New here? Start with this
What's a backtest?
You replay real historical candles and make the calls you'd make live — entries, exits, when to sit out — then see how it would've actually played out. No real money, no real risk, same decisions.
This covers testing process. For the rules themselves, see what makes a strategy good.
How to run one here
1
Pick a strategy
Choose one of ours, one you've forked, or skip it for a free replay with no rules attached.
2
Set the scene
Symbol, date range, starting balance — whatever conditions you want to test the strategy under.
3
Step through candle by candle
React only to what's on screen — enter, hold, exit — exactly like you would live.
4
Review the result
See how the calls played out, trade by trade, and decide what you'd change.
What makes a good backtest
Enough reps to mean something
Three trades tell you nothing. Run the strategy across enough setups that a pattern — not luck — is what you're seeing.
The same rule, every time
If you enter differently on candle 4 than you did on candle 40, you're not testing the strategy anymore — you're testing your mood.
No peeking ahead
Only react to candles you've actually reached in the replay. Knowing what happens next and pretending you didn't defeats the whole point.
Review the losses too
A good backtest ends with you looking at what went wrong, not just tallying the wins.
A worked example
9/20 EMA Crossover
Go long when the fast EMA crosses above the slow EMA, exit/flip when it crosses back below.
EnterFast EMA (9) crosses above Slow EMA (20)
ConfirmCandle closes clearly beyond the crossover, not just a wick
StopJust below the most recent swing low
Skip whenThe EMAs are flat or whipsawing with no clear trend