Once you have a sample of trades, from a backtest, forward test or your journal, a handful of numbers tell you whether the system works and how it behaves. Knowing which ones matter, and how they can mislead, is what turns a pile of trades into a decision.
| Metric | Formula | This sample | Rough guide |
|---|---|---|---|
| Win rate | Wins ÷ trades | 56% | Only meaningful with average win and loss |
| Average win / loss | Total won ÷ wins; total lost ÷ losses | 1.86R / 1.00R | Shows the reward-to-risk you actually get |
| Expectancy | Net result ÷ trades | +0.60R | Above 0 after costs; 0.2R+ is solid |
| Profit factor | Gross profit ÷ gross loss | 2.36 | Above 1 is profitable; 1.5+ is strong |
| Maximum drawdown | Largest peak-to-trough fall | 6.0R | Must be survivable at your risk per trade |
| Worst losing streak | Longest run of losses | 6 trades | Prepare for longer in live trading |
| Return / volatility | Average R ÷ standard deviation of R | 0.41 | Higher means smoother; the idea behind the Sharpe ratio |
Which metrics to watch
- Expectancy is the headline: is the system worth trading at all?
- Maximum drawdown and worst streak tell you whether you can survive it, and how to size it. At 1% risk, a 12R drawdown is about 11%.
- Profit factor is a quick sanity check that combines win rate and size of wins.
- The Sharpe ratio (return relative to volatility, usually annualised) matters most when comparing systems or reporting to others.
How metrics mislead
- Small samples: 20 trades can show almost any result by chance. Wait for 50–100.
- Before costs: always calculate after spreads and commissions.
- One big winner: check how the numbers look without your best trade.
- Changing conditions: split results by year or by market regime to see if the edge is stable.
Key takeaways
- Expectancy tells you whether a system is worth trading; drawdown and losing streaks tell you whether you can survive it.
- Profit factor (gross profit ÷ gross loss) above 1.5 is strong; return relative to volatility shows how smooth the results are.
- Use at least 50–100 trades, calculate after costs, and check results without the best trade.
- Measure in R to compare systems across markets and account sizes.