How we find similarities
Our first experiment compares six energy disruptions. The criteria are public, and the result can be reproduced.
Start with sources
Record dates, actors and indicators with references readers can inspect.
Compare the context
Assess the economy, institutions, technology and geography before claiming an analogy.
Follow the consequences
Record what happened over subsequent months and years, including results that challenge the first hypothesis.
Show the mistakes
Once forecasts are published, preserve their version, date, horizon and outcome.
How to read the result
Choose five features of an energy crisis: fuel, type of disruption, cause, location and intensity. The site compares them with the records in the archive. The more features they share, the higher the score.
A score of 80 out of 100 means the two cases resemble each other by these criteria. It does not mean the same outcome has an 80% chance of happening. We also show differences and sources. This early comparison covers only six energy crises and has not been validated for forecasting.
Calculation details
The initial weights are: fuel 20, type of disruption 28, cause 22, reach 12, intensity 18. Partial matches receive some points. Differences between eras can reduce a score by up to 8 points; if too much information is missing the site shows no score. These are editorial choices that still need retrospective validation.