Transparency

Transparency matters because crypto is already full of people pretending that their incentives are cleaner than they are.

Editorial independence and conflicts of interest

Proof of Analysis does not accept payment, free tokens, preferential allocations, affiliate compensation or any other benefit from projects in exchange for coverage. Projects cannot purchase a Deep Dive, influence a POA-100 score, review an unpublished verdict or negotiate the wording of an analysis.

The people behind Proof of Analysis may hold assets discussed in the publication. Familiarity with an asset can improve research while ownership can affect judgment, which makes disclosure essential. From 1 June 2026 onward, every publication containing an investable verdict will state whether a relevant position was held at the time of publication and whether any other material conflict existed.

Positions may change after publication as prices, fundamentals and portfolio constraints change. A published analysis records the judgment and disclosed interests applicable at its stated date. It does not provide a real-time account of subsequent portfolio activity.

Corrections, revisions and score changes

Research is a dated record. Market data, circulating supply, unlock schedules, protocol revenue and governance conditions can change shortly after publication. Each analysis should therefore be read together with its stated research cutoff.

Material factual errors are corrected when identified. A correction that changes the investment thesis, valuation or POA-100 score will be disclosed visibly in the article. Minor edits involving spelling, formatting or clarity may be made without a separate notice.

Scores are not rewritten merely because price performance later makes an earlier judgment appear inconvenient. A score changes when the underlying evidence changes or when a methodological error requires correction. Conflicting figures from projects, data providers or on-chain sources are disclosed where the discrepancy could affect the verdict.

Readers may report a suspected error through the contact address provided in the Legal Notice.

A public scoring framework

Every complete token assessment follows the same weighted POA-100 framework. The methodology separates fundamental quality, valuation, recommendation, risk and research confidence, allowing readers to see why a strong project can remain an unattractive purchase at the wrong price.

Proof of Analysis uses artificial intelligence (AI) in its research workflow. We use it to scan large volumes of material, compare claims across sources, summarize technical documents, stress-test frameworks, organize token data, surface contradictions, and prepare drafts for human review. Some phrasing, structure, or analytical sequencing may occasionally carry traces of that process. We prefer to say this plainly.

This publication is built by humans using machines. That distinction matters. AI assists the work, but it does not own the argument. It does not determine our verdicts, select our convictions, or grant a token the dignity of being called investable. Those calls remain ours, with all the occasional embarrassment that human judgment involves.

AI allows us to examine more material than a small editorial operation could process unaided. It lets us move faster through documentation, unlock schedules, ecosystem claims, on-chain data, governance archives, product descriptions, and the endless promotional fog of the crypto industry. That speed has value only when paired with skepticism. Raw acceleration without judgment is how mediocre research becomes confidently mediocre. The machine can gather, compress, compare, and provoke. The final assessment belongs to us.

We also recognize the risk. AI can produce fluent errors, elegant simplifications, and misplaced certainty. It can miss irony, context, incentives, and the small signs that often matter most in crypto. Every serious user of AI knows this. Every unserious user pretends otherwise.

Proof of Analysis is therefore neither a hand-carved monastery of pre-digital purity nor an automated content farm with a research costume. It is a human research publication using modern tools to think harder, cover more ground, and make better calls with fewer illusions.

Time in the market beats timing the market. We have been here since 2017.

That means almost ten years of crypto, three bull markets, three bear markets, exchange collapses, protocol failures, regulatory panic, liquidity manias, liquidation cascades, false revolutions, real breakthroughs, cults with tickers, tickers with cults, and enough market noise to make silence feel like alpha.

Surviving that long does not make us infallible, but it makes us harder to impress.

AI can read more than we can. It can process faster than we can. It can summarize, compare, classify, and detect patterns at a scale no human desk should pretend to match. Yet there are things the market teaches only by taking your confidence, your sleep, and occasionally your money. Anyone who has lived through several cycles has learned a form of judgment no model can fully extract from documents.

We use AI. We respect it. We also know what it lacks: scar tissue, embarrassment, memory under pressure, and the rare ability to recognize a familiar disaster wearing a cleaner interface.