Win Rate
Win rate is the most intuitive number, so it's everyone's first question: 'What's this signal's win rate?' But win rate is half the information. It tells you how many times you won — not how much changed hands on each win and each loss.
In numbers: take an account with a 90% win rate. Nine wins at +0.5% each make +4.5% — then one -10% hit. The total is -5.5%: a 90% win rate that loses money. Now take a 40% win rate: six losses at -1% each (-6%) and four wins at +3% each (+12%) net +6%. A low win rate can make money, and a high one can lose it.
Be extra careful, because win rate is an easy number to manufacture. Set a tight take-profit and skip the stop-loss, and the win rate soars on its own — every small uptick gets recorded as a win, while the failures are deferred as 'haven't sold yet.' If someone flaunts a high win rate without showing the risk-reward and the full equity curve, suspect exactly this trick.
Keep a baseline in mind too: on a pure up-or-down call, even a coin flip lands near a 50% win rate. So '55% win rate' may look impressive, but it's only a few steps from fifty-fifty. And whether that 55% came from enough cases (the sample size, N) decides whether it might just be luck.
So whenever you look at a win rate, always ask three things alongside it: the average size of wins versus losses (the risk-reward), how many cases produced the number (N), and whether anything is left after fees. Those three combined are the expected value (EV).
What the data actually shows
Barobara publishes the win rates of Bitcoin chart signals exactly as they are — what price actually did after each signal fired. Bottom line first: most signals sit near 50%, essentially a coin flip. You can see win rates by target on pages like RSI oversold (1h) or EMA golden cross (1d). These numbers are historical records, not predictions — and where the answer was fifty-fifty, that's exactly what we wrote. The full list is in the setup statistics.
Common misconceptions
"A signal group with a 90% win rate must be skilled?" A 90% win rate is easy to manufacture: set a very tight take-profit and never stop out. While they show off nine small wins, the one position with no stop-loss empties the account. Anyone who shows only a win rate while hiding the risk-reward and the equity curve deserves suspicion by default.
"A 50% win rate means breakeven?" Only when wins and losses are the same size. And since fees come out of every trade, in practice even 50% loses money.
FAQ
Q. What win rate makes a strategy good?
Win rate alone can't answer that. A 40% strategy can make money while a 90% one loses it. The bar is whether the expected value (EV) — win rate combined with risk-reward, sample size (N), and fees — is positive, and even that is only a past distribution, not a guarantee about the future.
Q. Can I just follow the signals with high win rates?
First check how many cases (N) produced that win rate, and the average size of wins versus losses. And a past win rate is exactly that — a past record. It's no guarantee the future looks the same.