Allow and Disallow tell compliant crawlers which paths they may request. The meaning is simple, but overlapping paths can create unexpected results.
Disallow examples
User-agent: * Disallow: /account/ Disallow: /search/
These rules request that crawlers avoid URLs beginning with the listed paths.
When Allow is useful
User-agent: * Disallow: /files/ Allow: /files/public/
This blocks the broader folder while making an exception for a public subfolder. Search engines usually compare the most specific matching rule.
Review the intention
Do not block a page only because it has low traffic. Robots.txt is mainly a crawl-management mechanism. For duplicate or low-value pages, canonical tags, noindex and improved internal linking may be more appropriate depending on the goal.