Backtesting has an honest answer and a marketing answer. The marketing answer is "yes, test your strategy on history and trade with confidence." The honest answer, for a discretionary Indian options or intraday trader, is "mostly no, and here's what to do instead."
What backtesting is, and what it's actually good for
Backtesting means running a set of trading rules against historical data to see how they would have performed. For a fully mechanical, rule-based system, this is genuinely useful: if your rules are precise enough to be coded, you can test them over years of data, check the sample size, and get a real sense of whether the idea has structure.
The key word is mechanical. Backtesting answers one question well: "would these exact, unambiguous rules have worked?" It works to the exact degree that your strategy is rule-based and your data is clean.
Why it deceives discretionary options traders
Most retail intraday and options trading is not mechanical. It's discretionary: you read the chart and decide. And the moment discretion enters, backtesting starts lying to you, in three specific ways.
Hindsight bias. Scrolling back through a chart, every setup looks obvious. You see the breakout that worked because you already know it worked. In real time, that same candle was one of five ambiguous possibilities. A manual backtest quietly tests your hindsight, not your foresight, and hindsight has a suspiciously high win rate.
It can't test your psychology. A backtest assumes you took every signal, held every stop, and sized every trade perfectly. But the thing that actually loses you money isn't the strategy, it's revenge, FOMO, and the stops you move. None of that appears in a backtest. You're testing a disciplined robot version of yourself that doesn't exist at 11 AM after a loss.
Options make it nearly impossible. To backtest an options trade properly you need historical option premiums, not just the underlying's price, plus the implied volatility at the time, plus realistic fills on a spread that's often wide. Most retail traders have none of this. Backtesting the Nifty's movement tells you almost nothing about what your specific option would have done, because theta and IV sat in between.
The trap even algo traders fall into
Even with clean rules and data, there's curve-fitting: adding conditions until the strategy looks perfect on past data. "Buy when RSI is below 30 and it's a Tuesday and the previous candle was green" might show a flawless backtest and fail instantly live, because you fit the rules to the noise of history rather than a real pattern. The more you optimise a backtest, the less you should trust it.
What actually validates your edge
For a discretionary options trader, the honest test isn't historical, it's forward. Trade your idea small, in real conditions, with real fills and your real psychology, and record what happens. After 30 to 50 real trades of a specific setup, you have something a backtest can never give you: evidence of how the setup performs when you trade it, not when an idealised robot does.
Your forward tradebook is the real backtest. It already includes the costs, the slippage, the theta, and the behaviour, the exact things a historical backtest leaves out. Use backtesting to sketch an idea if you like. Use your actual tradebook to find your edge.
SubTrades reads your forward tradebook, the real trades, real fills, and real outcomes, and shows you how each setup actually performs once theta, costs, and your own behaviour are in the mix. Auto detection of psychological patterns. Import your Zerodha, Dhan, Upstox, or Angel One trades and test your edge against reality. Free during the founding beta.