I’m currently watching Money Heist on Netflix. Since I tend to watch stuff whilst doing something else and my hearing isn’t great I have both English audio and English subtitles. It’s quite distracting in that even normal dialogue is phrased differently most of the time. Same meaning, just a completely different sentence. Some lines though are completely different in meaning: for example audio “For a moment I thought to myself, I guess he’s got a point” becomes subtitle “I thought about the Chinese and I hate people who spit.”
This isn’t the only show I’ve seen where the dubbed audio is radically different to the subtitles. Why wouldn’t they dub it then transcribe the dubbing as subtitles or vice-versa, keeping both consistent?
Subtitles are (or should) written with the purpose of making them easy to read when flashed quickly on screen. Dubs are (or should be) spoken with the purpose of matching them to the lip flaps on screen, while trying not to sound too artificial. The voice actor may even change the dialogue on the fly.
I’ll also bet that they use a form of voice recognition system for the subtitles, and these can have a tendency to interpret what is actually being said in various “interesting” ways. For movies and some other programs I like to use subtitles but there are lots of moments when the VRS gets the dialogue completely wrong or blanks out for a bit.
For older movies, especially old B movies from the mid 60s to the early 90s, stuff like “chopsocky” flicks from Hong Kong, spaghetti westerns and horror/thrillers from Italy, and films from the Philippine, South America and Japan, there was a widely practiced method of a company translating the foreign film’s original script into English, rewriting most of the original dialogue to fit the intended audience(US-UK), ignoring almost all the cultural details/local jokes found in the original script, and, in most cases, tone down the language, inappropriate remarks and jokes, and then give the newly revised script to B level voice actors, who sounded nothing like the original actors, who, then, deliver some of the cringiest dialog ever.
Bare in mind, most films made outside the UK, US, and France during that period of time didn’t record the dialog live on set, so even the original language of these movies is technically considered a dub.
The simplest answer is that there are two separate groups of people doing the dubbing and the subbing. The dubbing actors/actresses are chosen for their vocal abilities and they many not even be fluent in the foreign (to them) language, just speaking phonetically. By the same token, those doing the subtitles are chosen for their ability to transcribe their, usually native language into, in this case English. Again, just because they’re professionals doesn’t mean they have a full grasp of the complexity of the English language.
Why don’t they just use the same dialog for both? Probably time and cost restrictions. Both take multiples of the running time of the movie/video and may be done concurrently in completely different studios/offices.
Also, most languages have different spoken and written styles that can’t be completely, intelligibly be interchanged, especially since as Terminus Est stated, subtitles are best as brief and concise as possible.
As for the example cited in the OP, I chalk it up a bad translation for the dubbing or the subs. And it may well be possible that both are wrong.
If you want something that really messes with your head, watch at movie with Russian Gavrilov translation. Not only is the original dialog left in the background, there’s usually a single male and female voice actor (or sometimes just one voice for everyone!) with no regard for emotion, rhythm or pacing of the original!