A practical guide to manual evidence prioritization for a watchlist in TEMIRIN using configured records, explicit market identity, current controls, and human review.
August 3, 2026 · 4 min read
Evidence prioritization is a human review practice for deciding which stored items deserve attention first. Write the queue criteria in operational language: primary authority, time sensitivity, source ownership, collection freshness, explicit watchlist relevance, approaching contract deadline, unresolved source error, or a material change to a known record. The order should remain explainable from fields a reviewer can inspect rather than appearing as a forecast.
Use several lanes instead of one overloaded score. An official document published moments ago, a failed source that normally updates before an event, and a secondary analysis relevant to a later review may all be important for different reasons. Recording the lane and human reason prevents freshness from overwhelming authority and prevents a popular publisher from jumping ahead of the contract’s named resolution source.
Set a review expectation for each lane. A time-sensitive primary release may need immediate inspection, while a background methodology document can wait for scheduled research. The expectation is an operating convention, not a decision timer or market prediction. Its purpose is to expose neglected evidence and distribute attention without implying that queue position measures expected return.
Open the canonical URL or provider reference, publisher or account, observed time, supplied publication time, configured-source status, and duplicate history before promoting an item. A new primary record may deserve attention ahead of many summaries. Conversely, a fresh article that repeats an older announcement should remain on the same provenance chain rather than receiving artificial weight from its timestamp.
Source collection state changes the next action. A successful fetch with incomplete content calls for document review; a failed fetch calls for an operations check; a correction calls for comparison with the earlier version. Preserve each state and the reviewer’s response. Priority should help a teammate know what to inspect next, not compress evidence quality and source operations into an unexplained badge.
Place evidence beside a watchlist-derived review record only when it carries the supported explicit market identifier. Open the contract and verify the question, outcome, time window, and resolution authority. An urgent political, macro, sports, or crypto topic can resemble many contracts while satisfying none of their exact conditions. Unlinked material should remain available in the evidence collection for manual research.
Keep source reliability, evidence relevance, and market applicability as separate judgments. A highly authoritative source can be irrelevant to the selected outcome, while a narrow primary document can be crucial despite limited general coverage. The reviewer’s note should say which contract language makes the item useful. This creates a durable explanation for priority without claiming broad market association from shared words.
Compare the preserved trigger price with a separately time-stamped current price only after confirming market identity. A large move can increase urgency, but it may also mean the information is already reflected or the displayed market has become thin. Show public bid, ask, spread, and visible depth when present. None of those observations proves that the evidence is accurate or important.
Refresh the configured public-wallet snapshot before moving a record to the front of a decision queue. Visible exposure and portfolio concentration can change who should review it and how quickly. Keep address observation time, missing fields, and coverage limits visible. Source and category state, confidence setting, stake cap or override, free-capital percentage, price-move guard, and kill switch remain separate policy context.
A prioritized item becomes a paper order only after a user enters a direction, paper price, and notional. Store the per-order, rolling twenty-four-hour, per-market exposure, and portfolio-concentration results with that simulation. Keep the earlier priority reason, later in-app decision, alert attempts, journal note, and paper record distinct so their sequence remains visible during review.
Sample the top, middle, and bottom of the queue, plus unlinked, duplicated, corrected, failed, and ignored items. Trace why each occupied its position using facts available at the time and include records that received no action. If reviewers repeatedly misunderstand a lane or overlook an authority, revise the rubric prospectively. Do not rewrite earlier priority reasons after seeing the market move.
Useful facts include canonical authority, collection freshness, source ownership, explicit watchlist relevance, market timing, operational failures, and the reason a human recorded for the priority.
It remains available in the evidence collection for manual research and is not placed beside a watchlist opportunity until the relationship is explicitly verified.
No. Movement can affect review urgency, but provenance and relevance still require their own inspection. Both prices should remain time-stamped observations with available market context.
