Your Big Code Break Is Mostly Luck, 10-Year GitHub Study Finds

Researchers Yasir Zaki and Hazem Ibrahim wanted to know something simple: in open-source software, does success come from talent, timing, or plain luck? They pulled the entire public record of GitHub uploads from 2015 to 2025 — 102.2 million pieces of work by 6.15 million contributors — and measured success by how many other software projects came to depend on each one. Think of it as counting how many other businesses decided to build on your foundation.

The answer looks a lot like science, music, and books. When your biggest hit arrives is close to a coin flip across everything you make. There's a slight advantage to being early in your career, but it's small. What does carry over is a personal quality: some people reliably produce more influential work than others. Yet that lasting factor explains only about a fifth of why success keeps repeating. The rest behaves like momentum — a win gets you noticed, and being noticed gets you the next win. Reputation, not just raw ability, does a lot of the work.

Here's the part that matters most for anyone worried about AI. ChatGPT launched in late 2022, and researchers checked whether the rules of the game shifted afterward. They didn't. The measured change was plus 0.03, with a margin of error so wide it's indistinguishable from zero. The stable personal factor grew about as much as it did for an older group of coders who simply aged normally.

So what should you take from this? Be skeptical of reading GitHub track records as pure measures of merit — luck and momentum are baked in. And for all the talk of AI rewriting careers, this data suggests the ladder is, so far, shaped the same way it always was.