On Friday morning, before the open, Sam reread the end of his risk of ruin notes (Lesson 55: Risk of Ruin). His own settings, a 40% Win RateWin rate is the percentage of your trades that result in a profit — calculated as (number of winning trades ÷ total trades) × 100 — and while commonly cited as a measure of trading success, it means very little without also knowing the averRead definition → and 1:2 trades, had survived the simulation just fine at 1% risk.

That bothered him. Forty percent means losing six trades out of ten. Every instinct he had said a strategy that loses most of the time can't be a good one. And the ads in his feed kept promising "90% win rates."

So which number actually matters?

In 30 seconds

  • Win rate is the share of your trades that make money; on its own it says nothing about whether a strategy is profitable.
  • ExpectancyExpectancy is the average amount your strategy makes or loses per dollar risked — it's the single best number for evaluating whether your trading approach has a real mathematical edge.Read definition → is the average result per trade: (win rate × average win) minus (loss rate × average loss), minus costs.
  • A 40% win rate with wins twice the size of losses has positive expectancy, while a 60% win rate with wins smaller than losses can lose money.
  • In a DayTradingToolkit simulation, even a strategy with positive expectancy was behind after 20 trades for 41% of hypothetical traders, so short records mislead.

In this lesson you'll learn what win rate and expectancy mean, how to calculate expectancy in dollars and in R, why a 90% win rate can still lose money, and how many trades it takes before your numbers mean anything. It builds on Lesson 53: The Risk/Reward Ratio.

What is win rate?

Win rate is the percentage of your trades that close with a profit: winning trades divided by total trades, times 100.

It's the number people brag about, because it feels like a score. But it's like judging a shop by how many customers walk out with a bag. A shop that sells a hundred $1 items at a loss and a shop that sells ten $100 items at a profit tell you nothing about each other until you look at the size of each sale.

Your brain counts trades. Your account counts dollars. A win rate only becomes meaningful next to the size of the wins and losses that produced it, and the number that joins them is expectancy.

What is expectancy in trading?

Expectancy is the average amount a strategy makes or loses per trade over many trades, combining how often it wins with how much it wins and loses.

The formula:

Expectancy = (win rate × average win) − (loss rate × average loss) − average costs

It's easiest in R, where 1R is your planned Risk Per TradeRisk per trade is the maximum dollar amount you're willing to lose on a single position if your stop is hit — typically expressed as a percentage of your total account equity (commonly 0.5%–2% per trade) and used as the foundation for calcuRead definition → (the dollars you lose if the stop is hit, as set in Lesson 52: Position Sizing). Take a strategy that wins 40% of the time, makes 2R on winners, loses 1R on losers and pays 0.06R in costs every trade (roughly Sam's costs from Lesson 49):

(0.40 × 2) − (0.60 × 1) − 0.06 = +0.14R per trade

At Sam's $60 risk, that's about $8.40 per trade on average. Not exciting, and not guaranteed, but positive. Over 100 trades it would add up to roughly $840, with a lot of noise around it. You'll see exactly how much noise in a few sections.

Where does a number like +0.14R come from? Mostly from how often your setup beats a coin flip. As Lesson 53 showed, in a pure random walk a 1:2 target gets hit first about one time in three, whatever you do. Sam's 40% is a nearly seven-point edge over that one-in-three, and the edge is the only reason the arithmetic comes out positive.

Why can a 40% win rate beat a 60% win rate?

A 40% win rate can beat a 60% win rate when the winners are much larger relative to the losers, because expectancy depends on win size as much as on win frequency.

Strategy
A
Win rate
60%
Payoff (win ÷ loss)
0.8
Expectancy after 0.06R costs
+0.02R
Average per 100 trades at $60 risk
+$120
Strategy
B
Win rate
50%
Payoff (win ÷ loss)
1.0
Expectancy after 0.06R costs
−0.06R
Average per 100 trades at $60 risk
−$360
Strategy
C
Win rate
40%
Payoff (win ÷ loss)
2.0
Expectancy after 0.06R costs
+0.14R
Average per 100 trades at $60 risk
+$840
Strategy
D
Win rate
30%
Payoff (win ÷ loss)
3.0
Expectancy after 0.06R costs
+0.14R
Average per 100 trades at $60 risk
+$840
Strategy
E
Win rate
40%
Payoff (win ÷ loss)
1.2
Expectancy after 0.06R costs
−0.18R
Average per 100 trades at $60 risk
−$1,080

Calculated from the formula above. Hypothetical strategies.

