Five metrics chosen for what they prevent each other from hiding.

Not a dashboard

A dashboard adds metrics. This panel selects against them.

The instinct behind most trading dashboards is coverage: more numbers, more visibility. The five-metric panel was assembled by a different rule — a metric earns a place only by catching a specific failure the existing ones cannot see. That is a subtractive discipline rather than an additive one, and it is why the panel is five rather than fifteen. Numbers that merely restated something already visible were left out no matter how respectable they are elsewhere. The subtractive rule has a side effect worth stating: because nothing was included for completeness, there is no metric on the panel that can be safely ignored, which is a stronger claim than any dashboard can make about its contents.

The catches

Each number exists to stop another from lying.

Stated as pairs, the design becomes legible — every metric is somebody's auditor:

• Drawdown catches expectancy rewarding recklessness — edge harvested at ruinous cost.
• RAER catches drawdown rewarding timidity — safety bought by deploying almost nothing.
• RAPF catches profit that is luck rather than quality.
• Acceleration catches all of them missing slow decay in a still-rising curve.
• Expectancy catches the others being satisfied by a system with no edge at all.

The failure each pair prevents

Any single metric optimised alone produces a recognisable pathology.

The design is easiest to see in what happens without it. A trader optimising expectancy alone drifts toward high-variance methods that eventually deliver a drawdown they cannot recover from. One optimising drawdown alone drifts toward inactivity, and a system that never risks anything has excellent capital preservation and no reason to exist. One optimising profit factor alone concentrates into a narrow condition that works until it does not. Each pathology is the predictable consequence of a real metric taken seriously in isolation, which is the argument for the panel in its clearest form.

Why not more

A sixth metric would have to catch something the five miss.

The panel is closed rather than fixed, and the entry requirement is specific: a candidate must detect a failure mode none of the existing five can see. Most proposed additions fail this test on inspection — win rate is a component of expectancy, average R is a component of expectancy, Sharpe-style ratios largely restate RAER in a different scale. None of them is wrong; all of them are redundant against a panel already carrying that information. The bar is not novelty of the number but novelty of the blind spot it closes.

Reading them together

The panel is read as a shape, not as five scores.

In practice the five are read as a configuration rather than sequentially. Strong expectancy with deteriorating acceleration is one recognisable shape — the edge is intact and something is consuming it. Acceptable drawdown with weak efficiency is another — the account is safe and the risk it does take is being converted badly. Experienced operators stop reading the numbers individually and start recognising the configurations, which is the point at which the panel becomes fast to use rather than laborious.

When they disagree

Disagreement is the panel working, not the panel failing.

Traders new to the hierarchy often want the five to agree and treat conflict as a sign something is miscalculated. The opposite is true: agreement is the low-information state, and the disagreements are where every diagnosis comes from. Expectancy rising while efficiency falls, drawdown improving while acceleration flattens — each conflict points at a specific part of the machine. A panel whose members always agreed would be four redundant numbers and one useful one. In practice the most valuable weeks are the ones where two metrics move in opposite directions, because that is when the panel is telling the operator something they could not have worked out from the equity curve.