Trading Metrics That Actually Matter (And the One You Should Ignore)

Win rate is the most tracked and least useful number in trading. Here are the metrics that tell you whether you have an edge, and how to read them.

Daniël Vermes

Tradeflow Editorial Team

Aug 20, 2026

10 min read

Trading metrics are the statistics calculated from your trade history that tell you whether your strategy has a positive expectancy and where that expectancy comes from. The most useful are expectancy, profit factor, average R multiple and maximum drawdown. The most tracked is win rate, which on its own tells you almost nothing.

That last point is worth sitting with, because win rate is the number most traders quote when asked how their trading is going, and it is close to meaningless without context.

Why win rate misleads

A 70 percent win rate sounds excellent. It can also describe an account that loses money steadily.

Take ten trades. Seven win at 0.5R each, three lose at 2R each. That is 3.5R won against 6R lost, a net loss of 2.5R, from a win rate most traders would be delighted with. The pattern is common because it is psychologically comfortable: taking profit early feels responsible and letting losses run feels like giving the trade room.

The reverse is equally common. A 35 percent win rate where winners run at 4R and losers are cut at 1R produces 14R against 6.5R, a substantial profit from a strategy that fails two thirds of the time. Most trend following works this way, and most traders who try it abandon it during a losing streak that was entirely normal.

Win rate is only meaningful next to the average size of your wins and losses. On its own it measures how often you are right, and being right is not the same as making money.

The metrics that matter

Expectancy

The single most important number in your history. It tells you what you can expect to make, on average, per trade.

Expectancy = (win rate x average win) - (loss rate x average loss)

Express it in R multiples rather than currency and it becomes comparable across account sizes and position sizes. An expectancy of 0.3R means every trade you take is worth about a third of your risk unit, on average, across a large enough sample.

That is the number that tells you whether you have a business or a hobby. Positive expectancy plus enough trades equals profit. Negative expectancy plus enough trades equals ruin, however good the individual weeks look.

Profit factor

Gross profit divided by gross loss. A profit factor of 1.6 means you made 1.6 units for every unit you lost.

Below 1.0 is losing money. Between 1.0 and 1.25 is marginal, and likely to be erased by costs and slippage. Between 1.3 and 1.8 is a solid, sustainable strategy. Above 2.5 is either exceptional or a warning that your sample is too small, your period was too favourable, or your risk is concentrated somewhere you have not looked.

Profit factor is the most useful metric for comparing your own setups against each other, which is its real job. Ranked by profit factor, most traders discover one or two setups carry the entire account while three or four quietly break even.

Average r multiple

R is your risk unit: the amount you lose if the trade hits your stop. A trade that makes three times your risk is 3R.

Thinking in R rather than currency solves two problems at once. It makes trades comparable regardless of position size, and it separates the quality of a decision from the size of the bet, which is the distinction most traders' P&L hides.

Your average R across all trades is essentially expectancy stated another way, and tracking it by setup is where the useful information lives.

Maximum drawdown

The largest peak to trough decline in your equity. The first thing any prop firm risk manager looks at, and often the last.

Return without drawdown is meaningless. A 40 percent year with a 35 percent drawdown is leverage and timing. A 20 percent year with a 6 percent drawdown is a repeatable process. The second one attracts capital and the first one does not, regardless of which produced the bigger number.

Track average drawdown and recovery time alongside the maximum. How long it takes you to return to a previous high says more about your process than the depth of the hole.

Risk per trade, and its variance

Not a performance metric, and one of the most predictive numbers in any history.

Consistent risk of 0.5 to 1 percent across hundreds of trades indicates a system. Risk ranging from 0.2 to 6 percent indicates that your results came from a handful of oversized positions, which means your returns are a distribution with a fat tail, and eventually the tail points the other way.

Prop firms reject applications on this alone. It is worth checking on yourself before someone else checks it for you.

Sample size, or why your numbers are probably noise

Every metric above requires enough trades to mean anything, and most traders draw conclusions from far too few.

Around 30 trades in a single category gives you a rough signal. Fifty to a hundred is where a per setup conclusion becomes reasonably reliable. Below 30, you are measuring variance and calling it a strategy.

This is the most common analytical error in retail trading. A setup with a 70 percent win rate across 12 trades tells you nothing. The same win rate across 120 trades is worth building a strategy around. The number looks identical and the two facts are not remotely comparable.

The practical consequence: be much slower to cut a setup than you want to be, and much slower to scale one that has been working for three weeks.

Where the real information is

Aggregate metrics tell you whether you are profitable. They do not tell you why, and why is what you can act on.

