The thing about bad subtitle translation, it doesn't announce itself. It reads naturally. It feels fine. The error only becomes visible when someone who speaks the original language watches alongside you and goes quiet for a second before saying "that's not what she said at all."
Here are five real, documented Netflix subtitle errors, and what a professional translator would have done with each one.
1. Squid Game, The Error That Erased a Character's Entire Purpose
The Show: Squid Game (Korean → English) The scene: Mi-nyeo tries to convince other players to join her team.
| What the Subtitle Said | What She Actually Said |
| "I'm not a genius, but I still got it worked out." | "I am very smart. I just never got a chance to study." |
Why it Matters: These are not two ways of saying the same thing. The subtitle makes Mi-nyeo sound resourceful but unremarkable. The original makes her a specific type of character that Korean cinema knows well, someone who is highly intelligent but was denied education by poverty. That detail is the entire point of her being in the game. The wealth disparity theme of the whole show is compressed into that one line. The subtitle loses all of it.
What Went Wrong: The translator rendered the surface meaning without understanding the cultural reference. Korean audiences recognize "I never got a chance to study" immediately as a class and opportunity statement. English audiences got a shrug.
- How Polilingua Would Have Translated It: "I'm smart. I just never had the chance to prove it."
It preserves the implication, intelligence without opportunity, without requiring the viewer to know Korean social context. The character's motivation stays intact.
2. Money Heist, When "Hostia" Became Invisible
The show: La Casa de Papel / Money Heist (Spanish → English) The scene: A character threatens another in Season 2.
| Original Spanish | English Subtitle | Dubbed English |
| "...te voy a dar una somanta de hostias que no te vas ni a mantener en pie." | "I'll beat you senseless." | "I'm gonna beat you so hard you don't know what day it is." |
Why it Matters: "Hostia" is a Spanish expletive derived from the Eucharist, the communion host. It carries a specific register of Spanish profanity that mixes irreverence and violence in a way that is culturally loaded, particularly in Andalusian and southern Spanish speech. "Beat you senseless" is accurate in meaning but completely strips the line of its texture. It sounds like a threat. The original sounds like this specific character making a threat.
What Went Wrong: This is the most common error in Spanish-to-English subtitle work, reaching for the equivalent denotation without considering connotation, register, or the personality the dialogue is building.
- How Polilingua Would Have Translated It: "If I get out of this car, I'll hit you so hard you'll forget your own name."
Longer, yes, but subtitles in emotionally loaded scenes can afford the space. The violence is there. The character's voice is there. The threat lands the way it was written.
3. Squid Game (Again), The Expression That Made No Sense
The Show: Squid Game (Korean → English) The scene: Episode 8, a character speaks about recognizing trouble.
| What the subtitle said | What was actually said |
| "You always get in trouble to know it's trouble." | "You always eat it before you can tell whether it's sht or doenjang." |
Why it Matters: The original is a Korean proverb, a crude but vivid idiom about people who act impulsively without thinking. Doenjang is fermented soybean paste that looks identical to excrement. The expression is well-known in Korean and funny. The subtitle is grammatically broken and means nothing.
What Went Wrong: The translator attempted a literal translation of an idiom that has no English equivalent and produced something incomprehensible. This is the classic idiom trap, when translators know a language but not its proverbs.
- How Polilingua Would Have Translated It: "You never think before you act, you just jump in and hope for the best."
Idiomatic. Natural. Conveys the same meaning, acting without thinking, without requiring the English viewer to know what doenjang is.
4. Emily in Paris, The French That Was Never French
The Show: Emily in Paris (English → French subtitle for French viewers) The issue: French viewers watching with French subtitles reported dialogue that no French person would ever say, unnatural phrasing that read as a direct translation from English scripts rather than natural French speech.
Why it Matters: Emily in Paris is set in France, features French characters, and was watched by a substantial French audience. When French subtitles sound like translated English rather than natural French, it breaks the immersion for the audience that was presumably most invested in the show's authenticity.
