Calculator · Validation
Elusion & RecallWhat got left behind?
Estimate elusion, remaining responsive material, and recall from validation sample results.
97.73%
Recall estimate
Use after ranking or review to estimate what may remain in the unreviewed or discarded population.
Connect the eDiscovery Decoder MCP server and your agent can answer this with calculate_tar_recall_estimate and show its inputs and interval. The hosted tool reports Wilson intervals; this page uses exact Clopper-Pearson intervals, so the ranges can differ slightly.
Workspace
Elusion & Recall workspace
- Responsive found
- Missed: lower bound
- Missed: point estimate
- Possible up to the upper bound
- Other discarded documents
- Sampled, responsive
- Sampled, not responsive
- Elusion rate
- 0.27%
- 95% CI
- 0.07% - 0.68%
- Estimated remaining responsive
- 427
- Remaining responsive range
- 116 - 1,090
- Recall estimate
- 97.73%
- Recall range
- 94.41% - 99.37%
FormulaElusion rate = responsive in elusion sample / elusion sample size; recall = responsive found / (responsive found + estimated remaining responsive).
Glossary
Terms this calculator uses.
- Elusion
- A sample-based estimate of what responsive material may remain in the unreviewed set.
- Elusion Rate
- The proportion of documents in the discarded set that are actually responsive, estimated from a sample.
- Recall
- The share of truly responsive documents that were found by the workflow.
- Null Set
- The set of documents predicted non-responsive and set aside, also called the discard set.
- Validation Sample
- A random sample reviewed to estimate the quality of a completed review, such as its recall.
- Confidence Interval
- A range around a sample-based estimate that is likely to contain the true value, at a stated confidence level.
Use note
Educational and not legal advice.
Results are planning aids and can be wrong if inputs or assumptions are wrong. Users remain responsible for independent human and professional review.
Keep going
Other calculators.
Culling Impact
Did culling remove noise without losing responsive documents?
Richness & Yield
How rich is this collection, and what yield should reviewers expect?
Prevalence Results
How much responsive material is in this population, and how big is the review?
Precision, Recall & F1
How accurate are the model or reviewer calls?
Control Recall
Does a known control set confirm our recall?