Backtesting trading strategies: test the rule, then test the assumptions.

Start with a rule that can be stated before the result

Define entry event, filters, session times, exit logic, and any position/risk assumptions before running the test. If rules change every time the equity curve looks bad, the test is becoming an optimization exercise rather than an evaluation.

Know the data and execution model

Trade Ideas says OddsMaker uses one-minute candle data rather than tick-by-tick execution and does not account for slippage. It typically tests around 40–64 trading days depending on the alert/strategy. Those details matter when interpreting fast intraday setups.

Avoid overfitting by limiting degrees of freedom

Every additional threshold creates another way to fit historical noise. Prefer rules with a clear market rationale, change one variable at a time, and test whether performance depends on a narrow parameter value that may not persist.

Use backtesting to eliminate, not certify

A poor historical result can help reject a rule quickly. A strong result should lead to more questions: does it survive different periods, symbols, volatility regimes, and reasonable execution assumptions? A backtest should earn a strategy the right to be paper tested—not the right to receive real capital automatically.

Connect the test to the live scanner

One advantage of OddsMaker is that it is built around Trade Ideas alert strategies. After testing, the same underlying alert logic can move into live scanning and then simulation. Keep a written record of any differences between the test configuration and the live scan.

Keep a versioned test log

Give every backtest configuration a version name and record the exact filters, event, session window, exit assumptions, and data period. When you change one rule, create a new version rather than overwriting the old result. This makes it possible to tell whether an improvement came from a meaningful hypothesis or from repeated tuning.

The log should also record why a change was made before the next test runs. That small discipline reduces hindsight bias and makes later paper-trading comparisons much cleaner.