Verifying signal accuracy before live trading is a step most beginners skip, jumping straight from a first impression to betting real money. Without a systematic check, there's no way to tell a source with a real edge apart from a set of random forecasts that can look equally successful over a short stretch.
Why a short test doesn't show the real picture
A few winning trades in a row say nothing about signal source quality: pure random guessing of direction gets the right result about half the time with no system at all. To tell a real edge apart from chance, you need a statistically meaningful sample — at least 100 trades, ideally 150–200, collected across different days of the week and times of day.
How to organize a demo account test
A proper test requires discipline that is often missing when testing on a live account because of emotion. The recommended sequence:
- Pick a fixed position size for the whole test period so changing the amount doesn't distort the statistics.
- Log every signal in a separate table: time, asset, direction, result.
- Don't skip "inconvenient" signals or cherry-pick only the ones you like — that distorts the final win rate.
- Test the source for at least 2–3 weeks, covering different market conditions.
- Separately flag trades made during news events versus calm market periods.
How to calculate a real win rate
Win rate is the ratio of winning trades to total trades over the test period, without excluding failed attempts from the statistics. It's important to count based on actually acting on the signal at its stated time, not on how the trade would have looked with a more convenient entry. This "fitting to the outcome" is a common mistake that creates a false sense of high source accuracy.
What metrics matter besides win rate
Win rate alone doesn't give the full picture, especially with varying profit and loss sizes per trade:
- Profit factor — the ratio of total profit to total loss over the period.
- Maximum consecutive losing streak — shows how much capital is needed to survive a drawdown.
- Consistency by time of day — does the signal perform the same way morning, afternoon and evening.
- Dependence on news — does accuracy drop during major data releases.
Common mistakes when verifying signals
Testing most often gets distorted by a handful of systemic mistakes: too small a sample, testing only during "convenient" hours, ignoring the delay between signal publication and actual entry, and subjectively deciding which signals to count. Any of these makes the resulting statistics useless for deciding whether to move to a live account.
How to simplify the verification process
Manual verification requires discipline and time that beginners often lack. An alternative is using a signal source with transparent, fixed logic that can be verified once and then trusted without daily manual tracking. This approach is built into the trading robot for Pocket Option — details on the features page.
Conclusion
Verifying signal accuracy before live trading is a mandatory step that protects the deposit from sources with no real edge. Collect a sample of at least 100–150 trades, calculate win rate and profit factor without adjusting the numbers, and only then move to a live account. For more on combining signals with your own analysis, see the article how to combine signals with your own market analysis, and the robot's plans are here.