Compare
Put two sets of trades side by side and see whether the difference is real or chance.
Compare
Compare puts two sets of your closed trades side by side: long against short, this month against last month, one strategy against all other trades, or any two filters you define. It shows each set's numbers, the difference per trade, how likely that difference is to be chance, and both equity curves on one chart.
The page is in Analytics → Reports → Compare. Both sets come from your active account scope and filter, and follow your P&L display mode, breakeven threshold and timezone. The chosen comparison and anything you pick for it are part of the page address, so you can bookmark or share it.
Choose a comparison
The top row holds the controls. Pick a comparison from the Compare menu; next to it sits the one control that comparison needs, if any:
- Long vs short and This period vs previous need nothing else.
- Strategy vs the rest, Tag vs the rest and Symbol vs the rest add a menu of up to 50 strategies, tags or symbols that have trades in the view, most traded first. The page starts with the first one, so it always shows a result; pick another to compare it instead. When no trade in the view has a strategy or tag, the menu reads "None in this view" and the page says so.
- Before vs after a date adds a date picker. It starts at the first day of the current month in your timezone.
- Custom adds two buttons, one per set, labelled with what the set holds, such as "A: Long · ES" or "B: All trades in this view". Each opens the same filter controls as the global filter for that set.
On the right of the row, the key explains the colors and How to read this page opens this article.
Comparisons
| Comparison | Set A | Set B |
|---|---|---|
| Long vs short | Long trades | Short trades |
| This period vs previous | Trades in your date range | Trades in the period before it |
| Strategy vs the rest | Trades in the chosen strategy | Every other trade in the view |
| Tag vs the rest | Trades with the chosen tag | Every other trade in the view |
| Symbol vs the rest | Trades in the chosen symbol | Every other trade in the view |
| Before vs after a date | Trades before the chosen date | Trades from that date on |
| Custom | Trades matching the filters of set A | Trades matching the filters of set B |
This period vs previous reads the date range of your filter. A calendar range such as this week, this month or this year is compared with the whole previous week, month or year. Today is compared with yesterday. The last 30 or 90 days and a custom range are compared with the same number of days right before them. Without a date range, this month is compared with last month. Days are calendar days in your timezone.
Before vs after a date splits your date range at the start of the chosen day, so both halves stay inside it.
Custom sets start empty, so both hold every trade in the view until you edit them. A set's filters narrow the view; a trade that matches both custom sets counts in both. The page then says how many trades the sets share, and the confidence goes no higher than Early, because shared trades make the two sets look more alike than they are.
Date ranges
A set with its own start or end date, such as the two periods or a custom set with dates, replaces both bounds of the global date range for that set. A set without its own dates keeps the global range. Every other filter, such as symbols, tags or weekdays, still applies to both sets. A note under the page says when a set's range replaced the global one.
Read the page
The page leads with the answer in one sentence, such as "Long makes $42.10 more per trade than short trades.", above three figures: set A's average P&L per trade, set B's, each with its number of trades, and the difference per trade, set A minus set B, with its plausible range under it. The difference is the number the confidence rates; it is in green or red only when the confidence is Likely or Confirmed. Under the figures, short notes say when the sets share trades, when a set has too few trades, or when the result rests on one large trade. The confidence follows in words with both sets' trade counts.
Beside the answer, the chart draws both sets' cumulative P&L, each starting at zero. Set A is the solid line, set B the dashed one, both named above the plot. When the two sets overlap in time they share the calendar. When they do not, such as this month and last month, each curve starts on day 1 so the two lie over each other. A line under the chart says which of the two it is.
Side by side lists each set's net P&L, win rate, profit factor, average per trade, average win and loss, average hold time and, where trades have them, costs, average MFE and average MAE. Each column header names the set and how many trades and days it covers; for periods, a split date and custom sets it also states what the set holds, such as its date range. The difference column shows set A minus set B on every row, in the row's own unit: win rate in percentage points (pp), profit factor as a plain number, hold time as a duration. Only the average per trade is tested against chance, so only its difference is emphasized and colored.
Confidence
Compare rates the difference with the same rules as Habits:
| Label | Meaning |
|---|---|
| Not enough trades yet | One of the sets has fewer than 5 trades with a P&L value. |
| Early | The difference points one way but could still be chance, or it depends on a single large trade. |
| Likely | The difference is unlikely to be chance and covers at least 10 trades. |
| Confirmed | The difference covers at least 20 trades on at least 10 trading days and does not depend on its largest trade. |
| No clear difference | The two sets' results per trade are in line with each other. |
When the result rests on one large trade, the page says so under the answer.
When there is no answer
- No trades in this view have a strategy. (or a tag) appears when a "vs the rest" comparison has nothing to compare. Choose another comparison.
- Neither set has closed trades. appears when both sets are empty. Choose another comparison or widen the filter.
- Not enough trades to compare yet. appears when a set has fewer than 5 trades. The figures, table and chart still show what the sets hold.
- If a refresh fails, the page keeps the previous comparison with a Retry action.
Limits
- A difference describes your history. It does not predict how the next trades will go.
- In tick and point modes, only trades of your main instrument class carry a P&L value. A note on the page says when this applies and how many trades of each set are left out.
- Win rate, profit factor and the other figures are shown for both sets, but only the difference per trade is tested against chance.