The Truth About "87% Accuracy" — What It Means and What It Does Not

The Short Version

Wind Indicator V1.6 reports 75–87% accuracy across markets and timeframes. That number is the percentage of historical signals that reached a predefined take-profit level before hitting the predefined stop. It is measured on out-of-sample historical bars, not on optimized training data. And it is not the most important number to trade on.

The rest of this post is what the headline claim actually means, and how to interpret it when you are evaluating any signal product — not just ours.

How Accuracy Is Measured

For each historical entry arrow:

1. The entry price is recorded at bar close.
2. A stop level is defined using the system's ATR-based logic at the time of entry.
3. A take-profit level is defined at the same time.
4. The subsequent price path is traced. If TP is hit before SL, the signal is a win. If SL is hit first, a loss.
5. Signals where neither is hit within the expiration window are excluded.

The 75–87% range is the win rate across this definition, segmented by symbol group and timeframe. Crypto majors on 4h sit near the top of the range. Low-volume altcoins on 15m sit near the bottom, which is why the system does not publish signals for them.

Why The Range Is 75–87%, Not A Single Number

Anyone publishing a single accuracy number across all markets is either averaging over everything equally (which hides weak segments) or cherry-picking a best case. The honest answer is that accuracy varies by:

• Liquidity — major pairs outperform minors. Same logic, cleaner price action.
• Volatility regime — the system performs best in trending conditions, weakest in chop.
• Timeframe — higher timeframes have fewer but cleaner signals.
• Session — US/Europe overlap beats Asia session for most forex pairs.

Publishing a range instead of a point estimate is the more defensible claim.

Why Accuracy Alone Is The Wrong Metric

A signal can be 90% accurate and still lose money. Example: 9 wins of +1R each, 1 loss of -15R. 90% accuracy, net -6R.

The metrics that actually determine profitability are:

• Average win (R multiple). How much you make when you win, measured in units of risk.
• Average loss (R multiple). How much you lose when you lose. If stops are respected, usually -1R.
• Expectancy. (Win rate × average win) − (loss rate × average loss). This is the real edge.

Wind Indicator's published stop-and-target logic yields roughly 1:1.5 to 1:2 reward-to-risk on typical setups. Combined with a 75–87% win rate, expectancy stays comfortably positive — but the driver is the combination, not the accuracy alone.

Why 100% Accuracy Would Be A Red Flag

If a system published 95%+ accuracy, the right first question is not "wow, how?" — it is "what is the stop distance?" A system can engineer near-100% accuracy by placing stops absurdly wide, so almost every trade eventually reaches target. Then a single loss wipes out dozens of wins.

75–87% with honest stops beats 95% with invisible stops every time.

What You Should Actually Track

When you start trading Wind Indicator (or any signal system), track two things in your own journal:

1. Signals taken vs. signals missed. Most underperformance versus the headline stats is execution slippage — traders skipping signals they found uncomfortable.
2. Your own R distribution. Over 30–50 trades, your expectancy should converge toward the system's expectancy, not sit below it.

If your live results diverge significantly from the published range, the problem is almost always in execution, not the signal. That is diagnosable. The fix is usually "take every signal for the next 20 entries, regardless of how you feel about them," and compare.

The Honest Bottom Line

Accuracy is a headline number. Expectancy is the trading number. Wind Indicator V1.6 publishes the first because marketing requires it, and the second is what the math requires. Both are real. Neither is a guarantee.

Markets change. Volatility regimes shift. Every honest signal product should be evaluated, and re-evaluated, against your own live execution — not against the brochure.

Try it yourself with a 3-day free trial → Get TradingView access