The 3 backtest metrics I actually trust (and 2 I've stopped using)

Most backtest stories offer you 5 key metrics. I’ve stopped trusting 2 of them.
Here is the hierarchy I exploit to separate "backtest magnificence" from strategies that really survive the 2026 Indian markets—and why the numbers most individuals brag about are sometimes simply noise.

The 3 I Trust (The Sign)

1. Profit Factor: Gross income divided by gross losses. A Revenue Factor of 1.5 means your winners generated 50% greater than your losers. The Actuality: Whereas 1.0 is "break-even," I gained't contact a technique under 1.5 anymore. Why? Because after the 2026 STT hikes and slippage, a 1.2 Profit Issue often evaporates. You need a "buffer of security" to outlive contact with the stay change.
2. Most Drawdown (MDD): The deepest hole your fairness curve fell into (peak-to-trough). The Actuality: This is the "Psychology Metric." Robert Pardo (the authority on strategy analysis) famously noted that real-world drawdown is usually 1.5x to 2x worse than the backtest. In case your backtest exhibits a 10% dip, ask yourself: "Might I deal with a 20% hit to my precise bank account without turning the bot off?"
3. Sharpe Ratio: Return per unit of danger. The Actuality: Above 1.zero is "okay," but 1.5+ is the place the magic happens. I’ve seen methods with triple-digit returns that had a Sharpe of 0.4. That’s not a technique; that’s a lucky coin flip with excessive stress.

The two I Ignore (The Vainness)

1. Win Fee: That is the last word "newbie lure." A technique with a 75% win price sounds invincible. But if your common loss is 4x larger than your common win, you're quietly bleeding out. Win price with out Revenue Factor is simply vainness.
2. Complete Return: By itself, that is meaningless. Technique A: 60% Return | 40% Drawdown Strategy B: 45% Return | 8% Drawdown Technique B is the skilled selection every single time. It's the "smoother experience" that retains you in the recreation long sufficient to truly see those returns.

The Backside Line
I’m not saying Win Fee and Complete Return are useless—they are context. However they aren't the signal.
In my engine room at FlyTradr.com

, the hierarchy is straightforward: Profit Issue > Max Drawdown > Sharpe.

What’s your "Lead Metric" once you take a look at a report? Do you prioritize the graceful fairness curve (Sharpe) or the uncooked effectivity (Revenue Factor)?

https://preview.redd.it/j8yh3z4j7ixg1.png?1672&format=png&auto=webp&s=07ae5ce907791b7bd4a2aa55215dfbf30b93f8ec