When traders first start evaluating EAs or systematic strategies, they gravitate toward the same few numbers: win rate, net profit, and the total number of trades. These metrics are intuitive — they seem to answer the obvious questions: Is it winning? Is it making money? Does it trade enough?

The problem is that each of these metrics, used in isolation, can actively mislead you. Strategies that fail all three of these "obvious" tests can be genuinely profitable. Strategies that ace all three can be slow disasters in disguise.

Here's what beginners get wrong — and what to focus on instead.

01
Misleading Metric
Win Rate — "I need to win more than I lose"

Win rate is the first thing most beginners look at, and for an intuitive reason: winning feels like evidence that you're doing something right. A strategy with a 70% win rate must be better than one with a 40% win rate, right?

Not remotely. As we've covered in detail previously, what matters is the combination of win rate and the size of wins vs losses. A 70% win rate with an average win of $50 and an average loss of $200 produces a negative expected value — the strategy loses money over time regardless of its win rate.

70%
A "70% win rate" strategy with 1:3 risk/reward (risking $300 to make $100) has an expectancy of −$20 per trade. Over 100 trades, that's a $2,000 loss — while "winning" 70 times.

Win rate without risk/reward context is not just incomplete — it's actively deceptive. A strategy designed to maximize win rate will systematically take small profits and let losses run, which is one of the most reliable ways to blow an account.

Focus on this instead

Expectancy = (Win Rate × Avg Win) − (Loss Rate × Avg Loss). This number alone tells you more about a strategy's viability than win rate ever will. Any positive expectancy is a foundation. Any negative expectancy cannot be fixed by discipline or position sizing.

02
Misleading Metric
Net Profit — "It made money, so it works"

Net profit as the primary evaluation metric ignores how that profit was generated. Two strategies with identical net profits can have completely different risk profiles, consistency, and likelihood of continuing to work in the future.

More dangerously: a strategy with impressive net profit that was generated through one or two outsized trades has a fundamentally different story than one that produced consistent, repeatable results. The former may never replicate. The latter has evidence of a process that works.

$12k
A strategy showing $12,000 net profit over 2 years sounds strong. But if $9,000 of that came from three trades during the COVID volatility spike in 2020, the remaining 97% of trades produced $3,000 — or about $1,500/year. A very different strategy.
Focus on this instead

Profit Factor (gross profit ÷ gross loss) and Recovery Factor (net profit ÷ max drawdown) paint a far more complete picture. They measure the quality of the process, not just the accumulated result.

03
Misleading Metric
Trade Count — "More trades = more data"

A high trade count feels like statistical confidence. More data points means more reliability, right? Often, but not always — and in the wrong context, high trade counts can mask serious problems.

High-frequency strategies that generate hundreds of trades per month look statistically robust. But if those trades are all small-target scalps with 0.2–0.5 pip profit targets, spread and execution costs eat into every trade. The net profit may look fine in a backtest using fixed spreads — but live, with variable spreads and even minimal slippage, the edge disappears.

Conversely, a low trade count strategy — say, 3–4 trades per month — might look statistically suspect but actually has a well-defined, carefully selected edge. The issue isn't volume; it's whether each trade has a genuine positive expectancy after realistic costs.

Focus on this instead

Average profit per trade after realistic spread and slippage. If this number is less than 3× your average spread, the strategy is exposed to execution risk that backtests don't fully capture. A strategy with 50 high-quality trades can be far more reliable than one with 5,000 marginal ones.

I spent my first year in algo trading chasing high win rates and impressive looking net profit numbers. Every strategy I deployed failed live. It wasn't until I learned to read profit factor and expectancy that I started understanding why some strategies actually work.

The Quick Reference: Stop / Start

Stop Optimizing For
  • Win rate in isolation
  • Maximum net profit
  • Highest trade count
  • Smoothest equity curve at any cost
→
Start Optimizing For
  • Expectancy per trade
  • Profit factor (1.4–2.2 range)
  • Risk-adjusted return (recovery factor)
  • Consistency across time periods

The shift from beginner to intermediate trader in systematic trading isn't about finding a better strategy. It's about learning to evaluate strategies honestly — and being willing to discard ones that look good by the wrong metrics.

The metrics beginners ignore are the ones that determine whether a strategy survives. The ones they obsess over are largely stories told after the fact.

Evaluate Your EA by the Right Metrics

EA Analyzer Pro highlights expectancy, profit factor, and recovery factor from your MT4/MT5 backtest report — the numbers that actually matter.

Open EA Analyzer Pro →