Illustrative, before costs: the trader who wins less often has the better expectancy.

Strategy A wins most often and barely breaks even. Strategy B, a coin flip with even payoffs, loses money once costs are counted. C and D lose most of their trades and have the best expectancy on the table. Strategy E shows the flip side: a 40% win rate with a payoff of only 1.2 is a steady loser.

So a low win rate alone is harmless. It only hurts when the winners are small too.

Can a 90% win rate still lose money?

Yes. A strategy can win 90% of its trades and still lose money if the 10% of losers are large enough, because the average loss does the damage the win rate hides.

Here's a made-up strategy of the kind a flashy ad might describe:

Item
Win rate
Value
90%
Item
Average win
Value
+0.3R ($18 at $60 risk)
Item
Average loss
Value
−4R ($240)
Item
Expectancy before costs
Value
(0.90 × 0.3) − (0.10 × 4) = −0.13R
Item
Expectancy after 0.06R costs
Value
−0.19R, about −$11.40 a trade
Item
Win rate needed to break even before costs
Value
about 93%

DayTradingToolkit calculation. Hypothetical strategy, not a real record.

Nine trades out of ten feel great. Over 100 trades the strategy above still loses about $1,140. The one trade that goes wrong costs thirteen times as much as a typical win makes.

This is the pattern the U.S. Commodity Futures Trading Commission (CFTC) warns about. Its advisory on AI trading bots says scammers claim their algorithms can "yield 100 percent 'win' rates." A win rate shown without the average win, the average loss and a long, checkable record is a sales number.

Can you just pick a high payoff and a high win rate?

No. Win rate and payoff pull against each other: a closer target gets hit more often but pays less, and a distant target pays more but gets hit less often.

Move Sam's target from $1.00 to $2.00 above entry and the payoff doubles, but the price has to travel twice as far before the stop is hit. Fewer trades will make it. That's why a 1:5 target isn't automatically better than a 1:2 one, as Lesson 53 warned.

Win rate and payoff pull against each other; the right mix is the one your data supports.

The only way to know where your strategy sits is to measure it. Your target should come from the chart; your win rate and average R should come from your logged trades. And once you have a number, the next question is how much to trust it.

How many trades does it take to know your expectancy?

It takes far more trades than most beginners expect, because random streaks swamp the real edge over short runs.

The DayTradingToolkit simulation ran each strategy from the table for 10,000 hypothetical traders and counted how many were ahead after 20, 100 and 500 trades.

Strategy
A: 60%, 1:0.8
Expectancy
+0.02R
Ahead after 20 trades
59%
Ahead after 100
63%
Ahead after 500
69%
Strategy
B: 50%, 1:1
Expectancy
−0.06R
Ahead after 20 trades
42%
Ahead after 100
25%
Ahead after 500
8%
Strategy
C: 40%, 1:2
Expectancy
+0.14R
Ahead after 20 trades
59%
Ahead after 100
82%
Ahead after 500
98%
Strategy
E: 40%, 1:1.2
Expectancy
−0.18R
Ahead after 20 trades
25%
Ahead after 100
4%
Ahead after 500
0%

DayTradingToolkit simulation, fixed seed, independent trades, costs of 0.06R per trade. Hypothetical, not real trader data.

Read the 20-trade column. Four in ten traders with a genuinely positive strategy (C) were behind. And 42% of traders with a losing strategy (B) were ahead, feeling great about a system that loses over time. A few weeks of results can't tell those two apart.

"Ahead" only says which side of zero you landed on. It doesn't show how far results can swing, so a second run looked at the dollars. It followed strategy C at $60 of risk for 100,000 hypothetical traders:

Trades
20
Middle 90% of outcomes
−$552 to +$888
Median outcome
+$168
Trades
100
Middle 90% of outcomes
−$600 to +$2,280
Median outcome
+$840
Trades
500
Middle 90% of outcomes
+$960 to +$7,440
Median outcome
+$4,200

DayTradingToolkit simulation: 100,000 hypothetical traders, 40% win rate, 1:2, 0.06R costs, $60 risk, independent trades. "Middle 90%" runs from the 5th to the 95th percentile.

