Why Win Rate Alone Isn't Enough｜Risk-Reward and Expectancy

When people evaluate a trading method, the first thing most of them look at is the "win rate." Thinking "high win rate = good method" is a natural assumption. But if you judge a method by win rate alone, the conclusion can diverge from your actual P/L. You can have a high win rate and still watch your account shrink. It seems strange at first, but in trading it is not unusual.

This article explains why win rate alone is not enough, and lays out two perspectives worth checking alongside it when you evaluate a method: "risk-reward" and "expectancy." It does not recommend any specific method or show a way to win. It is purely a framework for how to look at a method.

What does win rate actually measure?

First, let's confirm what the win rate number represents.

Win rate is the percentage of your trades that ended in profit. If 7 out of 10 trades are profitable, your win rate is 70%. Because it is simple and intuitive, it is the first metric people reach for when evaluating a method.

But there is a pitfall here. Win rate only counts how many times you won. It does not include the size of each win or loss in money. You know you won 7 times and lost 3 times, but how much you gained in those 7 trades and how much you lost in those 3 is not captured in the win rate.

In other words, win rate shows only one side of a method.

Why win rate alone is not enough

If the size of your wins and losses differs, your P/L can be negative even with a high win rate.

For example, consider a method like this:

Because you win often, the win rate comes out high at 80%. But an occasional large loss wipes out all the small wins before it. As a result, the total ends up negative. This pattern tends to appear in trading where you take small profits quickly and put off cutting losses.

Conversely, your P/L can be positive even with a low win rate. Even if you win less often, if each win is large and you cut losses while they are small, a few wins can outweigh many losses.

What this shows is that you cannot judge whether a method is good or bad from a high or low win rate alone, because it leaves out the monetary size of each winning and losing trade. In fact, chasing win rate alone can push you toward a method that "wins small and loses big." Risk-reward, covered next, fills in this missing perspective.

※The figures above are examples to illustrate the concept and do not represent the results of any specific method.

The risk-reward perspective

Risk-reward expresses the "size ratio" between one win and one loss.

Specifically, it is the ratio between the profit from one win (the reward) and the loss from one loss (the risk). If your average win is $200 and your average loss is $100, your risk-reward is 2:1. That means your wins are twice the size of your losses.

When you look at risk-reward together with win rate, a method looks different. With a high win rate but poor risk-reward (losses larger than wins), overall performance can suffer; with a low win rate but good risk-reward (wins larger than losses), the overall result can be positive. You cannot judge from win rate or risk-reward alone; only by combining both can you get a clearer picture of a method's performance.

So how do you combine these two into "one view"? That is where expectancy comes in.

How to think about expectancy

Expectancy is a way of estimating what the average result per trade is likely to be if you keep using the same approach for 100 or 200 trades. It combines win rate and risk-reward to capture the average result per trade.

You think in terms of the win rate (the probability of winning) and the size of each win, together with the probability of losing and the size of each loss. Even with a high win rate, a large loss per losing trade lowers expectancy. Even with a low win rate, a large gain per winning trade can keep expectancy positive. It is easiest to think of expectancy as a single yardstick that combines win rate and risk-reward, rather than looking at them separately.

What matters here is that expectancy is not a number that predicts the future. It is a yardstick for evaluating past records, not a guarantee that "if this is positive, you will win." Market conditions do not necessarily stay the same, and there is no guarantee that past numbers will repeat in the future. It is safest to treat expectancy as a perspective for looking at a method from multiple angles, not as an indicator of guaranteed success.

Why you need a sufficient number of trades

Here, let's confirm the premise that makes both win rate and expectancy meaningful: the number of trades.

With just a few wins or losses, both win rate and expectancy are heavily influenced by chance. Win three times in a row and your win rate is 100%, but that number means nothing. That is why you need a reasonably large number of trades. Only after enough repetitions do win rate, risk-reward and expectancy start to mean something as numbers that reflect a method.

That said, building up enough trades through live trading alone takes time. Even if you wait, the setups that meet your entry conditions do not come around many times a day. This is where backtesting with historical charts (Chart Replay) helps: you can build up trades in a short time without risking real money. By rewinding past markets and trying the same method again and again, you can collect enough trades for the numbers to mean something in a short period. The roles of backtesting and real-time practice are covered in detail in a separate article, "Backtesting vs Demo Trading｜Practice Goals and Differences."

