How the record works
TakeTemps sourcing and scoring methodology
Every prediction starts with a source, stays OPEN until its outcome is known, and is graded against a stated resolution condition.
There are no points
TakeTemps does not award points. An analyst's record is the number of HIT and MISS receipts among settled predictions.
Accuracy = HIT receipts divided by settled receipts. OPEN receipts are excluded. A display of 67% (6/9) means six HIT receipts from nine settled receipts.
Receipt statuses
- HIT: the sourced prediction resolved correctly.
- MISS: the sourced prediction resolved incorrectly.
- OPEN: the result is not settled and does not affect accuracy.
Leaderboard qualification
Five settled receipts are required to qualify. Qualified analysts rank ahead of provisional analysts, then by accuracy and settled sample size.
Source and grading standards
- The prediction must be specific enough to resolve true or false.
- The source must identify the analyst and preserve the published wording or an explicitly labeled normalized poll response.
- The grading record includes the actual outcome and keeps a link to the original source.
- Duplicate versions of the same underlying prediction are not counted twice.
- Corrections are documented when attribution, wording, or grading changes.
Research automation
Automated tools can help locate candidate predictions and identify possible duplicates. A source-verification pass checks attribution, dates, wording, and resolution before publication.