Why use a ranking at all?
An equity universe contains different prices, sectors, volatility profiles and chart structures. A ranking layer creates a common comparison surface by translating several defined evidence families into one ordered output.
The score answers “which setup aligns more strongly with these rules?” — not “what will happen next?”
The evidence families stay conceptually separate
The public architecture combines four broad inputs while keeping their roles distinct:
- weekly coherence, which defines the admissible directional side;
- trigger confluence, which records daily structural evidence;
- payoff geometry, which compares objective and invalidation distances;
- structural space, which asks whether a usable objective exists.
Two setups can reach a similar rank through different evidence mixes. That is why the terminal keeps the underlying context visible instead of presenting the score as a self-explanatory answer.
Context comes after the base setup
The technical setup is assembled before market and sector context are applied. Context can increase, preserve or reduce the relative weight of the completed base ranking. It does not create a setup that was absent from the underlying evidence.
The final label needs more than a score
A high score still has to pass structural gates. The final classifier reads directional coherence, objective availability, rank, risk/reward and veto conditions together. A strong number cannot repair a missing objective or an incoherent side.
What the score cannot tell you
The rank depends on its inputs, data, rules and configuration. Change those components and the ordering can change. Historical testing can assess how a configuration behaved on selected data; it cannot certify future outcomes.
The internal point allocation remains proprietary. The public goal is to expose the architecture, definitions and limits necessary to interpret the visible output correctly.