Browse Metrics & analysis

Why two pages show different win rates

Two VEXA screens can show two different win rates for the same trades. Both are right. Here are the two formulas and how to tell which one you are reading.

Same on every platform Updated
On this page

What this does

VEXA shows a win rate in several places. Two of them divide by different numbers, so the same set of trades can produce two different percentages. Neither is a bug. They answer two different questions.

The two formulas

Win rate on the Metrics screen is the share of all your closed trades that made money.

win rate = winning trades / all closed trades x 100

Break-even trades stay in the denominator. They did not make money, so they pull the number down.

Win rate on the Trades screen is the share of your decisive trades that made money.

win rate = winning trades / (winning trades + losing trades) x 100

Break-even trades are removed from both sides. They are neither a win nor a loss, so they are dropped from the question entirely.

Open Profile → Trading settings before comparing figures. With settings off and complete fee evidence, recorded results normally use net P&L. Show gross P&L changes the displayed total before known recorded fees; Net P&L remains after fees. Default fees apply only where permitted, and a configured breakeven band can classify a small nonzero result as breakeven. Read any Provisional or Unavailable state: missing fee evidence is not zero fees. A broker-authoritative net figure can be known while its gross figure and fee breakdown remain unavailable.

The formulas below assume complete figures and matching classifications. With the breakeven band off, positive is a win, negative is a loss and zero is breakeven. When the band is on, results inside it can be breakeven even when nonzero.

A worked example

With Trading settings off and complete figures, take ten closed trades: six wins, three losses and one exactly-zero breakeven.

ScreenSumShown
Metrics6 / 1060.0%
Trades6 / (6 + 3)66.7%

Both describe the same ten trades. The Metrics figure answers “how often did I make money when I traded?”. The Trades figure answers “when a trade resolved one way or the other, how often was I right?”.

For this complete, matching cohort, the denominator difference disappears when there are no breakeven trades. Settings, incomplete evidence or differing cohorts can cause other differences.

Which screen uses which

Where you see itDenominatorBreak-even counts as
Metrics screen, Win Rateall closed tradesnothing, but it stays in the denominator
Trades screen, Win Rate tilewins plus lossesexcluded from both sides
Journal day card, Win Rateall trades that daynothing, but it stays in the denominator
Journal day card, W/L tilenot a rate, a pair of countsa loss
Dashboard, Performance Snapshotall closed tradesnothing, but it stays in the denominator

The Trades screen exists on the web, iPad and Android tablet. On iPhone and Android phone there is no separate Trades screen, so the pair you are most likely to compare is the Metrics win rate against the Journal day card.

What you should see

Two percentages that differ by a few points, with the Trades figure the higher of the two. The size of the gap tracks how many break-even trades you have: one break-even in ten trades moves it by a few points, none moves it at all.

If the gap is large, count your break-even trades. Filter the Trades screen by Status Break Even and read the result count. That count explains the denominator difference only when the settings, evidence and trade cohort also match.

If it doesn’t work

  • The gap is much larger than your break-even count explains. Check Trading settings and unavailable outcomes, then whether the two screens include the same trades. Check the date range and the account filter, which are shared across pages and persist between visits. See Date range, account and advanced filters.
  • One screen is stale. Trade data is cached for five minutes. See Refresh, caching and stale numbers.
  • The trade counts differ as well as the rate. A trade held over several days is filed under its entry date in the Journal and its exit date everywhere else. See How VEXA calculates every metric.

Related

Still stuck? Email support@vexatrade.ai.