Measuring trades honestly

BeginnerDuration ~30 min video + 30 min hands-onTools Python (pandas optional), A trade log — sample provided

Why this lesson

Win rate is the most quoted and least useful number in trading. A system can win 70% of its trades and still lose money; another can lose 60% of the time and compound beautifully. The metrics in this lesson — R-multiples, expectancy, profit factor, drawdown — are how you see past the seductive number to whether a strategy actually makes money. This is the core measurement skill the whole level is named for.

Watch

Three videos, in order. The first two build the R-multiple and expectancy from scratch; the third shows a real trader stress-testing his own metrics across 30 live systems.

Watch for: The R-multiple is a trade's profit or loss expressed as a multiple of the initial stop-loss risk. Key discipline (stated early): set the stop from strategy logic first, then derive position size to hit your target risk percent — not the other way round. Around 9:00–10:30 it maps a live strategy's R-distribution: a tight −1R stop cluster and ~+7R average winners. R-normalization is what lets you compare trades taken at different account sizes.

Watch for: Expectancy is the mean R-multiple across trades. A fully worked 8-trade example at 2:39–4:00 gives 6.3R total ÷ 8 = 0.79R expectancy. Watch two honest caveats: 8 trades is statistically meaningless (you need thousands), and expectancy hides both variance (a steady 1R system can beat a lumpy 2R one) and trade frequency.

Watch for: Davey shows real (not cherry-picked) stats across ~30 strategies he actually trades: win rates spanning 28–90% (median ~49%) and reward:risk medians around 1.7. The core lesson at 8:53–11:27 is a win%-vs-reward:risk profitability curve showing that both matter jointly — chasing a high win rate alone is a classic retail trap. Also note his distinction between plain average trade and the Van Tharp expectancy formula.

The explainer

Everything starts with the R-multiple. Take one trade. Its risk-per-trade is the distance from entry to your initial stop, in money — call that 1R. If your stop is 20 pips away on a position sized so those 20 pips equal $100, then 1R = $100. Now express the result as a multiple of that risk: a trade that made $300 is +3R; one stopped out for the planned $100 is −1R; one that ran to your stop but slipped to −$130 is −1.3R. The R-multiple strips out account size and position size, so a trade you took at $1,000 equity and one at $10,000 equity are directly comparable (Van Tharp’s framing, applied by Darwinex). Without this normalization, your later high-equity trades quietly dominate any naive profit-based average and destroy its statistical meaning.

Once every trade is an R-multiple, expectancy is just their average. Add the R-multiples, divide by the number of trades. In the worked example above, eight trades summing to 6.3R give an expectancy of 0.79R — meaning that, on average, each trade returns 0.79 times what you risked. Positive expectancy is the definition of an edge net of costs; zero or negative means you are, on average, feeding the market. Two honest limits, both stated in the video: expectancy over a handful of trades is noise (you need a large sample, covered in Level 2), and a single expectancy number hides how lumpy the returns are and how often you trade.

Now the headline: why a 70% win rate can lose money. Win rate and payoff ratio (the size of average winners versus average losers) are two independent levers, and expectancy needs both. Suppose you win 70% of trades but your winners average +0.3R while your losers average −1R. Expectancy = 0.70 × 0.3 + 0.30 × (−1) = 0.21 − 0.30 = −0.09R. You win most of the time and still bleed, because the rare losses are far larger than the frequent wins. Flip it: a system that wins only 40% but makes +2R on winners and −1R on losers has expectancy 0.40 × 2 + 0.60 × (−1) = +0.20R. This is the single most important idea in the lesson, and it’s why “high win rate” as a selling point is a tell, not a virtue.

Two more metrics round out honest measurement. Profit factor is gross profit divided by gross loss across all trades — above 1 means the winners outweigh the losers in absolute money, and it’s a quick sanity check that pairs well with expectancy. Drawdown is the decline from an equity peak to a trough, and reporting it honestly means stating two numbers: the depth (how far equity fell, in percent or R) and the duration (how long you were underwater before making a new high). Depth alone is a half-truth — a 20% drawdown that recovers in a week is a very different experience from a 20% drawdown that grinds for eight months, and the second is what makes people abandon systems at the worst time.

Finally, a per-trade diagnostic you’ll want later: MAE and MFE — maximum adverse excursion and maximum favorable excursion. For each trade, MAE is the worst unrealized loss it showed while open, and MFE the best unrealized profit. They tell you whether your stops sit too tight (large MAE on eventual winners) or your exits leave money on the table (large MFE versus actual exit). You don’t act on them yet; you just start logging them, because they’re the raw material for improving exits in Level 2. (Van Tharp’s SQN — expectancy divided by the standard deviation of R-multiples, scaled by sample size — is a natural next step here; note it and move on.)

