Glossary · Intelligence & research
Base rate
How often something happens in general, used to judge whether a specific finding is actually unusual.
What it means
A base rate is the underlying frequency of an event or condition in a relevant population. In analysis, it is the prior you should hold before examining a specific case. Neglecting it, by weighting a vivid specific detail more than the general frequency, is one of the most common reasoning errors. Good tradecraft states the base rate explicitly so the reader can judge how far a finding departs from it.
Why it matters in a dossier
Public records are full of things that look bad out of context. Most Massachusetts LLCs miss an annual report at some point. Most commercial properties on Route 18 have a code violation in their file. Most contractors have been sued. ICS reports a finding against its base rate: three suits in two years when the trade averages one in ten is a signal; one in ten is Tuesday. This keeps a dossier honest and keeps a client from walking away from a sound deal over noise.
In practice
A due-diligence brief on a Rockland restaurant tenant finds four Board of Health inspection failures in three years. Pulling the same records for the six nearest restaurants shows an average of 3.5 failures over the same period, all corrected on reinspection. The tenant is ordinary. The brief says so, with the comparison, and the landlord proceeds instead of losing a $42,000-a-year lease to a false alarm.
Straight answers
- Why does a researcher need base rates?
- To avoid treating normal as suspicious. A contractor with two small-claims cases in ten years sounds concerning until you learn that is typical for the trade. A town board member who owns property in town is the norm, not a conflict. Without the base rate, every finding looks like a red flag.
- Where do base rates come from in public-records work?
- From the records themselves. Pull the same search for a handful of comparable subjects, or count how often the pattern appears across a town's filings. Published statistics help when they exist. The point is to have a comparison in hand before deciding that one subject stands out.
Related terms
-
Confidence levels
Stating how sure the analyst is: high, moderate, or low, and why.
-
False positive
A record that appears to match the subject of a search but, on inspection, belongs to someone else.
-
Key assumptions check
A structured review of the beliefs a conclusion rests on, to see which would break it if wrong.
-
Structured analytic techniques
Repeatable methods that force analysis into the open so assumptions and gaps can be seen and challenged.
Where this term shows up
Terms that point here: Code violation records