CyberKit

Precision and Recall

Also called Fidelity.

Metrics for how accurate and how complete a detection is.

What it is

Precision is the share of a detection's alerts that are true positives, while recall is the share of actual threats that the detection catches. Together they describe how clean and how complete a detection is, and improving one often costs the other.

Why it matters

These metrics turn vague notions of detection quality into something measurable, letting teams reason explicitly about whether a rule is too noisy, too leaky, or appropriately tuned.

In practice

A rule with high recall but low precision that catches every intrusion but buries each one under a hundred false alarms.

What to look for

  • Precision: how many alerts were real.
  • Recall: how many real events the rule caught.
  • Tuning that raises one while quietly dropping the other.

Common mistake

Optimizing only for fewer alerts. That raises precision and can quietly cut recall.

Practice it nearby

CyberKit has no tool for this term itself. These are the closest places to practice it.

In the field

Well-known tools where this shows up on the job. They are listed to show the landscape, not as a recommendation.

  • Sigma Open source

    A vendor neutral rule format that converts into queries for most SIEMs. Our notes.

  • YARA Open source

    Matches byte and string patterns in files or memory, the standard for malware rules. Our notes.

  • Suricata Open source

    Runs network rules against live traffic or captures and logs every match. Our notes.

  • Splunk Enterprise Security Commercial

    Where many teams run, tune and schedule detection searches in production.

Read next

  • False Positive: An alert that flags benign activity as malicious.
  • False Negative: Malicious activity that no detection caught.
  • Tuning: Adjusting detection logic to cut noise without losing real catches.

Sources