Full disclosure: I left Twitter when it changed to X, so I’ve never used the platform.
X recently open-sourced its algorithm. This post dissects the code to find out how it works. The algorithm tries to maximize positive engagement and minimize negative engagement while prioritizing quality over quantity.
Controversial content persists despite penalties because volume overwhelms penalty. Rage bait works when the engaged audience vastly outnumbers the offended minority. Early engagement matters because of feedback loops, not because of any multiplier for fast likes.
Read the full article on Code Pointer.
Sources:
• TLDR. (2026, February 2). TLDR Newsletter. https://a.tldrnewsletter.com/web-version?ep=1&lc=f5cb7e17-003d-11ed-9258-0241b9615763&p=f506196a-0010-11f1-8219-038103267f56&pt=campaign&t=1770033693&s=b1ae5ccaa098e95025d02e286faf4b0178ff0f03d200dff5e776f132678924d7
• Yongkyun. (2026, January 28). [x-algorithm] How X Decides What 550 Million Users See. Code Pointer. https://codepointer.substack.com/p/x-algorithm-how-x-decides-what-550
