<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Detection and Response on Cyber Mixology</title>
    <link>https://sedward5.com/tags/detection-and-response/</link>
    <description>Recent content in Detection and Response on Cyber Mixology</description>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Thu, 23 Jul 2026 12:00:00 +0000</lastBuildDate>
    <atom:link href="https://sedward5.com/tags/detection-and-response/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Ranking Detection Rules with the Wilson Score Interval</title>
      <link>https://sedward5.com/wilson-scoring-for-detection-and-response/</link>
      <pubDate>Thu, 23 Jul 2026 12:00:00 +0000</pubDate>
      <guid>https://sedward5.com/wilson-scoring-for-detection-and-response/</guid>
      <description>&lt;p&gt;Detection engineering teams often inherit a deceptively simple problem: how do you decide which detection rules are actually your best?&#xA;Most teams already collect two useful metrics for every rule. First is &lt;strong&gt;precision&lt;/strong&gt;—the percentage of alerts that analysts ultimately determine are legitimate security activity. Second is &lt;strong&gt;alert volume&lt;/strong&gt; over a given period.&lt;/p&gt;&#xA;&lt;p&gt;Unfortunately, these metrics often point in different directions.&lt;/p&gt;&#xA;&lt;p&gt;A rule with 100% precision across 10 alerts sounds fantastic, but should you trust that result? Maybe. Maybe not. Ten alerts simply isn&amp;rsquo;t much evidence.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