After 20 trades, a trader running the same good strategy could be down $552 or up $888 and neither would be a surprise. The RangeA range (or trading range) is a price zone where a stock oscillates between a defined support floor and resistance ceiling without making meaningful new highs or lows — price is consolidating rather than trending, and the dominant trade is Read definition → narrows slowly. Only around 500 trades does the bottom of the range climb above zero.

A second calculation asks how many trades you'd need before luck alone rarely explains a positive average. The test is that the average result per trade sits roughly 1.65 to 2 standard errors above zero (a standard error is how far an average typically wobbles by chance).

Strategy
A: 60%, 1:0.8
Expectancy
+0.02R
Trades needed
about 5,300 to 7,500
Strategy
C: 40%, 1:2
Expectancy
+0.14R
Trades needed
about 300 to 425
Strategy
D: 30%, 1:3
Expectancy
+0.14R
Trades needed
about 465 to 660

DayTradingToolkit calculation using each strategy's known odds and 0.06R costs. Real logs are messier, so treat these as best-case counts.

The thin edge of strategy A is the lesson inside the table. A strategy that earns 0.02R a trade is real on paper, and you'd need thousands of trades to prove it to yourself. Nobody should build a plan on an edge they can't verify.

Sam checks his own record

Sam's practice log since Lesson 43 has 8 trades. All figures are hypothetical and in his paper account.

Group
Winners
Trades
4
Total
+$390.28
Group
Losers within his rules
Trades
3
Total
−$183.83
Group
CDEF (rules broken)
Trades
1
Total
−$600.11
Group
All trades
Trades
8
Total
−$393.66

Hypothetical practice account.

Sam's planned expectancy at $60 risk (hypothetical)

Win rate × average win (0.40 × $120)$48.00
minus Loss rate × average loss (0.60 × $60)$36.00
minus Average costs per trade$3.60
equals Planned expectancy+$8.40 a trade (+0.14R)

His winners averaged $97.57, about 1.6R, and his rule-following losers averaged $61.28, about 1R. Those look like his plan.

Then look at what the 8 trades say about his odds. He won 4 of 8, a 50% win rate. With so few trades, a 95% confidence range for that win rate (DayTradingToolkit calculation) runs from about 22% to 78%, which covers his planned 40% and also every number far from it. His seven clean trades made +$206.45, about $29.49 a trade, three and a half times his planned $8.40. That's luck talking, and it says nothing about whether he's better than the plan.

CDEF is the other story. One trade lost −$600.11, about 10R. At his planned $8.40 a trade, it takes roughly 71 average trades to earn that back.

Two lessons for Sam. Eight trades say almost nothing about his real expectancy. And the fastest way to wreck any expectancy is a single loss far bigger than 1R.

His note: "Win rate is a feeling. Expectancy is a number. I need 100 trades before I believe mine." A hundred trades is a sensible first checkpoint (82% of strategy C traders were ahead by then, against 25% of strategy B). Confidence takes several hundred, so he'll call his number an early estimate until then.

Myth and fact

Myth and fact

Myth A good strategy wins most of its trades.
Fact Profitability depends on win rate and payoff together. Strategies that lose most trades can have positive expectancy, and strategies that win most trades can lose money.

Your turn

Calculate your expectancy (15 minutes, practice log)

0 of 6 done

Check your understanding

Check your understanding

A strategy wins 30% of the time, wins average 3R and losses average 1R. Ignoring costs, what's its expectancy?

Check your understanding

Why can't you just set a far target and keep your win rate the same?

Check your understanding

A strategy advertises a 90% win rate with an average win of +0.3R and an average loss of −4R. What happens over many trades?

Key takeaways

  • Win rate tells you how often you feel good. Expectancy tells you whether you get paid.
  • Expectancy = (win rate × average win) − (loss rate × average loss) − costs.
  • Measure results in R so trades of different sizes compare fairly.
  • A closer target raises win rate and lowers payoff; only your log shows the real balance.
  • Twenty trades prove very little and a hundred is a first checkpoint, so protect your expectancy by keeping every loss near 1R.

