LessBother.

HUMAN WEB · CONVERSATION AUTHENTICITY LAB

Does this conversation look independent?

LessBother looks for coordinated behavior across accounts: repeated language, synchronized repetition and mutually reinforcing reply clusters. It does not label an individual Reddit user a bot from one strange comment.

PRIVATE BY DEFAULT

Analyze one thread — or compare several

Pasted thread datasets never leave this page. If the thread-data box is empty, you can enter up to three public Reddit thread URLs. LessBother fetches bounded public JSON for each and combines only normalized comment evidence locally, which makes recurring cross-thread account clusters much more informative.

For generic data, parent should contain the account being replied to. Native Reddit JSON is converted automatically. URL collection does not store the submitted Reddit URL or fetched thread payload.

No fake precision

The score is a coordination-risk index, not “73% chance this account is a bot.” Individual identity requires much stronger evidence.

Thread timing is normal

A Reddit thread naturally gets many comments quickly. LessBother does not flag that alone; it looks for synchronized repeated language across distinct accounts.

Longitudinal is stronger

This lab can now compare up to three Reddit threads in one run. Recurring account pairs and talking points across independent threads carry more weight than one dense discussion.