tlp-unclear/estimative-language.md

**1st April** - Collaboration Between TLP:CLEAR and CIA to Introduce a `-1` Value in Estimative Language Probability

The [TLP:UNCLEAR](https://tlp-unclear.org) classification announces a partnership between TLP:UNCLEAR and the CIA, aimed at enhancing the "[Words of Estimative Probability](https://www.cia.gov/readingroom/docs/CIA-RDP93T01132R000100020036-3.pdf)" a concept introduced by Sherman Kent.

Kent's Words of Estimative Probability have been expanded to include a new term. This term specifically denotes 'unclear' as a category characterized by negative probability.

| **Term**               | **Probability** | **Note**               |
|------------------------|-----------------|------------------------|
| Certain                | 100%            | Give or take 0%        |
| Almost Certain         | 93%             | Give or take about 6%  |
| Probable               | 75%             | Give or take about 12% |
| Chances About Even     | 50%             | Give or take about 10% |
| Probably Not           | 30%             | Give or take about 10% |
| Almost Certainly Not   | 7%              | Give or take about 5%  |
| Impossible             | 0               | Give or take 0%        |
| Unclear | -1 | Report back when, um, I don't know. When it makes sense.|

TLP-UNCLEAR iniative also review the existing [Probability Yardstick from UK Defense Intelligence](https://www.gov.uk/government/news/defence-intelligence-communicating-probability). The Probability scale divides the probability spectrum into seven distinct segments, each associated with specific terms. The division and term assignments were developed based on academic research and align with the general public's interpretation of these terms. The scale deliberately avoids a continuous structure to prevent giving an illusion of precise accuracy. 

In this model, the elements that are unclear are placed outside the standard probability scale. A value of `-1` is used to accurately represent such uncertainty, applicable in cases where an analysis does not fit into the conventional scale.

The selection of the value `-1` draws inspiration from the common practice in many Unix tools and libraries, where a negative return value signifies an error or an ambiguous state. This value effectively represents ambiguity, unclearness, and non-deterministic behavior, aligning with these existing conventions in computer systems where negative values often indicate atypical or exceptional conditions.