Reading a track record without fooling yourself

Most crypto signal services lead with a track record. Almost none of them give you the context to judge it. Independent reviews in 2026 found that advertised win rates commonly run 10–20% above third-party-tracked performance. Here is how to read any track record - ours included - without fooling yourself.

Win rate is the most over-weighted number

A 70% win rate sounds great until you learn the average loss is three times the average win. Win rate means nothing without average win vs average loss and position sizing. A 45% win rate with disciplined risk/reward can compound beautifully; an 80% win rate that occasionally gives it all back cannot.

Demand the losers

The single fastest scam filter: does the record show losing trades? A sequence of cherry-picked winning screenshots is marketing, not a track record. Look for a complete, timestamped log - every signal, recorded before the outcome was known, with losses sitting right next to the wins. That is the standard we hold ourselves to on the track-record page.

Sample size and survivorship

• Sample size - 20 trades prove nothing. Edges only become visible over hundreds of trades across different conditions.
• Survivorship bias - beware services that quietly delete bad calls, restart accounts after a blow-up, or only show their best strategy of many.
• Regime coverage - a record that only spans a bull run tells you nothing about how the system behaves when the market turns.

Drawdown is the number that tells the truth

Maximum drawdown - the worst peak-to-trough loss - is the figure that survives the marketing. It tells you what holding the strategy actually felt like and whether you could have stuck with it. We weight Calmar ratio (return ÷ max drawdown) above raw return precisely because surviving the bad periods is the whole game.

Demo vs live, and why we say so

We are transparent that our current numbers come from a demo accountin a validation phase - clearly labelled everywhere they appear. A demo record run with real strategy logic on live market data is a far higher standard than a backtest, but it is not the same as live capital, and we won't pretend otherwise. When we transition to a live fund, both records stay published side by side.

The verification bar

The 2026 best-practice standard is simple: if a track record isn't independently verifiable, treat it with deep scepticism. Building toward a verifiable, exportable public ledger is on our roadmap - see about. Until then, the protection we can offer is completeness: every signal, logged before the result, nothing hidden.