Learn how to read a trading backtest properly. A beginner-friendly guide to win rate, profit factor, number of trades, expectancy, and drawdown.
If you are new to trading, it is very easy to look at a backtest and focus on the wrong number. This guide breaks down the core metrics every beginner should understand before evaluating any strategy.
If you are new to trading, it is very easy to look at a backtest and focus on the wrong number.
A strategy might show an 80% win rate and still be terrible. Another might only win 42% of the time and still be very solid. A third might show an excellent profit factor, but only from 17 trades, which tells you almost nothing.
This is one of the biggest problems beginners run into when they start exploring trading strategies. They open a backtest report, see a few attractive numbers, and assume they have found an edge. In reality, they often have not learned how to interpret the results yet.
The goal of this article is to fix that.
By the end, you should understand:
If you are completely new to trading, this is exactly the kind of foundation you want before falling in love with a strategy.
A backtest is not just a scorecard. It is evidence.
That evidence is only useful if you understand what the numbers mean and how they relate to each other.
Most beginners naturally want a simple answer. They want to know whether the strategy is "good" or "bad." The problem is that trading performance is not captured by one number.
A strategy can have:
That is why I think beginners should start with a small group of core metrics rather than trying to interpret everything at once.
The first three I would focus on are:
Then, once you understand those, you should add:
Win rate is the percentage of trades that closed in profit.
If a strategy takes 100 trades and 58 of them are winners, the win rate is 58%.
That sounds simple, and it is. The mistake is assuming that a higher win rate automatically means a better strategy.
It does not.
A strategy can win very often and still lose money overall if the losing trades are much larger than the winning trades.
Imagine two systems:
At first glance, Strategy A looks better because 80% sounds impressive. But once you look at the size of the wins and losses, the picture changes very quickly.
That is why win rate on its own is not enough.
Win rate is useful because it tells you something about the rhythm of the strategy.
For example:
So win rate matters, but it only matters in context.
Profit factor is one of the most useful backtest metrics because it compares total gross profit to total gross loss.
Profit Factor = Gross Profit รท Gross Loss
So if a strategy made $20,000 in winning trades and lost $10,000 in losing trades, the profit factor would be 2.0. That means the system made two dollars for every one dollar it lost.
A profit factor of:
But again, context matters.
A very high profit factor from a tiny sample of trades is not very convincing. A more modest profit factor from a large, realistic sample can often be much more trustworthy.
Profit factor forces you to look at the total balance between wins and losses.
That is important because it helps answer a much better question than "How often does it win?"
It asks:
> "Does the strategy make more from its winners than it gives back through its losers?"
That is a far more useful lens.
This is one of the most overlooked metrics for beginners.
A backtest with only a small number of trades may tell you almost nothing.
If a strategy shows:
that might look exciting, but it is very weak evidence.
Why?
Because with very few trades, randomness can dominate the result.
A small sample can make an average strategy look amazing or a good strategy look terrible. It simply does not give you enough information.
The more trades a strategy takes, the more confidence you can usually have that the results are showing something real rather than just a lucky patch.
That does not mean "more trades" automatically makes a strategy good. It just means the evidence becomes more meaningful.
For beginners, this is a crucial shift in mindset:
> You are not just looking for a profitable result. You are looking for a result that has enough evidence behind it to be worth taking seriously.
If two backtests both look promising, but one is based on 35 trades and the other is based on 600 trades, the second one deserves far more attention.
That does not prove it will work live. But it gives you a much stronger foundation for analysis.
This is the main lesson.
A backtest becomes much more useful when you stop reading one number in isolation.
You want to read the metrics together like this:
That combination already gives you a much better read on a strategy than win rate alone.
At first glance, this looks excellent.
But 22 trades is a very small sample. The strategy may still be interesting, but you should treat it as an idea that needs much more testing, not as a proven edge.
A beginner might dismiss this immediately because the win rate is below 50%.
That would be a mistake.
This system has:
It may not be perfect, but it is far more credible than the first example.
Once you understand win rate, profit factor, and trade count, the next metric I would learn is expectancy.
Expectancy answers this question:
> How much does the strategy make, on average, per trade?
This is one of the clearest ways to think about whether a system actually has an edge.
A strategy with positive expectancy should, in theory, make money over a large enough sample if traded consistently under similar conditions.
That is why expectancy is so important. It takes you beyond "this looks good" and into "this appears to have a statistical edge."
You do not need to become obsessed with the formula immediately as a beginner. What matters first is understanding the idea:
That is one of the biggest reasons why win rate can be so misleading on its own.
A strategy can be profitable and still be completely unsuitable to trade.
That is where drawdown comes in.
Drawdown shows how much the strategy fell from a previous equity peak during the backtest.
This matters because traders do not experience a backtest as a spreadsheet. They experience it as a sequence of wins, losses, frustration, and uncertainty.
A strategy might have decent profit factor and good long-term returns, but if it suffers brutal drawdowns along the way, many traders will not be able to follow it properly.
So once you have checked:
you should also ask:
This is where the difference between a strategy that looks good and a strategy you can actually trade starts to become very real.
A high win rate feels safe, but it can hide terrible risk/reward.
A strong result with too few trades is not strong evidence.
Net profit matters, but it does not tell you how the strategy behaved to get there.
A strategy that is technically profitable may still be untradeable for you in practice.
One clean backtest is not enough. You need broader evidence, realistic assumptions, and proper testing.
If I were helping a beginner interpret a backtest, I would tell them to ask these questions in order:
1. Is it profitable at all? Start simple. Is the result above break-even after realistic costs?
2. How many trades are in the sample? If the trade count is very small, be careful. You may not have enough evidence yet.
3. What is the profit factor? This helps you see whether total gains are meaningfully outweighing total losses.
4. What is the win rate, and does it fit the strategy style? Do not judge it in isolation. Judge it alongside average win and average loss.
5. What does the drawdown look like? Could you realistically live through that?
6. Does the result make sense? Sometimes a backtest looks amazing for the wrong reasons: overfitting, unrealistic fills, too many filters, or rules that would be difficult to execute live.
That last point matters a lot. A beautiful backtest is not always a truthful one.
If you are new and want a simple way to think about backtests, use this:
| Metric | What it tells you | |---|---| | Win rate | How often it wins | | Profit factor | Whether the wins meaningfully outweigh the losses | | Number of trades | Whether you have enough evidence to trust the result at all | | Expectancy | Whether the system appears to have an edge on average | | Drawdown | How painful the strategy may be to trade in real life |
That is a much more useful framework than chasing whichever number looks best.
This is worth saying clearly.
A good backtest does not guarantee future profits.
It only tells you how a defined set of rules would have performed on historical data under the assumptions of the test.
That still makes backtesting extremely useful. But it is useful as evidence, not as certainty.
The real purpose of a backtest is not to prove that a strategy will make money forever. It is to help you answer better questions, such as:
That is a much healthier way to use historical results.
The biggest mistake beginners make with backtests is looking for one magic number.
There is no magic number.
A good backtest is a combination of:
If you take nothing else from this article, remember this:
> Win rate tells you how often. Profit factor tells you how well. Number of trades tells you how much evidence you have.
Read those together, not separately.
That simple shift will already put you ahead of most beginners.
If this article helped, the next useful step is to learn how backtests can go wrong even when the numbers look attractive.
Recommended next reading: