Mechanism, with the instrument and timeframe removed: after the narrowest-range day of the last 7 (NR7), a daily CLOSE below that NR7 day's low is a stop-run into the liquidity pocket under a tight range. Buy it and hold for a fixed number of days. This is the OPPOSITE sign of the Crabel NR7 breakout that was abandoned earlier. That design assumed continuation; the pre-study shows the downside break of a compression day reverses. Universe: the 9 Bybit USDT linear perps with at least 4 years of Bybit daily history: ETH, SOL, XRP, DOGE, ADA, LINK, AVAX, BNB and LTC. None were dropped after the fact, and all 9 were individually positive. BTC was excluded because of corpus concentration. Each name is traded independently as an event panel, not a neutral book. PRE-STUDY (Bybit 2021-11 to 2026-09, events non-overlapping per name, net of 0.11% RT). NR7 downside-break long, 3-day hold: n=785, mean gross +1.24%, median +0.78%, win rate 57%, positive in 5 of 6 years including 2022; only 2025 was negative (-0.48). The CONTROL (any close below the prior day's low) earned +0.20% gross with calendar Sharpe 0.13. The NR4/7/10 x 2-5 day surface is smooth, with calendar Sharpe 0.43-1.22. Capped book (20% per event, max 5 concurrent): 694 trades, net +1.09%, PF 1.56, Sharpe 1.17, DD 32.6%. A -10% intraday stop cut Sharpe to 0.80, so there is no tight stop. Sizing is 12.5% per leg to bring DD toward ~20%.