New here? Start with this
What makes a strategy good?
Not every set of rules is worth trading. This is about the rules themselves — before you ever test them — so what you bring to a backtest is worth testing in the first place.
What makes a strategy good
An unambiguous entry
Two people reading the rule should reach the same decision. If it needs a feeling to apply, it's not a rule yet.
A stop decided upfront
Every trade needs a predefined "I'm wrong" point set before you enter — not one you find in the moment, under pressure.
One real edge, not five vague conditions
Stacking conditions until only two trades a year qualify isn't precision — it's overfitting to the handful of examples you eyeballed.
Repeatable across time
If it only works on the exact days you designed it around, it's curve-fit to history, not a strategy you can trade forward.
How to build one
1
Start with one idea
A single observation about how price behaves — a breakout, a pullback to a moving average, a reversal pattern.
2
Write the entry rule first
The exact condition that would make you take the trade, stated plainly enough that nothing is left to interpretation.
3
Add the stop before anything else
Decide where you'd be proven wrong now, not after you're already in the trade and reluctant to admit it.
4
Add a confirm and a skip
Confirm filters out noise around the entry; skip keeps you out of conditions the strategy was never built for.
5
Test it before you trust it
This is where Backtesting 101 picks up — run it against history before you run it with real money.
Common mistakes to avoid
Adding rules until it fits the past perfectly
A rule set tuned to match every winning trade in your sample will fit that sample — and nothing else.
Discretionary "feel" wording
"If it looks strong" isn't a rule — it's a mood, and it'll change under pressure.
No stop, or a vague one
"Somewhere below" isn't a level you can act on when the trade is moving against you.
Copying a strategy without knowing why each rule is there
A rule you don't understand is one you won't follow the first time it's uncomfortable to follow.
A worked example
9/20 EMA Crossover
Every rule below exists to answer one question — why it earns its place:
Enter9-period exponential moving average crossing above the 21-period one — the core signal this strategy trades.
ConfirmA close beyond the crossover point, not just a touch — a wick through it happens constantly and means nothing alone.
StopThe most recent swing low is your invalidation point if the crossover fails.
Skip whenWhen the two EMAs keep crossing back and forth in a few candles, there's no real trend to trade yet.