Expectancy and the R-Multiple System

Most new traders optimize for the wrong metric. They chase a high win rate because winning feels like winning. But a 75% win rate with 0.5:1 reward-to-riskReward-to-riskDistance to target ÷ distance to stop. Minimum workable setups are typically 2:1 or better.Read in glossary → is a losing strategy. A 40% win rate with 3:1 is a wildly profitable one. The number that actually determines whether you make or lose money is expectancy - and the unit you measure it in is R, the R-multiple. Once you start thinking in R, most of trading's "paradoxes" dissolve. A 7-loss streak isn't scary. A missed trade isn't a catastrophe. Every trade becomes a simple bet of 1R with a variable payoff distribution.

What is an R-multiple?

1R = the dollar amount you risk on a trade. Everything else is measured in multiples of that.

• If you risk $100 on a trade and win $200, that's a +2R trade.
• If you risk $100 and lose $100, that's a −1R trade.
• If you risk $100 and lose $50 (stopped out partway), that's −0.5R.
• If you risk $100 and win $50, that's +0.5R.

Why this matters: every trade is now measurable in the same unit regardless of dollar size. A trader with a $5,000 account who makes +2R made "as much" as a trader with a $500,000 account who made +2R. The strategy is identical; the dollars scale.

More importantly: R removes the emotional charge. A -1R loss is supposed to happen. It's the expected cost of doing business. A +3R win feels the same whether it's $30 or $30,000 when you think in R.

The expectancy formula

Expectancy tells you, per trade, how many R you can expect to make over a large sample.

• WinRate = fraction of trades that win (e.g., 0.40 = 40%)
• LossRate = 1 − WinRate
• AvgWinR = average size of winning trades in R (e.g., 2.5R means winners average 2.5× the risk)
• AvgLossR = average size of losing trades in R (usually ≈ 1R if you honor your stops)

Example - high win rate, bad strategy

A scalper with 75% win rate, average win 0.5R, average loss 1R:

E = (0.75 × 0.5) − (0.25 × 1.0) = 0.375 − 0.25 = +0.125R per trade

Positive, but barely. 100 trades nets 12.5R. On a 1% account, that's 12.5% - before commissions, slippage, and psychology draining.

Example - low win rate, great strategy

A trend follower with 40% win rate, average win 3R, average loss 1R:

E = (0.40 × 3.0) − (0.60 × 1.0) = 1.20 − 0.60 = +0.60R per trade

100 trades nets 60R. On a 1% account, that's 60% - nearly 5× the scalper's bottom line despite losing 60% of trades.

Example - the common losing strategy

Retail average: 50% win rate, average win 1R, average loss 1.5R (bigger losses because stops get widened):

E = (0.50 × 1.0) − (0.50 × 1.5) = 0.50 − 0.75 = −0.25R per trade

Losing strategy. Every trade costs 0.25R on average. After 100 trades, -25R = -25% of the account. This is how most retail accounts die, despite winning half their trades.

The win-rate vs R-multiple grid

This is the table that should be on every trader's monitor. It shows the break-even win rate for each reward-to-risk ratio, plus the win rate needed to produce 20R/100 trades (a good working target).

A trader running 2:1 setups only needs to be right 34% of the time to be profitable, and 40% of the time to make 20% on the account every 100 trades.

This is why pros talk about being "wrong often." The combination of strict 1R losses and outsized winners is what makes low win rates profitable.

Why most traders can't run a 2:1 system

The math above is easy. The execution is brutal. Here's why:

• With a 40% win rate, you will have stretches of 5, 6, 7 losses in a row. These are not strategy failures - they are statistically normal.
• Each loss feels like "something is broken" even when nothing is.
• The trader starts deviating: widening stops, holding past targets, cutting winners early to "lock in something."
• Each deviation kills the 2:1 statistics. Soon the avg win is 1.5R, the avg loss is 1.3R, and expectancy is barely positive.

The edge isn't in the setup. The edge is in the trader's ability to keep honoring the 1R stop and the 2R target over hundreds of trades while feeling like the strategy is failing.

This is why discipline is the actual edge. The math is public knowledge. Executing the math is rare.

Tracking R, not dollars

Your journal should track every trade in R, not dollars. Here's the minimum data per trade:

After 50-100 trades, you can compute:

• Average win (R): sum winners ÷ count winners
• Average loss (R): sum losers ÷ count losers (should be ≈ −1R if you honor stops)
• Win rate: count winners ÷ total trades
• Expectancy: from formula above

Now you have actual data about your strategy instead of narrative. The number is either positive or negative. If positive, scale up (carefully). If negative, either the strategy is broken or you're not executing it correctly.

