Show a beginner two trading systems and ask which is better.
System A wins 70% of its trades. System B wins 35%.
The beginner picks A instantly — 70 beats 35, obviously.
A professional asks one question before answering anything: "What happens on the wins, and what happens on the losses?"
That question is the entire subject of this chapter, and the unit that makes it answerable is called R.
R simply means: one unit of risk. If you risk ₹10,000 on a trade (Chapter 2's formula, in action), then that ₹10,000 is 1R. Win ₹20,000 on that trade, and you made +2R. Lose the full ₹10,000, and you lost −1R — because your stop, correctly placed, caps the loss at exactly one unit, never more.
Once every trade is measured in R instead of rupees, something powerful happens: trades of wildly different sizes become directly comparable. A ₹5,000 win on a small position and a ₹50,000 win on a large one might both be exactly +2R — the same quality of outcome relative to what was risked, however different the numbers look on a statement. R strips away the noise of position size and shows you the only thing that actually matters: how the trade performed against its own risk.
Now go back to Systems A and B, look at the figure, and measure them in R:
System A wins 70% of the time — but its average win is only +0.6R, and its average loss is a full −1R. Small wins, full losses.
System B wins only 35% of the time — but its average win is +2.2R, a big multiple of its risk, while its average loss stays the disciplined −1R.
Calculate what actually matters — expectancy, the average result you can expect per trade, over many trades:
EXPECTANCY = (Win% × Avg Win in R) − (Loss% × Avg Loss in R)
System A: (0.70 × 0.6R) − (0.30 × 1.0R) = 0.42R − 0.30R = +0.12R per trade
System B: (0.35 × 2.2R) − (0.65 × 1.0R) = 0.77R − 0.65R = +0.12R per trade
Identical expectancy. The high-win-rate grinder and the low-win-rate hunter arrive at almost exactly the same place, per trade, over a large enough sample — despite looking like completely different animals day to day. Win rate alone told you nothing. It needed its partner — the average size of wins and losses — before it meant anything at all.
This single formula quietly explains two of the most common trading personality mismatches in this Academy's schools:
The trend-follower's curse (you met this family in the Algorithmic Trading school): trend systems often look like System B — losing more often than they win, each loss a disciplined −1R, waiting for the occasional +5R or +10R trend that pays for a dozen small losses. The strategy can be genuinely excellent and still feel miserable to run, because human wiring (the Psychology school's bent curve) hates losing most of the time, regardless of the math working out. Traders routinely abandon perfectly good System-B strategies right before the big win arrives, purely because the losing streak felt unbearable.
The high-win-rate trap: System-A-style strategies feel wonderful — you're 'right' most of the time — which is exactly why beginners gravitate toward them without checking the other half of the equation. A strategy that wins 80% of the time but risks 3R to make 0.3R has negative expectancy despite an enviable win rate; one bad trade erases many good ones. "I'm right most of the time" and "I make money" are different claims, and only expectancy connects them honestly.
So here is the practical use of this entire chapter, and it belongs in every strategy review you'll ever do:
Never evaluate a strategy — yours, a course's, a friend's — on win rate alone. Demand all three numbers: win rate, average win in R, average loss in R. Calculate expectancy. A strategy with modestly positive expectancy, traded with disciplined sizing (Chapter 2) and proper stops (Chapter 3), across enough trades to let the average show up (Chapter 5, next, is about exactly how many trades 'enough' requires) — that is a real edge, whatever its win rate looks like on the outside.
And notice something quietly reassuring buried in this chapter: you don't need to be right most of the time to make money. You need your wins, on average, to be worth more than your losses cost — measured honestly, in R, over a real sample. That reframe alone changes how it feels to take a loss inside a good strategy: not a failure, just −1R, exactly as planned, one data point in an average that's still working in your favour.
One number is still missing from the picture: how bad a losing streak can get before the averages show up — and what that means for how you should really be sizing. That's next.
Key Takeaway
R = one unit of risk; measuring every trade in R makes wins and losses across different position sizes comparable. Expectancy = (Win% × Avg Win in R) − (Loss% × Avg Loss in R) — the only honest measure of a strategy, because a 70%-win-rate system and a 35%-win-rate system can carry identical expectancy. You don't need to be right often; you need your wins to be worth more than your losses, on average, over a real sample.
Think About It
Calculate your own last 20 trades in R: win rate, average win in R, average loss in R, expectancy. Are you actually profitable per trade — or does it just FEEL that way because you win often?
Risk Lab — Calculate Your Real Expectancy
Pull your last 20–30 trades. For each, convert the result into R: (profit or loss) ÷ (amount originally risked on that trade). Losses at a properly hit stop should cluster near −1R; if yours are scattered wildly below −1R, that's Chapter 3 talking to you.
Compute: win rate, average winning R, average losing R.
Run the expectancy formula. Is the number positive? Is it meaningfully positive, or a coin flip's width from zero?
Then write one sentence comparing your GUT FEELING about your trading ('I feel like I'm doing well/badly') to what the actual expectancy number says. If they disagree, trust the number — it's the only one that survived contact with your memory intact.