Win rate by setup: how to find out which of your setups actually works

A short, fixed tag list and four numbers per setup, with the sample-size and moving-label traps to avoid.

A trading journal earns its keep when it lets you ask a pointed question: which of my setups actually works for me? Not which one feels best or which one I read about, but which one shows an acceptable result across the trades I really took. The answer needs two ingredients, a consistent label on every trade and enough trades under each label to say something. This page explains how to tag trades by setup, how to compute win rate and average result per setup, and how to avoid the two classic mistakes: tiny samples and moving labels.

Start with a short, fixed list of setups

The most common reason setup analysis fails is a messy tag list. If you invent a new name for every trade, no group ever gets big enough to read. Choose a short list, perhaps four to eight names, each defined in one sentence of plain language: what you need to see on the chart for the label to apply. Write the list down, keep it stable for at least a quarter, and add an Other label for trades that fit none. More on choosing labels is in how to tag trades in a trading journal.

Tag at the time, not months later

Tagging from memory biases the result, because you remember a trade as the setup that would have explained its outcome. Tag when you open or close the trade, while the reason is fresh, and record it before you see the result if you can. The journal's setup field and notes are meant for exactly this, and screenshots taken at the moment add evidence. If you have a backlog, tag from your notes and screenshots, and mark older entries as lower confidence.

What to calculate for each setup

MeasureHowWhy
TradesCount entries with that setupSample size comes first
Win rateWinners divided by tradesContext; ignores size
Average winMean of winning resultsPart of the payoff picture
Average lossMean of losing resultsPart of the payoff picture
Expectancy per tradeNet result divided by tradesOne figure that combines the above
Plan followedShare of trades marked as following the planSeparates the setup from your discipline

Expectancy per trade is simply the average net result per trade for that setup, after swap and commission. For the definitions, see win rate, profit factor and expectancy. A setup with a modest win rate can still be the best performer if its average win is much larger than its average loss, which is why win rate alone should never decide the ranking.

A worked example (hypothetical)

These numbers are invented to illustrate reading the table. They do not describe any real strategy.

SetupTradesWin rateAverage winAverage lossExpectancy
Setup A4652%+14.00-11.00+2.02
Setup B1867%+6.00-15.00-0.96
Setup C580%+20.00-9.00+14.20
Other2236%+9.00-12.00-4.44

Setup B has the best win rate among the established setups, yet it loses money on average, because its wins are small and its losses are large. Setup C has the best expectancy of all, but five trades is far too few to build anything on; a couple of lucky trades could explain it. Setup A is the quiet workhorse: a decent sample, a coin-flip win rate and a positive average. The Other row is the most telling. If a fifth or more of your trades fit no defined setup and they lose on average, the cleanest improvement may be to stop taking them, which is a question about discipline rather than about any particular pattern.

Avoiding the two classic mistakes

Cross-check with the rest of the record

A setup does not exist in a vacuum. A setup that works only on one symbol, in one session or in one month is a narrower finding than one that works across several. Cross-check against history by symbol, history by hour and monthly results. Also look at the emotion field: if one setup is mostly taken when you feel rushed, the result may reflect your state, not the pattern.

AI Trading Journal is a Windows app for MetaTrader 5 only. It reads closed trades from the terminal on your PC, adds new ones every 5 minutes, and keeps the journal in a file in your Documents folder. Each journal entry can carry a setup, an emotion, a plan-followed flag, a rating from 1 to 5, tags, notes and screenshots, and the Journal page supports search and filters, so you can pull up all trades with one setup. The Excel report lists the trades with live formulas; check its Trades tab in your version for the columns you need to build the table above. The app is read-only: it never places, modifies or closes a trade, and it gives no trading advice. The 7-day trial needs no card, and the full version is $19.99 once, with no subscription.

Limits

Frequently asked questions

How do I find my best setup?
Tag every trade with one setup from a short fixed list, then compare trades, win rate, average win, average loss and expectancy per setup, with the count beside each result.
Is win rate enough to rank setups?
No. A high win rate with small wins and large losses can still lose money. Use average result per trade as well.
How many trades per setup do I need?
There is no magic number. Decide a minimum before you look and treat anything below it as inconclusive.
Can I change a setup definition later?
You can, but start a new count, because the old and new trades are no longer the same thing.
Does the app support setup tags?
Yes. Each entry has a setup field, tags, notes and screenshots, and the Journal page can search and filter by them.

Related guides

Tag your setups: free for 7 days

Free for 7 days, no card required. After that, a one-time $19.99 purchase. No subscription. Also on the Microsoft Store.

One-time $19.99, no monthly fee · your MT5 login never leaves your PC · trades go into your own journal on your PC, not through our servers.

Start free 7-day trial