The useful move is breaking every metric down by dimension.

By setup. Rank profit factor by setup and compare it to how often you take each one. The setups you take most often are frequently not the ones that pay, because habit and expectancy are unrelated.

By session. Morning versus afternoon, London versus New York. Many traders have an entire session running at negative expectancy and have never isolated it.

By day of week. Sounds superstitious and often is not. Thursday afternoons show up in a lot of trading histories for reasons that have to do with the trader rather than the market.

By hold time. If your winners average four hours and your losers average twenty minutes, you are cutting winners early. If it is the reverse, you are letting losers run.

By state. If you have logged how you felt, group by it. Nearly every trader has a state that reliably costs money and cannot name it without data.

By combination. This is where the edge usually hides. A setup that is marginal overall might be strongly positive in the morning and strongly negative in the afternoon, which averages out to nothing and looks like a mediocre setup rather than two different results wearing one label.

That last one is also the reason most traders never find it. Five setups, three sessions and four states is sixty combinations, and nobody checks sixty combinations in a spreadsheet.

Getting these numbers without building them

Every metric here is arithmetic on data your broker already recorded. The obstacle is never the maths, it is the assembly: exporting, cleaning, categorising, and then rebuilding the whole thing every time you want to look at a different cut.

In Tradeflow the metrics calculate themselves from synced trade data, and you get at the breakdowns by asking rather than by building filters. Show me profit factor by setup for the last ninety days. Which session has the worst expectancy. Break down my R multiples by hold time.

The combination question is the one worth trying first, because it is the analysis nobody does by hand and where the answers tend to be most surprising. Conversational AI is on every plan from $19 a month.

What to do with the numbers

Metrics are only useful if they change something, and the temptation after reviewing them is to change four things at once. Do that and you learn nothing, because you cannot attribute the result.

One change per review. Written down, tracked the following week. The setup that has been below break even across fifty trades gets a rule attached or gets cut. The session running negative gets removed from your schedule. The hold time pattern gets a mechanical fix, such as a minimum time in trade before you are allowed to close a winner.

We covered the process in the 20 minute weekly trading review.

Key takeaways

Win rate on its own is close to meaningless. A 70 percent win rate can be a losing strategy and a 35 percent win rate can be a strong one. Expectancy in R multiples is the single most important number in your history. Profit factor between 1.3 and 1.8 is a solid strategy. Above 2.5 usually means your sample is too small. Return without drawdown is meaningless. Prop firms and investors look at drawdown first. Inconsistent risk per trade is the fastest way to get rejected by a firm, and it is visible in your own data. Around 30 trades per category gives a rough signal, 50 to 100 for a reliable one. Below that you are reading variance. Aggregate metrics tell you whether you are profitable. Breakdowns by setup, session and state tell you why.

Frequently asked questions

What is the most important trading metric?

Expectancy, expressed in R multiples. It tells you what you can expect to make per trade across a large sample, which is the number that determines whether your strategy is a business or a hobby.

What is a good profit factor?

Between 1.3 and 1.8 is a solid, sustainable strategy. Below 1.0 is losing money and between 1.0 and 1.25 is likely to be erased by costs. Above 2.5 is possible but usually indicates a small sample or a favourable period rather than a superior strategy.

Is a high win rate good?

Not on its own. A 70 percent win rate is a losing strategy if your average loss is four times your average win. Win rate only becomes meaningful next to the size of your wins and losses.

What does R multiple mean?

R is the amount you risk on a trade, meaning the loss you take if the stop is hit. A trade that returns three times that amount is 3R. Thinking in R makes trades comparable regardless of account or position size.

How do you calculate expectancy?

Multiply your win rate by your average win, then subtract your loss rate multiplied by your average loss. Expressed in R multiples, a result of 0.3 means each trade is worth roughly a third of your risk unit on average.

How many trades do I need before my metrics mean anything?

Around 30 in a single category for a rough signal and 50 to 100 for a reliable one. This is the most common analytical error in retail trading. A 70 percent win rate across 12 trades tells you nothing.

Which metrics do prop firms look at?

Maximum drawdown first, then consistency of risk per trade, then trade count and duration. Returns matter considerably less than the shape of the path that produced them.

Should I track metrics by setup or overall?

Both, but the breakdowns are where the actionable information lives. Overall metrics tell you whether you are profitable. Per setup and per session metrics tell you what to cut.

See your real numbers

Tradeflow syncs every trade from 600+ brokers and platforms and calculates your expectancy, profit factor, R multiples and drawdown automatically. Then break them down by setup, session or anything else by asking.

Start free for 7 days. From $19 a month afterwards.