Fifteen trades into a new setup. Twelve winners. You've already worked out what doubling your size, the amount you risk on each trade, would look like.
Before you do that, it helps to know how easily fifteen trades can fool you, even when the setup is good.
Can fifteen good trades be luck?
Yes, and more often than feels right.
Flip a coin fifteen times. Most of the time you get something close to half heads. Now and then you get twelve. Nothing is wrong with the coin. Fifteen flips is simply too few to tell a fair coin from a slightly bent one.
Your trades work the same way. A setup that only breaks even over the long run can still hand you twelve winners out of fifteen on a good stretch. A setup that really works can hand you a string of losers. Over a short run, luck is louder than the setup.
There's a second way a small group fools you. Say you're up after fifteen trades, but most of that came from one big winner. Take that one trade out and you're flat. Fifteen trades isn't enough to know whether winners like that come around often, or whether you got lucky once.
So what is the number?
There isn't one. Anyone who gives you a single round number is guessing.
The shape of it is this. A few dozen trades give you a hunch. A hunch is worth watching, and not worth trading bigger on.
Somewhere well past a hundred trades, all taken the same way, the picture gets clear enough to lean on. How far past depends on the setup, and the next two sections are about that.
Why so many? Because trading results jump around from one trade to the next far more than most people expect. In a small group, one big winner or a short run of losses can drag the total a long way. The more trades you have, the less any single one can move it.
What pushes the number up?
How many ideas you tried before this one. Say you tried ten ideas this year and kept the one that looked best. Part of why it looks best is luck: one of the ten was always going to come out on top, even if none of them work. So a setup you picked from a long search needs more trades than one you only ever tried once.
Nobody counts this. You test a new entry, then a new filter, then a different time of day, and by Friday you've forgotten Tuesday's idea existed. The honest number of ideas you tried is nearly always higher than the one you remember.
A small edge. An edge is a way of trading that comes out ahead over many trades. If your setup makes only a little more than it loses, that small difference hides inside the normal ups and downs for a long time. A big edge would show itself sooner, but big edges are rare.
What pulls the number down?
A clear rule, written down before the trades. If you wrote down what the setup is before you started counting, every trade is a test of the same thing. If you worked out the rule by looking back at trades you'd already taken, you're partly describing luck that already happened.
Taking it the same way every time. Same entry, same stop (the price where you get out at a loss), same way of handling the trade once you're in. Each trade taken a little differently means you're testing several slightly different setups, each on a handful of trades.
Why does filtering your journal make it worse?
Say you have two hundred trades in your journal. That's a decent number.
Now you filter. Only the London session, the hours when the London market is open. Only the days when the bigger chart was moving mostly one way. Only the trades you graded A+, your own mark for how clean the setup looked. Each filter sounds sensible. By the end, two hundred trades have become nine.
Nine trades can say anything. If they look amazing, you've learned very little.
The A+ filter is the sneaky one. If you gave the grade after you saw how the trade went, the grade already knows the result. Of course those trades look good. A grade only tells you something if you gave it before the trade played out.
Every filter feels like you're getting more precise. You're also getting fewer trades, and that matters more.
What do you do while you wait?
Keep your size small. Trade the setup at a size where being wrong for the next few dozen trades doesn't hurt.
Keep logging. Every trade, the same way, including the boring ones and the ones you'd rather forget. A journal that only holds the trades you're proud of will tell you what you want to hear.
Don't change the rules halfway. Widen your stop a little partway through and you no longer have one pile of trades. You have two smaller piles, one for each version, and each is too small to say anything. If you want to try a change, write it down as a new version and count it on its own.
A setup that worked for a while and then stopped looks a lot like a normal losing stretch. Why your setup stopped working shows how to tell them apart.
Questions people ask next
Is thirty trades enough to test a strategy?
Thirty trades is enough to see whether you can follow your rules. It is rarely enough to know whether the strategy works. Treat thirty good trades as a reason to keep going, not as proof.
Do backtested trades count?
They help you learn the setup, and you can count them separately. Trades you took with real money, as they happened, are the stronger test, because in a backtest you already half know how the chart ends.
Should I size up after a winning streak?
Not because of the streak alone. A streak is exactly what luck looks like over a short run. Size up slowly, after many trades taken the same way, not after a good week.
What if I never get enough trades?
Then look for ways to get more of the same trade: a second market that behaves alike, or a slightly looser rule that you write down and test as its own version. Or accept a slower answer. A slow honest answer beats a fast lucky one.
Is "too early to tell" a real answer?
Yes. It might be the most common honest answer in trading.
A setup can be real and still not provable yet. Every setup starts there. "Too early to tell" is better than a confident yes that turns out to be luck, and much cheaper than sizing up on one.
For what having an edge means in the first place, see what a trading edge is. Once you have enough trades, turning your journal into a trading plan shows how to write down what worked and test it on the trades that follow.
EdgeFlow puts the number of trades next to every line of your plan and says when it's too early to tell, instead of showing a percentage that looks more certain than it is.