Risk & Reward
The math that lets a strategy lose most of its trades and still make money — and why your bot refuses trades that don't offer enough reward. Amber terms are tappable.
2.1Stop counting wins
Beginners judge trading like a test score: more wins = better. That instinct is wrong, and it's the single most expensive instinct in trading. What matters is the combination of your and how big your wins are compared to your losses. Watch what happens across ten trades where every loss risks 1 unit and every win pays 2:
Those units have a name: — one R is simply the amount you risked on the trade. Lesson 1's calculator fixed your R at 1% of balance; from now on we measure every result in R. A trade that hits a 1:2 target is +2R; a stopped-out trade is −1R, always.
2.2The break-even see-saw
Win rate and reward size trade off against each other. The more your winners pay, the less often you need to be right. Here's the win rate you need just to break even at each reward ratio:
sl_pips × 2.0 in the code.Notice what this kills: the fantasy of needing to be right all the time. A system with a modest win rate and disciplined 1:2 exits beats a "90% win rate" system that lets one loss wipe ten wins — the classic beginner account-killer.
2.3Expectancy — your edge in one number
is the average result you can expect per trade, combining win rate and reward into a single figure:
⚙ Try it — does a system have an edge?
Try 55% at 1:1 (small edge), 30% at 1:2 (negative — no amount of discipline saves a system with no edge), 40% at 1:3 (strong). Positive expectancy is the only permission slip for trading a strategy.
2.4Losing streaks — the test you must survive
Here's the uncomfortable consequence of a 40% win rate: losing streaks are normal, not a sign of failure. Over 100 trades at 40%, a run of 6+ straight losses is more likely than not. The question is what a streak does to your account — and that depends entirely on Lesson 1's golden rule:
This is why — the drop from your account's peak — is watched so closely, and why "boring" 1% risk is what lets a positive-expectancy system survive long enough for its edge to show up.
2.5Where you'll meet all this in IronGate
Every backtest you run prints exactly these numbers. From a real run of the backtest engine:
That run was a strategy on random synthetic data — and the numbers correctly say "no edge here" ( below 1, negative ). You now know how to read a system's report card and reject it. When a backtest on real data shows profit factor above ~1.3 with a healthy trade count and tolerable drawdown, it earns the next test — and everything in between is Lesson 8's story.
1) Measure results in R, not dollars or wins. 2) A system is judged by expectancy, never win rate alone. 3) Position sizing is what turns a survivable streak into a fatal one — or not.