Green, yellow, red — and why the color never acts alone.

What the color is

An expectancy threshold verdict on the blended week.

The week status is mechanical: the Overall Blended EV — the four branch EVs combined under the live profile weights — lands against fixed expectancy thresholds, and the week tags GREEN, YELLOW, or RED. Nothing discretionary enters the tag. That mechanicalness is the point: a verdict computed the same way every week is comparable across weeks, and the dashboard's distribution reads — green percentage, status counts, current streak — only mean something because no week's color was negotiated. The tag answers exactly one question: was this week's blended expectancy inside the tradable zone? It deliberately does not answer why, or what to do, which is where the rest of the doctrine takes over.

The second axis

Stability turns three colors into six operating states.

The dashboard's action logic pairs every color with a volatility read, and the pairing changes the prescription completely. GREEN and stable is the best operating state — stay systematic, and pointedly, do not over-adjust. GREEN with high volatility earns a warning instead of a celebration: good EV in a noisy distribution, don't confuse variance with skill, keep risk inside gate and pool rules. YELLOW stable is a thin but reliable edge — tighten selectivity, review mechanics. YELLOW volatile means thin and unstable — avoid aggression, hunt for sample distortion, missed quotas, news effects, branch drift. RED stable is the uncomfortable one: cleaner evidence of genuine weakness, reduce or pause per doctrine and diagnose at the branch level. RED volatile is the dangerous week — respect brakes first, and explicitly, do not chase random fixes.

The look-through

A blended color can hide a branch story.

The weekly diagnostic questions the manual prescribes all share one move: look through the blend. Was the week GREEN because multiple branches were healthy, or because one strong branch masked deterioration elsewhere? Was RED caused by actual expectancy decay, or by too few trades and high volatility doing what small noisy samples do? Did Trend No-Partial actually participate, or did avoidance quietly remove the fat-tail exposure the profile assumes? The blend is honest arithmetic over its inputs, which means it faithfully transmits their distortions — a profile-weighted average has no mechanism for noticing that its healthiest component is carrying a corpse. The color starts the weekly read. The branch columns finish it.

• A GREEN week with one deteriorating branch is a monitoring obligation wearing a good color.
• A RED week with high adherence, controlled MAE, and a thin sample is often variance — check gate, benchmark, and CP3 before touching rules.
• The Weekly Summary warning applies verbatim: do not let one strong branch hide a weak system.

What the color cannot do

The status informs posture; it never touches authority.

The scorecard sits in the evidence family, and its verdict inherits that family's limits in both directions. A RED week cannot force deployment below what the gate requires — the reduce-or-pause response follows doctrine, not despair — and a GREEN week grants precisely nothing: not a tier, not a pool expansion, not an override's justification. The strong operating rule from the console applies one level up: tactical and weekly evidence can never override a restricted gate. The color's actual consumers are posture and attention — next week's selectivity, this month's diagnostic focus, the review record's persistence trail. The capital consequences of a week arrive through the throttle's structural inputs on their own schedule, weighed against everything else the authority chain reads.

The key idea

The verdict is designed to be humble, and that's why it can be trusted.

Everything around the color exists to keep the compression honest: mechanical thresholds so weeks stay comparable, the stability axis so noise doesn't impersonate signal, the branch look-through so blends can't hide stories, and the authority boundary so a good week can't buy risk. What remains is a verdict simple enough to read in one glance and disciplined enough to mean the same thing every time — which, across a year of Fridays, is exactly what a weekly health monitor is for.