Operator brief · 386

Any single metric can look healthy while the machine deteriorates.

The key idea

The testable claim

Every metric has a blind spot with a name.

'One number always lies eventually' is a slogan until the lies are enumerated. They can be. For each of the five metrics there is a specific, ordinary deterioration — not an exotic edge case — that the metric will not register while it is happening. Setting them out one at a time converts the argument for the panel from an appeal to thoroughness into a list of concrete holes, each of which is plugged by a specific other member. Enumerating the blind spots also supplies a defence against the panel growing indefinitely: a proposed sixth metric has to name the specific deterioration it detects that none of the five can, and most candidates cannot.

Expectancy alone

Edge intact, cost of harvest doubling.

A system can hold its expectancy per trade precisely while the drawdown required to produce it deepens substantially — typically when variance rises without the average changing. Expectancy reads flat and correct throughout. The account experiences a materially worse system: deeper holes, longer recoveries, more gate compression, and a trader far more likely to abandon it. Nothing in the expectancy figure contains this information, because expectancy is an average and the deterioration happened entirely in the distribution around it.

Drawdown alone

Shallow drawdown bought by deploying almost nothing.

The mirror case. Drawdown readings improve steadily while the account quietly stops taking risk — fewer trades, smaller sizes, more sitting out — and the metric registers the improvement as health. It is the signature of a trader who has been frightened by a previous drawdown, and it produces a curve that is admirably flat and going nowhere. Efficiency is what catches it: risk deployed and return generated, read together, reveal an account that has purchased its safety by declining to operate.

FigureTwo deteriorations, neither visible in its own metric

Schematic. Expectancy holds while the cost of harvesting it rises. Drawdown improves while deployment collapses.

RAPF alone

Excellent profit quality inside a narrowing condition.

Risk-adjusted profit factor can strengthen while the range of conditions producing that profit contracts sharply — a system increasingly dependent on one regime, one session, or one branch. The metric registers improving quality and it is not wrong; the profit genuinely is high quality. What it cannot see is concentration, and concentration is a fragility rather than a defect, so it produces no bad reading until the supporting condition disappears. Branch attribution and regime classification are what surface it while the numbers still look excellent.

RAER and acceleration alone

Efficient conversion of a shrinking edge, and momentum without direction.

Efficiency can hold or improve while the underlying edge decays, because the ratio describes conversion rather than magnitude — converting a smaller edge well reads identically to converting a larger one adequately. Acceleration has the opposite blind spot: it describes the second derivative and is silent about level, so a system accelerating from a poor base looks like a system with momentum. Neither is a bad metric. Both require a companion that reports magnitude, which is what expectancy supplies.

What follows

The panel is a minimum, not a thorough option.

The practical conclusion is that five is the floor rather than a generous allowance. Each metric has a demonstrated hole; each hole is closed by a specific other member; removing any one reopens a hole with no other coverage. Traders who trim the panel almost always trim to expectancy and drawdown — the two most intuitive — and thereby lose the ability to detect concentration, poor conversion, and decay inside growth. That is three blind spots traded for a shorter review, and it is a bad trade. The trimming instinct is worth understanding rather than merely resisting — it comes from the review taking too long, which is a real problem with a different solution: automate the assembly, not the panel.

Connected inside MARS

The Five Metrics →

Risk-Adjusted Efficiency Ratio →

Risk-Adjusted Profit Factor →

Drawdown →

Every brief documents the same shipped system.

The complete MARS package — eleven workbooks, three TradingView indicators, the full manual library — $497.

Get MARS Pro — $497 Explore the Platform