Frequently asked questions

What is a good win rate for day trading?
There isn't one number. A 35% win rate can be excellent with large winners, and a 65% win rate can lose money with small ones. At a 1:1 payoff and normal costs, 60% is solidly profitable. At 1:0.5 you'd need about 67% just to break even before costs, so 60% loses. Check the break-even win rate for your average payoff (Lesson 53 has the table) and aim to be clearly above it.
What is a good expectancy per trade?
There's no official benchmark. Anything above zero after costs means the plan has paid, and Sam's planned +0.14R is $8.40 a trade at $60 of risk. What matters more is whether the number holds up over enough trades. A big figure on a short record deserves suspicion: after 20 trades, the middle 90% of results for the same strategy ran from −$552 to +$888.
Should I track win rate, profit factor or expectancy?
Make expectancy in R your main number, and keep average win, average loss and your deepest DrawdownA drawdown is the peak-to-trough decline in account equity — it measures how far you've fallen from your high-water mark, either from a single losing trade or a series of losing trades.Read definition → beside it. Profit factor is your total winning dollars divided by your total losing dollars, and above 1.0 means the strategy made money over that period. It's the same information as expectancy in a different shape (strategy C's is 1.33 before costs), and it's just as shaky over a small number of trades. Win rate is an input to those, not a score.
How can I improve my expectancy without changing my strategy?
Start by removing the losses that break your rules. Without CDEF, Sam's 8-trade log goes from −$393.66 to +$206.45. Then cut costs (each 0.06R of cost comes straight off your edge) and skip setups whose payoff sits below the break-even line from Lesson 53. Lesson 58: Where to Place a Stop-Loss shows how stop placement keeps losers near 1R.
Should I trust services that advertise very high win rates?
Be skeptical. The CFTC warns that scammers promote trading bots with claims of huge returns and even 100% win rates, and FINRA's day-trading risk disclosure (Rule 2270) tells customers to be cautious of claims of large profits from day trading. A win rate without average win, average loss and a long, verifiable record tells you nothing about profitability.
Does expectancy change over time?
Yes. Markets change, and a strategy's win rate and payoff drift with them. Recalculate on a rolling basis, such as your last 100 trades, and compare it with your longer record.

Up next: is three trades really three trades?

On Monday, Sam's scanner shows three chip stocks setting up at once, each with a clean stop, a correct size and a 1:2 ratio. Three good trades, $60 of risk each. But all three moved on the same industry news. If the sector turns, do they lose separately, or all together? Lesson 57: Correlation Risk finds out.

Disclaimer

This lesson is for educational purposes only and is not financial advice. Expectancy figures and simulation results are hypothetical, rely on fixed assumptions and independent trades, and don't predict the results of any strategy. A positive record over a small number of trades may be luck. All Sam examples and figures are hypothetical. Trading involves the risk of loss, and nothing here guarantees any result. For our complete disclaimer, visit https://daytradingtoolkit.com/disclaimer.

Article Sources

This lesson relies on a CFTC advisory, FINRA's day-trading risk disclosure and original DayTradingToolkit simulations and calculations.
  • CFTC: AI Won't Turn Trading Bots into Money Machines - Scammers claim trading algorithms can yield huge returns and "100 percent 'win' rates."cftc.gov
  • FINRA Rule 2270: Day-Trading Risk Disclosure Statement - Be cautious of claims of large profits from day trading.finra.org
  • DayTradingToolkit simulation (October 2026): five hypothetical strategies, 10,000 traders each, 20, 100 and 500 independent trades, costs of 0.06R per trade, fixed seed 56.
  • DayTradingToolkit simulation (October 2026): strategy C (40% win rate, 1:2), 100,000 hypothetical traders, $60 risk, 0.06R costs, fixed seed 565656; dollar outcomes after 20, 100 and 500 trades.
  • DayTradingToolkit calculations (October 2026): the 90% win rate example, Sam's win-rate confidence range (Wilson interval, 95%) and the trades-needed table (known odds, normal approximation).