We were about to ship a performance write-up claiming three queries had gotten 2-4x slower. All three numbers were wrong, and the reason is one every benchmark author should worry about: the same binary , on the same machine , running the same query , gave different answers depending on what tim...
Source: [Dev.to](https://dev.to/samyama-ai/we-nearly-published-a-2x-regression-that-didnt-exist-3bii)