How Many Trades Do You Need to Backtest a Strategy?

Enough that your result could not plausibly be luck — which is usually far more than the “100 trades” rule of thumb suggests. To separate a genuine 55% win rate from a 50% coin flip at 95% confidence takes roughly 500 trades. A much stronger edge needs far fewer. The weaker your edge, the more evidence it takes to prove it exists.

That last sentence is the part traders get backwards. The strategies people are least sure about are exactly the ones they test least.

Why “100 trades” became the standard answer

It’s a reasonable-sounding number: big enough to feel serious, small enough to be achievable in a few weekends of replay. It also has no statistical basis whatsoever.

The number of trades you need is not a fixed constant. It depends on how big the effect is that you’re trying to detect. Proving that a coin lands heads 90% of the time takes a handful of flips. Proving it lands heads 52% of the time takes thousands. Trading strategies live at the difficult end of that spectrum, because real retail edges are usually small.

The actual numbers

Here’s what a 55% observed win rate means at different sample sizes, using 95% Wilson confidence intervals — the standard method for proportions, and the one that behaves properly at small sample sizes:

Trades	Observed win rate	95% confidence interval	Rules out a coin flip?
20	55%	34% – 74%	No
50	55%	41% – 68%	No
100	55%	45% – 64%	No
200	55%	48% – 62%	No
500	55%	51% – 59%	Yes

Read the 100-trade row again. You have done a hundred replayed trades, you are winning 55% of them, and the honest summary of your evidence is “my true win rate is somewhere between 45% and 64%.” That range contains comfortably profitable strategies and comfortably losing ones. You have not yet learned whether your strategy works.

At 200 trades you still can’t quite exclude 50%. It takes around 500 before the interval clears break-even and you can say the edge is real rather than probable.

Bigger edges need less evidence

This cuts the other way too, and it’s the genuinely useful part:

• A strategy winning 65% of trades separates from a coin flip in well under 100 trades.
• A strategy winning 52% may need several thousand.
• A strategy whose edge comes from payoff rather than win rate — 40% winners at 3R — needs a different calculation entirely, because the thing being measured is no longer a proportion.

So the honest answer to “how many trades do I need?” is: enough that the confidence interval around your result no longer includes break-even. That is a moving target, and it is the right target.

Win rate is the wrong metric anyway

Win rate appears in every sample-size example, including this one, because proportions produce the cleanest maths. It is not the metric you should be acting on.

A 40% win rate at 3R is an excellent strategy. A 70% win rate at 0.5R loses money. Win rate on its own cannot distinguish them, which makes it the single most over-weighted number in retail trading.

Use expectancy — the average result of one trade, in R or currency. It incorporates both how often you win and how much you win, which is what actually determines whether the account grows. Expectancy has wider variance than a simple proportion, so if anything it needs more data before you can trust it, not less.

The mistake that quietly destroys sample size

Everything above assumes you’re evaluating one number over your whole history. The moment you start breaking results down, your effective sample collapses.

Split 60 trades across five trading sessions and you have twelve trades per bucket. Twelve. At that size, the difference between your “best” and “worst” session is almost certainly noise — and it will look completely convincing on a bar chart.

This is how traders end up abandoning perfectly good setups because of a losing streak in one arbitrary slice of their data. Every filter you apply — by pair, by session, by weekday, by setup — divides the evidence again.

Practical rules:

• Look at the trade count in each bucket, not just the percentages
• Be suspicious of any bucket below a few dozen trades
• Ask whether the pattern survives removing your two best and two worst trades
• Prefer one large conclusion over five small ones

What this means for how you backtest

The uncomfortable implication is that meaningful backtesting takes longer than most people plan for. Several hundred trades is a real commitment — which is exactly why the process needs to be efficient and, more importantly, honestly recorded. Five hundred trades you half-remember are worth less than fifty you logged properly.

Two things make that sample achievable:

1. Speed. Manual replay with proper speed controls and keyboard shortcuts turns months of chart time into hours.
2. Automatic recording. If logging each trade is manual work, you will stop doing it around trade 40 — right before the data starts being useful.

BacktestFlow is built around both. Replay won’t advance past an unjournaled close, so the record is complete by construction, and around 35 statistics build themselves as you go. Flow Coach, the analysis layer, runs significance tests before making any claim — it measures confidence intervals and minimum group sizes, so it won’t tell you a session is your best one on twelve trades.

The short version

There is no magic number. The sample you need depends on the size of the edge you’re trying to prove, and small edges need a lot of evidence: roughly 500 trades to establish a 55% win rate as real, far fewer for a strong edge, far more for a marginal one. Track expectancy rather than win rate, watch what happens to your sample every time you filter, and stop treating 100 trades as a finish line.

---

This article is educational and is not financial advice. Trading involves risk, and past performance — including backtested performance — does not guarantee future results.