Strategy Backtest Baseline — Metrics & Procedure (V2.7)

Scope: define what to measure and how to run a baseline pass. No performance claims.

1) What to measure (minimum)

Trade counts

• Total trades
• Trades per day/week
• Long vs short split

Outcome metrics

• Win rate
• Loss rate
• Profit factor (if strategy outputs PnL)
• Expectancy (if available)

Risk metrics

• Max drawdown (absolute + %)
• Avg MAE/MFE (if tracked)
• Avg bars in trade

Execution-truth metrics (custom)

V2.7 includes a TP1-before-SL counter model (separate from broker PnL):

• Signals
• Resolved outcomes = TP1 hit before SL vs SL hit before TP1
• Tie rule: if TP1 & SL hit same bar ⇒ count as loss (configurable)

Record:

• TP1_before_SL_winrate = wins / (wins + losses)

2) Baseline test matrix

Run each symbol/timeframe across identical dates (example):

• Timeframes: 5m, 15m, 1h
• Sessions: filter ON vs OFF
• Mode: Aggressive vs Safe
• Confirm-on-close: ON (baseline)

3) Procedure (TradingView)

1. Open the V2.7 Strategy script:
   • EnvoTrades-Smart-Signals-V2_7-futuristic-score.SHIP.strategy.pine
2. Set Bar magnifier policy (if applicable) consistently.
3. Set process_orders_on_close=true (already set in script).
4. Fix the date range.
5. Export results:
   • Strategy Tester → Performance Summary
   • List of Trades (CSV if available)
6. Capture screenshots of HUD stats for the same endpoints.

4) Reporting template

For each (symbol, timeframe, date range, mode):

• Total trades
• Net profit (if used)
• Max drawdown
• Profit factor
• TP1-before-SL winrate
• Notes: any blocking reasons dominating (SESSION/RANGE/RSI/BOS)

5) BLOCKED / NEED

NEED (to run an actual baseline pass):

• Symbols to test
• Date ranges to standardize
• Session settings preference (NY/Asia/London)
• Whether to include spreads/commissions assumptions