The through-line: never judge a strategy by its win rate. Convert to R, average to expectancy, cross-check with profit factor, and always state drawdown as depth and duration. And remember Lesson 1.3 — every one of these must be computed on net results, or you’re measuring a fantasy.

Do it

You’ll compute the full metric set from a real-looking trade log — first the logic, then optionally in code.

Here is a 12-trade log, already expressed in R-multiples (net of costs):

+2.1, -1.0, -1.0, +3.4, -1.2, +0.4, -1.0, +5.0, -1.0, -1.1, +0.5, +1.9

1. Win rate. Count positive trades ÷ total. (5 wins of 12 ≈ 42%.)
2. Expectancy. Sum the R-multiples and divide by 12. Is it positive?
3. Payoff ratio. Average of the winners ÷ absolute average of the losers. Notice it’s the winners’ size carrying this system, not their frequency.
4. Profit factor. Sum of positive R ÷ absolute sum of negative R.
5. Max drawdown (in R). Walk a running cumulative total, track the running peak, and record the largest peak-to-trough drop. Note which trades you were underwater across — that span is the drawdown duration.
6. Write one sentence: does a 42% win rate here mean the system is bad? (It shouldn’t — check expectancy, not win rate.)

Optionally, verify in Python:

import numpy as np

R = np.array([2.1, -1.0, -1.0, 3.4, -1.2, 0.4, -1.0, 5.0, -1.0, -1.1, 0.5, 1.9])

win_rate    = (R > 0).mean()
expectancy  = R.mean()
payoff      = R[R > 0].mean() / abs(R[R < 0].mean())
profit_fac  = R[R > 0].sum() / abs(R[R < 0].sum())

equity      = R.cumsum()               # cumulative R
peak        = np.maximum.accumulate(equity)
drawdown    = equity - peak            # <= 0
max_dd      = drawdown.min()           # most negative

print(f"win rate    {win_rate:.0%}")
print(f"expectancy  {expectancy:.2f}R")
print(f"payoff      {payoff:.2f}")
print(f"profit fac  {profit_fac:.2f}")
print(f"max dd      {max_dd:.2f}R")

Confirm your by-hand numbers match. Keep this file — Lesson 1.5 feeds net R-multiples into these exact formulas.

Terms introduced

Terms introduced in this lesson

• R-multiple
• risk-per-trade
• expectancy
• win rate vs payoff ratio
• profit factor
• drawdown
• MAE/MFE

Each term links to the glossary, where it is defined by the way you just used it.

Check

Check yourself

1. An R-multiple expresses a trade's result as:

   the profit in dollarsthe number of pips gaineda percentage of the accounta multiple of the initial risk (the distance to the stop)

2. Expectancy is:

   the profit factor times the win ratethe win ratethe largest winning tradethe average R-multiple across all trades

3. A system wins 70% of the time but has expectancy below zero. The most likely reason is:

   winners are small relative to losers (poor payoff ratio)the profit factor is above 1it trades too rarelythe win rate is too high to be real

4. Profit factor is:

   net profit divided by number of tradesthe average winner divided by the average loserwin rate divided by loss rategross profit divided by gross loss

5. Reporting drawdown honestly requires stating:

   only the deepest percentage lossboth the depth of the equity decline AND how long it lastedthe win rate during the drawdownonly the number of losing trades

6. MAE and MFE measure, for each trade:

   the worst and best unrealized excursion while the trade was openthe margin and equity usedthe average entry and exit pricethe commission and swap paid

Check answers

You can move on when you can… compute R-multiple, expectancy, drawdown (depth AND duration), profit factor, and explain why a 70% win rate can lose money.

Mark this lesson complete

Go deeper

• QuantStart’s articles on performance measurement and risk metrics — the best free written corpus on expectancy, drawdown, and the Sharpe family; effectively a free textbook.
• The Darwinex “Van Tharp” series (the first two videos above) continues into the System Quality Number if you want the variance-aware extension of expectancy.
• The book that owns this topic: Ernie Chan’s Quantitative Trading ch. 3 is the cleanest short treatment of measuring a strategy honestly; Van Tharp’s Trade Your Way to Financial Freedom is the origin of R-multiples and expectancy if you want the primary source.