The "average loss ≠ −1R" diagnostic

If your tracked average loss is significantly bigger than 1R (say, −1.4R), you have a specific problem: you're not honoring stops. That's nearly always widening stops mid-trade, hoping for a reversal. The fix isn't at entry. The fix is mechanical stops placed at the broker that you don't cancel.

If your average loss is less than 1R (say, −0.7R), you're cutting trades early out of fear. You're exiting at −0.7R before the stop because it "feels wrong." That saves small dollars but usually also means your winners get cut early too - destroying the asymmetry that makes the strategy work. The fix is to trust the stop you placed.

Targeting an average loss of exactly −1R is a core discipline. It proves you're honoring stops both when you're wrong and when you're scared.

Expectancy requires sample size

A 20-trade sample tells you nothing about expectancy. Variance dominates.

Rough heuristics:

• < 30 trades: Don't change anything based on results. Not enough data.
• 30-100 trades: Indicative but not definitive. Look for red flags (avg loss >> 1R) but don't overhaul strategy.
• 100+ trades: You have data. Compute expectancy and trust it.
• 500+ trades: You have solid data across many market regimes. Expectancy is probably real.

The trader who runs 10 live trades, hits a 2-loss streak, and "decides the strategy doesn't work" has learned nothing about the strategy and everything about their own variance tolerance.

A worked expectancy calculation

50 tracked trades over 3 months:

• Winners: 22 (44%), total +48R, average +2.18R
• Losers: 28 (56%), total −30R, average −1.07R

Analysis:

• Positive expectancy: 0.36R per trade → over the 50 trades, +18R net.
• Avg loss is −1.07R, slightly wider than the ideal −1R. Investigate: is the extra 0.07R slippage, or widened stops? If slippage, live with it. If widened stops, fix the habit.
• Avg win is 2.18R - solidly above 2R target. Good asymmetry.
• Win rate at 44% is consistent with a 2:1 RR strategy (break-even is 33%, so 44% has margin).

Verdict: the strategy is working. Scale the size (within risk policy), keep journaling, run another 50 trades. If expectancy stays above +0.25R across 200+ trades, you have a durable edge.

Common questions

What if my expectancy is negative after 100+ trades? You have three options: fix the execution (if the data shows you're widening stops / cutting winners), switch strategies, or accept that this isn't the right trading style for you. Don't blame the market.

Does expectancy account for costs? The base formula doesn't. Subtract estimated per-trade cost (commissions + average slippage + spreadSpreadThe difference between the best ask and best bid. Effectively the round-trip cost paid to market makers on every trade.Read in glossary →) from the expectancy to get the net number. For most retail setups this is 0.02-0.08R per trade - material at higher frequencies.

Can expectancy change over time? Yes. Strategies degrade (edges fade as markets adapt), improve (as you get better at execution), or drift (as the regime changes). Monitor expectancy on a rolling window (last 100 trades) rather than assuming the all-time number is current.

Should I calculate expectancy for setups separately? Absolutely yes. "My strategy" is usually 3-4 distinct setups (breakout, pullback, reversal, range-trade). Computing R per setup reveals which setups you're actually profitable in and which are drag. Journal tags make this possible.

Key takeaways

• R-multiple = the unit you risk per trade. Every outcome measures in R. Makes trades comparable across account sizes and setups.
• Expectancy = (WinRate × AvgWinR) − (LossRate × AvgLossR). Positive = edge. Negative = bleed.
• Win rate alone is meaningless. A 75% WR strategy can be losing; a 35% WR strategy can be excellent.
• Break-even win rate drops as R-multiple climbs: 50% at 1:1, 33% at 2:1, 25% at 3:1.
• Most retail accounts die with a ~50% WR but avg loss > avg win, because stops get widened.
• Track every trade in R. After 100+ trades, compute expectancy. Trust the number, not the narrative.
• Avg loss should be ≈ −1R. Higher = not honoring stops. Lower = cutting trades too early.
• Variance is massive below 30 trades; indicative at 100; reliable at 500+.

Up next: DrawdownDrawdownPeak-to-trough decline in account equity. Resets only at new equity peaks. Recovery math is asymmetric.Read in glossary → Math and Recovery - the asymmetric cost of losses, recovery curves, and the drawdown cap that tells you when to stop trading and debug.