Skip to main content

​
What strategy improvement really means

Improving a strategy is not about:
  • maximizing backtest returns
  • chasing the best parameter set
  • reacting to every bad drawdown
It is about:
  • improving robustness
  • aligning behavior with intent
  • reducing unnecessary risk
  • making outcomes more consistent
Good iteration is conservative, not aggressive.

​
The iteration mindset

Every change should answer one question:
What problem am I trying to fix?

​
A safe iteration loop

1

Start from a stable baseline

Keep a version you trust as a reference.
2

Change one thing at a time

Multiple changes destroy traceability.
3

Re-evaluate behavior, not just metrics

Check trades, exposure, and drawdowns first.
4

Test across ranges

Ensure improvements are not regime-specific.

What to change first

Overfitting & guardrails

Parameter ranges

When not to optimize