Every metric on your dashboard, and how it works

Profit factor — gross win over gross loss

Gross profit divided by gross loss over the window. Above 1.0 means the system pays for itself; the honest band is roughly 1.1 to 3.0 — below that the system loses, and sustained values above 3 are usually a martingale or curve-fit fingerprint, which is exactly what the automated health checks look for. It pairs with win rate: a 40% win rate with a 2.5 profit factor is a trend-follower; the same factor with 90% is something else entirely.

Win rate and expectancy — the pair that matters

Win rate alone flatters: 90% winners means nothing if the losses are ten times the wins. Expectancy completes it — the average result per trade, expressed in R (multiples of your initial risk) and in dollars. A +0.15R expectancy means each trade, win or lose, contributes fifteen cents of expectancy per dollar risked on average. Compounded over a year of activity, that is the edge; everything else on the dashboard is its shadow.

Max drawdown — relative, absolute and floating

The deepest peak-to-trough fall on the closed equity curve, shown as a percentage and in dollars. Siddani also samples equity including floating positions and draws that curve separately — closed-trade drawdown hides open pain, and the gap between the two curves is exactly where grid and martingale accounts hide their risk. The vetting panel flags when floating pain runs far past the closed record.

Sharpe — daily returns, annualized

The mean daily return divided by its standard deviation, annualized. It answers "how smooth was the ride to the return" — and it is checked for coherence against CAGR: spectacular growth with a weak Sharpe rarely belong together, and the health check flags exactly that combination.

Sortino and Calmar — the downside cousins

Sharpe penalizes all volatility — including the upside kind that makes you money. Sortino replaces total volatility with downside volatility only: the deviations that actually hurt. It is computed alongside Sharpe (annualized from the same daily series) and the two are read together — a high Sortino next to a modest Sharpe means the volatility you carry is mostly on the profitable side. Calmar completes the trio: CAGR over max drawdown, growth per unit of worst pain. All three are shown in Deep Analysis; none of them can be gamed by frequency the way win rate can.

Recovery factor and the ulcer index

Recovery factor is net profit divided by the maximum dollar drawdown — how many dollars of profit each dollar of worst pain bought. Above 10 with a short history is usually still ahead of you. The ulcer index is the root-mean-square of drawdowns: it punishes drawdowns that are deep and long, which a single max-drawdown number cannot see.

Z-score — are your wins and losses independent

A streak-dependency statistic over the win/loss sequence: near zero means outcomes behave independently; strongly negative means wins and losses cluster (streaks), which changes how you should size and when you should stop. |z| above 2 gets flagged by the health checks.

CAGR and time under water

Compound annual growth rate, suppressed entirely when the record spans under a month — annualizing noise is how fake track records are born. Time-under-water shows the longest stretch below an equity peak and where you are now: the emotional cost of a system that the win rate never shows.

Size escalation — the martingale detector

Average position size after a loss versus after a win, normalized by each symbol's median size so cross-market averaging stays fair. Ratios above 1.3 mean you press after losses; above 1.8 is escalation. The health check reads it directly from your fills — no self-reporting involved.

MAE and MFE — entry and exit quality

For every closed trade, the venue candles covering its life are fetched and the extremes recorded: how far it went against you (MAE) and how far it ran your way (MFE), direction-aware and R-normalized with the same risk baseline as the trade itself. Entry efficiency, exit efficiency and the scatter of every trade turn "cutting winners early" from a feeling into a number.

Monte-Carlo — the shape of your edge

The Risk Lab resamples your own R-sequence hundreds of times into simulated runs: the probability of touching 5%, 10% or 20% drawdown, median outcome, and the fan of first paths. Not a prediction — a shape. It answers "what does my edge feel like at its worst" before your account finds out live.

The health checks — nine signals, always on

Sample size, activity ratio, profit-factor band, recovery factor, Sharpe-CAGR coherence, win/loss shape, sizing after losses, streak dependency, and the equity-versus-balance drawdown gap. The same checklist that exposes fake signal pages on the internet, pointed at your own account: pass, watch or flag, with the reasoning shown. Flags are questions, not verdicts — small samples never produce confident ones.