The Win Rate Illusion
An EA with a 90% win rate blows accounts. An EA with a 40% win rate prints money.
You're measuring the wrong number. Here's the trap: your backtest shows 180 winning trades and 20 losers. Feels great. But if those 20 losers average -3% each and those 180 winners average +0.2% each, your math is: (180 × 0.002) - (20 × 0.03) = 0.36 - 0.60 = -0.24. Negative. You're watching the one metric that feels good (accuracy) while the metric that kills you (expected value) runs silent.
This is how account blowups happen with beautiful backtests.
Expected Value Is the Only Metric That Matters
Expected Value (EV) = (Win Rate × Average Win) - (Loss Rate × Average Loss)
This one formula separates traders who survive from traders who get liquidated. It doesn't care about how many trades you win. It only cares about the math.
Let me show you the difference:
- Scenario A (90% win rate): 9 wins at +0.5% each, 1 loss at -4%. EV = (0.9 × 0.005) - (0.1 × 0.04) = +0.0005 per trade. Barely positive.
- Scenario B (50% win rate): 5 wins at +2% each, 5 losses at -0.8%. EV = (0.5 × 0.02) - (0.5 × 0.008) = +0.006 per trade. 12x better.
Scenario B makes money. Scenario A will eventually wipe you. But Scenario A is the one traders chase because the win rate number is beautiful.
Why Backtests Hide the Real Problem
MT5 shows you win rate on every backtest report. It's right there, big and shiny. Traders optimize for what they can see.
Your backtest shows 89% win rate and your brain says "profitable." Your backtest shows 40% win rate and your brain says "broken." But the 40% win rate with proper risk-reward has beaten the 89% win rate with bad risk-reward every single time.
The professionals know this. They look at the equity curve shape, not the win percentage. They look at maximum drawdown, not average win. They look at profit factor (gross profit ÷ gross loss), not accuracy.
Most traders optimize for the metric they can see instead of the metric that matters. That's the only reason they fail.
One Losing Trade Erases Fifty Winning Trades
This is the psychology that kills accounts.
You set your EA to win on 90% of trades. Feels bulletproof. You watch it go 18-1, 35-2, 47-1. Dopamine hits. But that one loss? If your risk-reward is inverted (small wins, big losses), that single trade erases 30-40 winning trades.
Now you're underwater. The EA is still "working"—still winning 90% of trades. But the account is dying because the one trade type that dominates is the loser. Flip the ratio. Small win rate, big wins. Those traders sleep because they know the math: 50% losers, but losers are small. Ten wins pay for fifteen losses.
How to Build Real Edge
Stop chasing win rate. Start here:
- Define your risk per trade. Usually 1-2% of account per trade. Non-negotiable.
- Set your minimum risk-reward ratio. 1:2 minimum (lose $1 to make $2). 1:3 is better. 1:5 is professional.
- Calculate expected value before you trade. If EV is positive with your real numbers, you're in. Negative? Your strategy doesn't trade yet.
- Backtest for equity curve shape, not win rate. Smooth curve up beats spiky curves with drawdown spikes.
- Stress test the drawdown. If the EA loses 30% of your account in a bad month, it's not an edge—it's a lottery ticket.
This is why most DIY traders fail. They skip step 3. They build an EA, see it win 80% of trades, and deploy it. Six months later they're confused why their account is at 40% when the backtest showed 95% accuracy.
The professionals build EAs with expected value locked in from day one. Alorny builds custom MT5 Expert Advisors with these metrics as the foundation. Every backtest report includes the real numbers: expected value, risk-reward ratio, and drawdown under live market stress. No hidden losses. No win-rate theater.
The Cost of Measuring the Wrong Thing
Here's what being wrong about this costs:
Deploy a negative-EV EA with 2:1 leverage because the backtest showed 85% accuracy, and the expected cost is your entire account. Not "might be." Will be. It's statistical certainty over enough trades.
The traders who win are the ones who can watch their EA win only 45% of trades and sleep soundly. Why? Because every trade has a 1:3 risk-reward ratio. Math beats hope. Math beats luck. Math beats the ego boost of 80+ accuracy.
This is the edge that separates the 5% of traders who compound wealth from the 95% who blow up.
Key Takeaways
- Win rate is a vanity metric. Expected value is the only metric that predicts account growth.
- A 50% win rate with 1:3 risk-reward beats a 90% win rate with 1:0.8 risk-reward every time.
- Professional traders optimize for profit factor and drawdown, not accuracy.
- One losing trade can erase 30+ winning trades if your risk-reward is inverted.
- If you're chasing win rate instead of expected value, your account blowup is not a matter of if—it's when.