Signal sellers often quote an accuracy figure such as "85% winners". On its own that number says almost nothing. It does not say how many signals it covers, over what period, whether losing signals were left out, or what payout applied. This guide explains how to read an accuracy claim and how to test one yourself.
Accuracy Is Not Profit
A signal is right or wrong. A trade is profitable only if the share of winners beats the break-even rate for the payout. Break-even is 1 divided by (1 plus the payout as a decimal): about 55.6% at an 80% payout and about 52.6% at 90%. A source that is right 55% of the time loses money at an 80% payout. Work through your own numbers with the payout calculator and the win rate calculator.
Sample Size Matters
Twenty signals with 17 winners looks impressive, but the range of true rates that could produce that result is very wide. Small samples are dominated by luck. See strategy sample size and statistical significance for the idea, and use the signal accuracy tracking tool to see the uncertainty around your own record.
How Claimed Figures Get Inflated
Common ways a headline figure can mislead include:
- Counting only the signals that were posted as winners, or deleting losing ones.
- Counting a signal as a win if any of several entries or martingale steps recovered, which hides the size of the losses. See martingale risk explained.
- Using a backtest or cherry-picked screenshots instead of a live, time-stamped record. Screenshots are easy to alter.
- Ignoring the payout, the delay between the signal and your entry, or what happens on ties.
Ask whether the source shows a complete, time-stamped record that you could verify independently. If it does not, treat the figure as unproven. The signal provider verification guide and red flags checklist set out what to look for.
Testing a Source Yourself
The only reliable approach is your own record. Follow every signal in real time on a demo account or on paper, log each one including those you would have skipped, and keep going until the sample is large enough that the range around the hit rate is narrow. Compare that range with the break-even rate. Even a good result does not guarantee the next signals will behave the same way.
Limits
This page is educational and does not rate or recommend any signal provider. Nothing here is advice, and signals can lose money. See also are binary options signals reliable? and fake trading signals scam warning.
Frequently Asked Questions
What is a good signal accuracy rate?
There is no universal number. It must be compared with the break-even rate at your payout, and with the uncertainty from the sample size.
Can a provider show verified results?
Only a complete, time-stamped record that you can check independently is meaningful. Screenshots alone are weak evidence.
Does a high accuracy claim mean low risk?
No. A claimed rate does not remove the risk of loss, and may hide large losses behind frequent small wins.
Why test on demo first?
So that you can judge a source without risking money, and see how delays and execution affect results.