--- url: "https://predictiontoolkit.com/methodology/" publisher: "PredictionToolkit" --- [← Polymarket tools]() # How we review Polymarket tools By PredictionToolkit · Updated 2026-10-05 PredictionToolkit organizes distinct offerings across trading, interfaces, analytics, research, development, infrastructure, capital, external data and learning. One offering keeps one stable profile even when several categories apply. ## Selection scope A research lead or website review does not automatically become a published or tested product. We reconcile identity and official destinations, describe a useful task and limitations, and assess category membership separately from suggestions. ## Distinct offerings Named API channels and components remain within a parent offering unless evidence establishes a separate usable product. Unknown pricing stays unknown; it need not exclude an otherwise useful listing. ## Evidence and testing Provider sites, documentation and repositories support our descriptions. Documentation review is separate from hands-on testing, security audit or endorsement. Any reported test names its action and date; detailed source and review records remain in editorial evidence. ## Dates and changes Fact-check dates differ from article edits and builds. Service fees differ from venue fees; forecasts differ from resolution observations; announced shutdowns differ from observed closures. Unsupported old prices leave current fields but retain dated history, as the Polyfollow correction and Struct cutoff illustrate. ## Comparisons and order Alternatives need a stated task difference; complements need distinct roles. Collections give a criterion and limitation per inclusion. Catalog and category sorting defaults to Highest rated, then unrated; ties use name and stable ID. Alphabetical sorting is available. Homepage editorial picks are separate. Confirmed commercial relationships are disclosed; unknown does not mean absent. ## Editorial rating scale - 1 — Unavailable for the intended task. - 2 — Limited evidence, major gaps or a dependency needing replacement. - 3 — Useful for a defined task with meaningful constraints. - 4 — Strong documented fit and clear conditions. - 5 — Exceptional assessment support, including hands-on validation. Half points fall between these levels. Editorial scores assess task usefulness, documented scope, setup clarity and access/cost transparency; they are not benchmarks, audits or return forecasts. The main score is labelled Editorial until 5 published user reviews. At 5, their arithmetic mean becomes primary, to one decimal place, excluding the editorial score. Below 5 after moderation, Editorial becomes primary again. ## User reviews One published review is allowed per confirmed email and tool; changes replace it. Email confirmation proves address control, not product use. Reviewers describe their own experience, disclose business connections and consent to publishing their chosen name and review. Emails remain private. Confirmation and moderation precede scoring. Positive and negative reviews follow equal rules; spam, duplicates, impersonation and irrelevant content are excluded. External discussions and testimonials do not count as submitted reviews. ## Corrections and media Send the page, affected claim and dated primary source to info@polytradingbot.com. Ownership claims require evidence and do not grant listing control. Website submission forms are unavailable. Corrections update the versioned record; images need provenance, dates, accurate captions and reuse grounds. A saved screenshot alone is insufficient. Request a review change or withdrawal from its submitting email. ## Continue researching - [Polymarket directory updates]() · context: See dated fact corrections. - [Struct Polymarket API shutdown]() · context: Read the shutdown announcement context. - [Editorial policy]() · context: Read editorial policy. - [Privacy policy]() · context: Read data-handling terms.