Memory is the worst record-keeping system available
Ask any trader how they are doing and you get a story built from the two trades they remember: the big winner and the painful loss. A journal replaces the story with an arithmetic. It tends to reveal something uncomfortable and useful - that the strategy is roughly break-even and one category of trade, usually the impulsive one, is paying for all of it.
Why everything is measured in R
Recording results in dollars makes trades incomparable, because a $400 win taken at four times normal size is not better than a $100 win taken correctly. Expressed in R - multiples of the amount risked - every trade is on the same scale, across markets and across account sizes. A +0.3R average over fifty trades is a real edge. A big dollar month that was secretly +0.1R at triple size is a warning.
The three breakdowns that change behaviour
- By setup: almost everyone has one named setup that is quietly negative. Removing it is usually worth more than any new indicator.
- By weekday and time: results cluster. Thin sessions and news days show up here long before they show up in your intuition.
- By how the trade ended: a pile of manual exits next to very few targets reached means you are closing winners early, which quietly destroys the reward-to-risk you planned.
Review it weekly, not after every loss
Reading a journal immediately after a bad trade produces conclusions written by the loss. Set one weekly slot, look at the group totals rather than individual trades, and change at most one rule per review - otherwise you cannot tell which change did anything.
Where the trades come from
The fastest way to fill this is the Practice Arena: live prices, enforced stops, and every closed trade arriving with the reason you typed before entry. Then price the next one honestly in the risk and reward planner.
The six numbers a journal must produce
| Metric | What it answers | Healthy range |
|---|---|---|
| Expectancy (R) | Is this worth repeating? | above +0.10R |
| Profit factor | Gross wins ÷ gross losses | above 1.2 |
| Win rate | Only meaningful next to R | any, if R fits |
| Average win vs loss (R) | Are you cutting winners early? | win > loss |
| Max drawdown | What you must live through | under 20% |
| Trades | Is the sample real yet? | 30+ to read, 100+ to trust |
If your average loser is larger than your average winner while your win rate is above 50%, you have the shape that ends accounts: lots of small wins paid for by occasional large losses. It works until one trade does not, and the journal shows it long before the market does.
Why R and not dollars
Recording results in money makes trades incomparable, because a $400 win taken at four times normal size is not better than a $100 win taken correctly. Expressed in R - multiples of what was risked - every trade sits on the same scale, across instruments and across account sizes. A +0.3R average over fifty trades is a real edge. A big dollar month that was secretly +0.1R at triple size is a warning.
Review weekly, change one rule at a time
Reading a journal immediately after a bad trade produces conclusions written by the loss. Set one weekly slot, look at group totals rather than individual trades, and change at most one rule per review. Change two and you will never know which one did anything, which is how people spend years adjusting a system without ever improving it.
Keep your own copy
Everything here lives in your browser and nowhere else, which is why we can promise we never see it. The trade-off is that clearing site data erases it, so export to CSV regularly. The fastest way to fill the journal honestly is the Practice Arena, where every closed trade arrives carrying the reason you typed before entry.
Frequently asked
What should a trading journal record?
At minimum: the setup you traded, entry, stop, exit, the amount risked, and the result expressed in R. Add the reason you took it and a one-line note written before the outcome is known. Prices alone produce a spreadsheet; the reason is what lets you group trades later and find the one that is actually losing money.
What is expectancy and why is it the headline number?
Expectancy is your average result per trade measured in R - units of what you risked. It combines win rate and reward-to-risk into the only figure that answers the real question: if I take this trade a hundred more times, am I ahead? A win rate on its own says nothing, and a profit figure says nothing about the risk taken to get it.
Is my journal data private?
Completely. It is stored in your browser using local storage and never leaves the device - there is no account, no sync and no server copy, so we could not read it if we wanted to. The trade-off is that clearing your browser data erases it, so export to CSV regularly.
How many trades before the numbers mean anything?
Around thirty before the shape is suggestive, and closer to a hundred before it is evidence. Below that, a run of luck in either direction dominates the statistics. The breakdowns by setup and weekday need even more, because each one splits the sample further.
Can I log real trades, not just practice ones?
Yes. Use "Log a real trade" to enter entry, stop, exit and quantity by hand - the journal prices it in R the same way and marks it as real, so your practice record and your live record sit side by side without being confused for each other.
Trading is high risk and most retail accounts lose money. This tool is education only, not financial advice.