Analytic standards
This site publishes independent analysis of information warfare, influence operations and cognitive security. This page explains how that analysis is produced, so you can judge how much weight to give it.
Sources
Every analysis is built on sources a reader can check. In order of preference:
- Primary documents: official doctrine, government and parliamentary reports, court records, platform transparency reports and original datasets.
- Peer-reviewed research and reports from established research centres, such as the NATO StratCom Centre of Excellence, the Oxford Internet Institute and the EU’s East StratCom Task Force.
- Quality journalism with named reporters and documented evidence.
- Open-source material (OSINT), such as public posts, archived pages and metadata, used only when it can be verified independently.
Anonymous claims, single-source stories and material from parties with an obvious stake in the outcome are treated with extra caution. When they are used, the text says so. Every analysis ends with a list of its sources.
AI tools are used to assist with research and drafting. Every analysis is reviewed, fact-checked against its sources and edited by the author, who is responsible for its content.
Facts and judgments
The analyses separate three things:
- Facts: what is documented and sourced.
- Assessments: what the author concludes from those facts. These are signalled with phrases such as “in my assessment” or “the evidence suggests”.
- Forward-looking judgments: what is likely to happen. These are uncertain by nature and are labelled that way.
Key judgments and confidence levels
Many analyses open with a short list of key judgments: the main conclusions, stated plainly. Each list carries a confidence level.
Confidence describes the quality of the evidence, not the probability of an event:
- High confidence: the judgment rests on high-quality, corroborated sources. It is still a judgment, not a certainty.
- Moderate confidence: the information is credibly sourced and plausible, but not corroborated well enough for a firmer judgment.
- Low confidence: the information is scant, fragmented or comes from sources of uncertain reliability. The judgment is offered because it is useful, not because it is settled.
This approach follows the practice of professional intelligence analysis, adapted for public writing.
Corrections and updates
Errors of fact are corrected openly. Significant corrections are noted at the end of the article with the date of the change. When an analysis is substantially revised, the “Updated” date at the top of the page changes.
If you spot an error or have evidence that challenges a judgment, please get in touch. Substantive disagreement is welcome.
Independence
This is a personal publication. The author works in the cyber defense field, but the analyses rely only on publicly available information. They contain no classified, confidential or employer-internal material. The views expressed are the author’s own and do not represent any employer or institution.
The site carries advertising. Advertisers have no influence over topics or conclusions, and there is no sponsored analysis.