What Went Wrong: This is a workflow problem as much as a translation problem. When a show is originally written in English but set in a French-speaking environment, subtitle translators sometimes work from the English script without enough context about how French characters would actually speak. The result is technically correct but culturally hollow.
- The Polilingua Principle Here: Native-speaker review is not optional for subtitles aimed at the target language's home audience. A French translator reviewing French subtitles for French viewers of a Paris-set show needs to be asking, "Would a Parisian say this?", not "Is this the correct translation of the English line?"
5. The West Country → America
The Show: Unnamed British series on Netflix The documented error: The West Country, a region of England covering Somerset, Devon, and Cornwall, was translated in subtitles as "America."
Why it Matters: The West Country is not America. The West Country is specifically English, has a distinctive dialect that is recognizable to British audiences, and the reference carries regional meaning. "America" has none of that. Someone read "the west" and guessed.
What Went Wrong: Geographic and cultural references are one of the hardest categories in subtitle translation. A translator unfamiliar with British regional geography will not recognize "West Country" as a specific place with specific cultural connotations. They will translate what they see, not what it means.
- How Polilingua Would Have Handled It: Depending on the context, either keep "the West Country" with a brief descriptor in the subtitle if the show needs the audience to understand it's a British region, or find the closest culturally equivalent shorthand. What you never do is replace a specific regional reference with a different continent.
Why This Keeps Happening
None of these errors are random. They follow a pattern that anyone who has worked in subtitle translation recognizes immediately.
The most common cause is speed. Subtitles on streaming platforms are outsourced under brutal deadlines at low per-word rates. One thread on Hacker News put it simply, subtitles are "a job that gets outsourced, done by the lowest bidder, under tight deadlines." When a translator is rushing through a 10-episode season with a 48-hour turnaround, nuance is the first thing to go.
The second cause is cultural distance. A translator working from Korean into English doesn't just need to know the language, they need to know what "never had a chance to study" means in the context of Korean class dynamics. They need to know what doenjang is. They need to understand why that joke about fermented soybean paste is funny. None of that comes from a dictionary.
AI makes this worse, not better. Automated subtitling tools produce text that sounds plausible. They don't produce text that is culturally accurate. An AI will translate "mi arma" as "my weapon" every single time, because that's what the dictionary says. It will never flag that in Seville, it means something else entirely.
And then there's register, the hardest thing to get right and the easiest to get wrong. Every character in a well-written show speaks in a specific way. Chandler Bing does not sound like Ross Geller. Mi-nyeo does not sound like the Front Man. When a subtitle flattens all of them into neutral, grammatically correct sentences, the show loses something real. You just can't see what you're missing.
What Professional Subtitle Translation Services Actually Include
Professional subtitle translation services are not the same as automated captioning or per-word translation. A professional subtitling workflow does not start with translating sentences. It starts with understanding the show, the characters, their social positions, their regional origins, their relationships. The translator reads the script, watches the content, and builds a terminology glossary before touching a single subtitle line.
The translation then goes through a native-speaker review, not by a bilingual person, but by someone who watches television in the target language and will immediately flag unnatural phrasing. For content aimed at specific regional audiences (French subtitles for a Paris-set show, Spanish subtitles for a Latin American market), that reviewer needs to be from that region.
What you get at the end is not a translation. It is a parallel version of the story, one that delivers the same emotional beats, the same character nuances, and the same cultural references to an audience that has never heard the original language. That's the standard. It's also why it costs more than a per-word rate and a deadline.
The difference between a subtitle error that goes viral for the wrong reasons and a subtitle translation that disappears into the story, because it works, is professional subtitling services that treat the source material as a text worth understanding, not just a file worth processing. Video subtitling services at the professional standard include native-speaker review, cultural adaptation, and format delivery for every major platform. The errors above exist because that standard was skipped.
Need Professional Subtitle Translation for Your Video Content? Polilingua provides subtitling services in 200+ languages, produced by native-speaking linguists who watch, read, and understand the content before translating a word. ISO 17100:2015 certified. Request a Quote