A 400 status shows up when the server can't process a request because something about it is malformed, before the server even gets to the question of whether the requested content exists. A broken URL with invalid characters, a corrupted query string, or a request the server simply can't parse are typical causes.
It's rarer to encounter than a 404 in normal browsing, since it usually points to something wrong with how the request itself was built rather than a missing page: a bug in how a link was generated, a script constructing malformed URLs, or a browser extension or proxy interfering with the request. When 400 errors show up in crawl reports or server logs at any real volume, it's worth tracing where the malformed requests are actually coming from, since fixing a page that never gets a legitimate request in the first place does nothing for the underlying cause. Checking the request itself, rather than the destination, is usually where the actual fix lives.