Look at several metrics together to evaluate a method

To sum up so far, no single number determines how a method should be evaluated. In practice, you look at a combination of metrics like these:

• Win rate (the percentage of winning trades)
• Risk-reward (the ratio of the size of one win to one loss)
• Expectancy (the average result per trade)
• Profit factor (the ratio of gross profit to gross loss)
• Maximum drawdown (the largest decline in your account)
• Standard deviation of P/L (how widely results vary from trade to trade)
• Recovery factor (how much profit you made relative to drawdown)

If you pull out just one number, you see only one side of a method and are more likely to misjudge it. Only by putting several metrics side by side can you see whether a method fits your capital and risk tolerance.

For example, two methods with positive expectancy can have very different performance characteristics. A method that makes a large profit only once in 10 trades and takes small losses the rest of the time (low win rate, large reward) and a method that wins steadily most of the time but occasionally takes a fairly large loss (high win rate, small reward) have completely different distributions of individual wins and losses (P/L distributions). Even with the same overall expectancy, how bumpy the growth of your account is will differ. To know which kind of swings you can tolerate, you need to look from multiple angles. This "bumpiness" can also be checked numerically as the standard deviation of P/L, which shows how widely results vary from trade to trade. The larger the standard deviation, the more individual results swing, even at the same expectancy.

A high win rate is not bad in itself. The problem is judging by win rate "alone." If you look at win rate, look at risk-reward and expectancy at the same time. This habit is the first step toward avoiding misreading win rate.

How to check the numbers for your own method

So how do you find these numbers for your own method? The key is to record your trades and aggregate them by method.

If you record your trades and calculate figures grouped by method, you can see trends not only in win rate but also in risk-reward and expectancy. The number of trades mentioned in the previous section can also only be counted once you record them.

In ENTRIQ, you can group and analyze trades with Strategy Tags. For example, if you tag "buying the dip" and "double top" separately, metrics such as win rate, risk-reward, profit factor, expectancy, recovery factor, standard deviation of P/L, P/L distribution and cumulative P/L are compiled automatically for each method, so you can compare the two side by side. For guidance on labeling trades once you have defined where each trade begins and ends, also see the separate article "What Counts as One Trade? How Counting Changes Your Stats." The specific steps for aggregating and analyzing are covered in detail in the separate article "Review Your Trades with AI Analysis."

And one thing worth remembering: these numbers are not for grading a method like a test score and stopping there. Numbers are not only for evaluating trades; they are hints for thinking about how to improve your method next. If your win rate is low, review how you enter; if your losses are large, rethink where you place your stop loss. Using the numbers as a starting point for improvement is the real purpose of recording and analyzing.

Frequently asked questions

If my win rate is high, won't I make money?

A high win rate is one favorable factor, but it alone does not determine your P/L. Depending on the size of each win and loss (risk-reward), you can end up negative overall even with a high win rate. You need to look at both win rate and risk-reward together.

What win rate is good enough?

You cannot judge from win rate alone. Even with a 60% win rate, your P/L can be negative depending on risk-reward, and even with a 40% win rate, you can end up with a profit if your risk-reward is good. Rather than setting a target based on win rate alone, it is important to also look at metrics such as risk-reward and expectancy.

What is the difference between risk-reward and expectancy?

Risk-reward is a number that expresses "the ratio of the size of one win to one loss." Expectancy combines the size of those wins and losses with the win rate to look at "what is likely to happen on average per trade." A simple way to think of it: risk-reward is a ratio of sizes, while expectancy is an overall view that also includes the win rate.

Does positive expectancy mean I will make money?

Not necessarily. Expectancy is a yardstick for evaluating a method from past records and does not guarantee future profits. Market conditions do not necessarily stay the same, and there is no guarantee that past numbers will repeat. It is appropriate to use expectancy as a perspective for looking at a method from multiple angles.

How many trades do I need for the numbers to be reliable?

There is no clear standard, but with only a few trades, results tend to be skewed and the numbers are unstable. The more trades you accumulate, the more meaningful win rate and risk-reward tend to become. Since building up trades through live trading alone takes time, one option is to supplement the number of trades with backtesting on historical charts.

Because win rate is an easy number to understand, it is tempting to look at it alone. Simply adding the perspectives of risk-reward and expectancy changes how a method looks. Start by recording your own trades and looking at your method through several metrics.

※This article is provided for informational purposes and does not guarantee the effectiveness or profitability of any specific method or metric. The figures shown are examples to illustrate the concepts and do not represent investment results. All investment decisions are your own responsibility.

Reproduce this validation yourself

With ENTRIQ's chart replay, you can trade through past charts using the same rules.