Read your stats without fooling yourself
A dashboard is a mirror with opinions. The order you read it in decides whether you learn something or just find a number that agrees with you.
You leave with A reading order that resists self-flattery
Start at expectancy, not win rate
Expectancy is the average result of one trade, losses included. It answers the only question that compounds: does taking this trade a hundred times make money? Read it per setup, not per account. An account is a portfolio of habits; setups are the habits.
Then look at the shape, not the total
Two equity curves can end at the same number, one built from steady singles and one from a lucky outlier. The R distribution tells them apart. If the month lives on one trade, you do not have an edge yet; you have a story.
Cross it with discipline
- Split every stat by followed-plan versus broke-plan. The gap is the price of your impulses.
- Rank setups by discipline score before P&L: the leak is usually a setup you like more than it likes you.
- Check the time axis: most accounts have one hour of the day that quietly pays for the rest.
The number you want to see is the one to distrust first.
Let the sample size veto everything
Under thirty trades per setup, treat every conclusion as a rumor. reps keeps some panels locked until the sample is real for exactly this reason: an early number is temptation dressed as information.
Done if
- You know your per-setup expectancy, or know the sample is too small to trust.
- You found the setup where discipline, not edge, is the leak.
- Every under-thirty-trades conclusion is filed as a rumor.