A note on the trader we borrowed three of these ideas from: his discretionary results may well be real. But whatever makes them real lives in the parts that don't systematize — level selection, timing feel, news awareness. The mechanical rules he states, applied mechanically, lose. Worth knowing before you copy anyone's playbook verbatim.