i What this calculator does
Fifty trades at a 60% win rate feels like proof of skill. Statistically it often is not: a coin flip produces 30 or more wins out of 50 about one time in twelve. This calculator applies the test a fund's investment committee would apply to your record. From the number of trades, the win rate and the average win and loss, it computes the expectancy per trade, the t-statistic of that expectancy against zero, the probability that a zero-edge process would have done as well, and the number of trades you would need at the same statistics to reach the conventional threshold.
It is not a verdict on whether you can trade; it is a statement about how much the evidence so far can carry. It belongs with the expectancy calculator and the article on backtesting.
This calculator is for educational purposes only. Results are estimates and may vary depending on market conditions, spreads, commissions, platform settings, and exchange rates. It should not be considered financial advice.
What the test actually measures
Every trade is a draw from a distribution with some average, the expectancy, and some spread. The question is whether the average you observed could have come from a distribution whose true average is zero. The t-statistic answers it: it divides your expectancy by its standard error, which shrinks with the square root of the number of trades. A t of about 2 means a zero-edge process would match your record only about 2.5% of the time, which is the threshold most quantitative funds use before they allocate capital to a strategy.
Why most records fail it
Trading results have a large spread relative to their average. A strategy with a 55% win rate and a 1.3 reward-to-risk ratio has an expectancy of about 0.27 R per trade against a standard deviation of about 1.1 R, so it needs roughly 65 trades to reach t = 2, and that assumes the statistics are honest and stable. A 52% win rate at even money needs over 2,000 trades. This is why backtests with hundreds of trades are taken more seriously than live records with dozens, and why a strong first quarter proves very little.
How to use the result
A record that has not reached the threshold is not a failure, it is an unfinished experiment, and the right response is to keep risk per trade small, as the risk per trade calculator suggests, until the sample is large enough. A negative expectancy is different: more trades only confirm it. The chart shows how the probability of luck falls as the sample grows at your current statistics, which tells you how long the experiment has to run.
Spotted an outdated figure? If a rate, limit or formula in this calculator no longer matches its source, email us the page and the source. A material correction is dated on the Updates Log.
Frequently asked questions
Why does the test use expectancy rather than just the win rate?
Because a 40% win rate with large wins beats a 70% win rate with large losses. Expectancy combines the win rate with the size of wins and losses, which is what decides whether the account grows. The win-rate-only test is shown separately for comparison.
What does a p-value of 10% mean?
That one in ten traders with no edge at all would show a record at least as good as yours by chance. Convention treats 5% as the threshold for taking a result seriously and 1% as strong evidence.
My record is 20 trades with 75% wins. Why is the test unconvinced?
Because 20 is a small sample. Fifteen wins out of twenty happens to a coin flipper about 2% of the time, which sounds rare until you remember how many traders there are. Twenty trades cannot distinguish a genuine 60% edge from a lucky 50%.
Does the test assume anything unrealistic?
It assumes trades are independent and that your win rate and average sizes are stable. In reality traders change size and strategy, and wins cluster, which makes real records noisier than the model. The honest reading is that the real evidence is weaker than the number shown.
Can I enter R-multiples instead of rand?
Yes. Enter the average win and loss in R-multiples (1 R = your risk per trade) and the expectancy and standard deviation come out in R as well. The t-statistic is the same either way.
Why does the chart show 800 trades?
To show how slowly the probability of luck falls. Doubling the trades only reduces the standard error by about 30%, so proving a small edge takes a very large sample.
Want to embed this calculator on your site?
Embed this calculator
Free to use on your own site. Copy the code below and paste it where you want the calculator to appear.
<iframe src="https://www.tradeanswers.co.za/embed/luck-or-skill-calculator.html"
width="100%" height="720" style="border:0;max-width:520px"
title="Luck or Skill Calculator - TradeAnswers" loading="lazy"></iframe>
<p><a href="https://www.tradeanswers.co.za/luck-or-skill-calculator.html">Luck or Skill Calculator</a> by <a href="https://www.tradeanswers.co.za/">TradeAnswers</a></p>