Edge Discovery: Find Where Your Trading Actually Works
Edge discovery means looking through your own trades to find what made you money. Which combination of the confluences you look at, in which kind of market, with which way of getting in and out? One overall win rate cannot tell you that. Splitting your trades into groups and comparing them can, as long as each group holds enough trades.
Why is one overall number not enough?
Say your win rate, how often you win, is fifty percent. That is a summary. It hides that you might win most trades in the morning and lose most in the afternoon, or do well when the bigger chart is trending and badly when it is not. Edge discovery is about finding those differences in your own trades.
Step one: decide which trades count
Start by choosing which trades to look at. Real-money trades only, or also forward-test trades (new trades you took as a test, often with small size) and backtest trades (trades you worked out on old charts)? Practice trades can change the picture, so you choose. You also see how many of your trades have a result and how many of your confluences you filled in.
Step two: find your core setup
Your core setup is the small group of confluences that has to be there before you take a trade. EdgeFlow shows combinations of your setup boxes that really came up together in your trades, with how each one did. You pick one, or build your own. You can remove one box at a time and see what happens to the rest without it.
Step three: where does it work?
Next, the same core setup is split by the market around the trade: trending or sideways, the session, the time of day. The groups sit side by side. For each one you can leave it out of your plan, keep it as its own branch, or leave it alone, and you can always undo that.
Here is an everyday example. A setup might win often when the bigger chart is trending and lose often when it is sideways. Over all trades together it looks average. Split, it turns out to be a good setup that only belongs in one kind of market.
Step four: how you got in and what you did after
Then come your entry, your stop (the price where you get out at a loss), your target (where you planned to take profit) and what you did once you were in. If you traded the same setup in more than one way, the ways sit next to each other with their numbers.
Step five: one plan on one page
What you chose becomes one checklist: the setup, where it works, how you get in, what you do after. Then you freeze it, so its numbers stop moving.
Step six: test it on trades it has not seen
A pattern found in your past trades will always look good on those same trades, because that is where it came from. The fair test is trades taken after you froze the plan.
Those new trades are compared with the rest of your trading in the same weeks, so an easy month for the whole market does not flatter the plan. The answer is one of three: the plan is ahead of your other trades, it is too early to say, or it has stopped working.
Why not draw conclusions from six trades?
Because six trades can say anything. Flip a coin six times and five heads is nothing special. Every group in EdgeFlow shows how many trades it holds, and a group that is too small is marked as too early to tell. A small group is an idea to test, not a result. How many trades before you can trust a setup explains why, and why your setup stopped working covers what to check when a plan stops working.
Common questions
What is edge discovery in trading?
Edge discovery is looking through your own trades to find which combination of conditions, in which kind of market, made you money. You split your trades into groups that share something and compare how each group did.
How many trades do I need for edge discovery?
You can start looking with a few dozen, but treat what you see as ideas to test. Results become something to lean on only when each group holds many trades, usually well past a hundred in total.
Can I use imported trades?
Yes, once you have ticked the confluences you looked at on each one. An import brings in prices and results, but not your confluences, so those trades cannot be compared until you fill them in.
What does it mean when EdgeFlow says too early to tell?
It means a group has too few trades for its result to mean much. A handful of trades can look great or awful by luck alone. Keep logging and look again later.
Does edge discovery guarantee a profitable plan?
No. It shows what happened in the trades you logged and lets you test a plan on new trades. If nothing in your trades stands out, it shows you that too.
Related reading
Guides on this topic
- Best Trading Journals for Finding Which Setups WorkWhich trading journal shows which of your setups work? What to look for, and how EdgeFlow, Edgewonk, Tradezella, TraderSync, Tradervue and TradesViz compare.
- How to Backtest a Strategy by Hand, Without CodeYou can backtest without code: write your rules, scroll back on a chart, hide the future, and log each test trade the same way as a live one. Step by step.
- Why Your Setup Stopped Working (and What to Check First)A setup that worked for months and now loses has three likely causes, each with its own fix. How to tell them apart using your own journal.
- How Many Trades Before You Can Trust a Setup?Fifteen good trades can be luck. How many trades you need before a setup means something, what pushes that number up, and what to do while you wait.
See which of your confluences made money
Log your trades, tick the things you checked before you entered, and see which of them did the work in your own trades.