Skip to main content

Definitions before conclusions

Research notes that keep the denominator attached.

Polytrack publishes code-reviewed notes on how to interpret monitored trades, severity, wallet activity, aggregates, rankings, and alert delivery. Each note names its sources, states what the evidence can support, and preserves the limitations that should travel with any quotation.

Published by PolytrackReviewed July 14, 2026RSS feed

Current library

Four reusable interpretation notes.

Method note6 min read

Severity is a screening tool, not a verdict.

A useful alert system reduces a large stream to a smaller review queue. It does not turn unusual activity into proof. This note explains the reasoning boundary behind Polytrack severity labels and the checks a reader should make after a record is flagged.

Read the note
Data interpretation6 min read

A wallet address is a research key—not a verified identity.

Wallet-level research is valuable because one address can connect trades, positions, and repeated patterns. The same convenience creates an attribution trap. This note separates address-level evidence from claims about people, teams, services, or ownership.

Read the note
Data interpretation5 min read

Monitored volume is not exchange-wide volume.

Aggregates become misleading when their denominator disappears. This note defines the difference between volume in Polytrack's monitored records and total Polymarket activity, then gives a checklist for reading rankings and trends consistently.

Read the note
System transparency5 min read

Telegram alerts have a delivery chain, not an instant promise.

“Real-time alert” language can hide several independent systems. Polytrack instead documents the path from upstream availability to ingestion, classification, eligibility, Telegram delivery, and the subscriber's device so users know what a message can and cannot promise.

Read the note

The publication contract.

A public note should be easy to quote without becoming less true after the quote is shortened.

  • A direct answer and a defined unit of analysis.
  • Visible author, publication date, and modified date.
  • Primary or canonical sources with a reason for each citation.
  • A separate limitations section, not a hidden disclaimer.
  • No claim that monitored coverage equals exchange-wide data.
  • No identity, intent, outcome, or profit conclusion from a classifier label.

Start with the full interpretation contract.

Methodology explains how records and metrics are built. Data Sources shows provenance. Limitations defines the claims those facts cannot support.

Read methodology