High Win Rate in Crypto Contracts ≠ High Profitability? Introduction to Expected Value Thinking
Many traders think a higher win rate means more stability, but when you do the math, a system with a 90% win rate but a terrible risk-reward ratio can still lose money in the long run. What determines whether you ultimately make money is not win rate, but 'expected value' — the average amount you expect to earn per trade.
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
1. Prerequisite: Calculate Your True Win Rate and Risk-Reward Ratio
Before discussing expected value, you need two sets of real data.
Situation A: You have a trading history — Open your trade log (at least the last 50 trades) and calculate two things:
Win rate: Number of winning trades ÷ Total number of trades
Average risk-reward ratio: Average profit per trade ÷ Average loss per trade
Situation B: You don't have enough historical records yet — Use a demo account or backtest to run 50–100 trades and record the same two sets of data.
What counts as a finished task: You hold a concrete set of numbers, like "55% win rate, 1.2:1 risk-reward ratio", rather than "I feel my win rate is quite high".
Common reason for failure: When calculating, you only count the number of winning trades and ignore trading fees and slippage. Fees will shrink your real risk-reward ratio. According to backtesting models, with a theoretical 50% win rate, after accounting for fees the actual win rate may drop to 46.8%, flipping expected value directly from positive to negative. Therefore, use your "actual net profit/loss" rather than "theoretical profit/loss" when gathering statistics.
2. Step Two: Calculate Expected Value — One Formula Tells You If You Can Make Money
The formula for expected value (EV):
EV = (Win rate × Average win) – (Loss rate × Average loss)
Where loss rate = 1 – Win rate.
Example:
Win rate 55%, average win 200 USDC, average loss 150 USDC
EV = (0.55 × 200) – (0.45 × 150) = 110 – 67.5 = +42.5 USDC
This number means: over the long run, you can expect to earn an average of 42.5 USDC per trade. If EV is positive, the system is worth sticking with. If EV is negative, a high win rate is useless.
What counts as a finished task: Plug in your own data and calculate your EV. If EV is negative, stop — do not continue trading with your current system.
Risk reminder: The calculated EV is a long‑term average. Short‑term actual results can deviate significantly from EV — this is called "variance". For example, if you flip a coin 10 times, you might get heads 8 times (80%), even though the true probability is only 50%. Trading randomness is much larger than a coin toss; a positive EV does not guarantee the next trade will be a winner.
3. Step Three: Understand the Relationship Between Win Rate and Risk-Reward Ratio — The Seesaw Effect
If price movements are roughly random to you (you have no edge in direction judgment), the relationship between win rate and risk-reward ratio resembles a "seesaw": lowering the risk-reward ratio can raise the win rate, and vice versa, but the long‑term expected value will not become positive.
Typical combinations where expected value = 0 (excluding fees):
Risk-reward 1:1 requires a win rate > 50% to be profitable
Risk-reward 2:1 requires a win rate > 33% to be profitable
Risk-reward 3:1 requires a win rate > 25% to be profitable
Formula: Breakeven win rate = 1 ÷ (1 + Risk-reward ratio)
If you take a trade with a 100 USDC profit target and a 50 USDC stop loss (risk-reward 2:1), you only need a win rate above 33% to break even. Conversely, if you always take profits at 10 USDC but let losses run to 100 USDC before cutting (risk-reward 0.1:1), you need a win rate above 90% just to avoid losing money.
What counts as a finished task: You can state: "The breakeven win rate of my system is __, and my actual win rate is higher/lower than that."
4. Step Four: Use the Kelly Criterion to Determine Position Size — Don't Go All In
Once you've confirmed a positive EV, the next step is "how much to risk per trade." The Kelly Criterion gives you the optimal position fraction:
f = (bp – q) ÷ b
f = Suggested fraction of capital to risk
b = Risk-reward ratio (Average win ÷ Average loss)
p = Win rate
q = Loss rate (1 – p)
Example: Win rate 55%, risk-reward 2:1
f = (2 × 0.55 – 0.45) ÷ 2 = (1.1 – 0.45) ÷ 2 = 0.325
This means, in theory, using 32.5% of total capital per trade is optimal. In practice, however, it is recommended to use one‑quarter to one‑half of the Kelly value as your actual position size — because real markets involve slippage, fees, and unpredictable black swan events, and using the full Kelly amount carries far too much risk. Keeping risk per trade at 1%–5% of total capital is a more prudent approach.
What counts as a finished task: You have calculated your theoretical position fraction and then deliberately cut it in half to serve as your actual maximum per‑trade position size.
Risk reminder: Leverage does not change the profit or loss amount of a position; it only changes the margin required and the liquidation distance. Opening a 10,000 USDC position with 1x leverage and 100x leverage produces exactly the same profit or loss, but the margin of error differs by 100 times. Even a high‑win‑rate system cannot survive a full‑position liquidation.
FAQ
Q: My win rate is only 40%, but my risk-reward ratio is 3:1. The EV is positive — can I really make money? In theory, yes. 40% × 3 – 60% × 1 = 0.6 (an average profit of 0.6 units per trade), which is indeed a positive expected value. However, the problem with such systems is that losing streaks can come densely. Monte Carlo simulations can model the possibility of 5 to 8 consecutive losses; if your capital cannot withstand that drawdown, a positive EV is irrelevant to you.
Q: How much do fees impact my expected value? Significantly, especially for high‑frequency short‑term trading. Fees and slippage simultaneously compress your "actual win rate" and your "actual risk-reward ratio". If you don't deduct the per‑trade fees before calculating, the EV you compute is artificially inflated. To quote a backtesting model's conclusion: a system with a 50% win rate before fees may see its true win rate drop to just 46.8% after fees, causing the expected value to turn negative.
Q: In the EV calculation, is "average win" a simple arithmetic mean or weighted? Use the arithmetic mean of each profit amount. But be aware: if your system occasionally produces one oversized winner (e.g., a 10x risk-reward trade), the average will be pulled up, and actual stability may not be as good. It is advisable to also look at the "median profit"; if the median is far lower than the average, it means your gains are concentrated in a few large trades, and the strategy has relatively high variance.
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
Next Steps
After completing these four steps, you should have a set of data: true win rate, true risk-reward ratio, EV value, and the Kelly‑suggested position size. Take these numbers and review all your trades from the past week — which ones deviated significantly from the system's average win rate and risk-reward ratio? If the deviation is too large, it means your execution mindset was distorted. If your EV is positive but your equity curve is still losing, it means your sample size is not yet sufficient. Keep going, and let the law of large numbers help you realize that positive expected value.
