"The translated version of a game is never as good as the original."
"The German version of this game sucks!"
"The text does not fit in the textbox! Why did the translator not write shorter text?"
"I know for a fact that these kanji mean 'fire' and 'mountain'! Why has it been translated as 'volcano'?!?"
We all heard them, many of us said them.
I know that not all game translations and localizations live up to your high standards. In some cases I will totally agree with you that the translator has done a bad job. But often this is not what actually happened. What you may not know are all the reasons that can make a localization a bad one, even if done by a great professional. And many times things are simply out of the translator's hands.
Reason #1: Tight deadlines
When a developer promises a launch date to their fan community, they want to stick to it. An earlier launch means the developer begins to recuperate development costs earlier. But due to unexpected issues during development, files aren't always ready for translation soon enough. That's why developers press really hard to get a translation done rather yesterday than tomorrow.
The text might still be buggy—which can lead to bugs in the translation. When translation files are sent late, they are also often split among several translators rather than just one. Since every translator has their own style, this leads to inconsistency in the text and a lack of time to have it all properly proofread by the same professional proofreader. On top of that, there is often no time to receive and implement replies to the questions the translators had about ambiguous text.
Reason #2: Tight budget
Do you remember a time when there were 3 major consoles and games were still special and rare? It seems to me that more effort was invested back then to make a game a great one. However, when 500 games are launched daily for iOS alone (March 2016), development schedules and budgets contract significantly, and resources are often allocated to the most high priority items. Unfortunately, localization is often one of the secondary choices to invest resources. Indeed you will find many games who are completely or partly machine-translated.
Related post: Why Robots Don't Sell Apps in Germany
A good translation costs good money. After all, going to college and gaining deep knowledge of a language—with all its fine nuances—is a long, time-consuming investment. And since money is something many small developers lack, they might decide to go with someone who studied German for a year in high school and will do it for cheap.
I don't condemn fan translations. Fans know and love the game—and doing something with love generally means doing it well. I see it as a better option than paying a wannabe-translator. But not all gamers are linguists or good writers. Not all of them might be prepared for the challenges of translating a game.
Related post: How to Translate a Game
Reason #3: Lack of communication
Have you ever played the game Chinese whispers? One person whispers something to the next person who will whisper it to the next who will then whisper it to the next one, and so forth. Ten people later, the original message is hardly ever recognizable.
It is not so different when a game designer communicates his or her wishes through several middlemen. A message regarding a change in the game might have to go from the Japanese designer to a coordinator to a Japanese-English translator to a European producer to a project manager and two translation coordinators. By the time it reaches the final translation team—if at all—it might not be what the designer originally intended to communicate. It is not rare that the developer makes crucial changes in the game (age rating, ending, renaming characters, etc.) without the translators knowing a thing about it.
Reason #4: Lack of information
Imagine organizing a party for 200 people. You would ask yourself: What is the party about? How old are the guests? Do they work in the tech industry, or are they retired academics? Are there any vegetarians?
Receiving information about the characters—what they are like, how they look, what their profession is, what they long for, what they fear, yeah even if they are male or female—is often very hard information to get. But knowing these things is crucial for the outcome. Many developers are not prepared to send out such information, although they surely have it somewhere. One of the first things you learn when taking writing classes is that every character has a back story. Even if the reader never knows about it.
The more information the translator gets, the better he can translate the developer's vision. Age rating, characters, background about the game universe... Those are all pieces of information that matter. They help make characters pop and bring happiness to translators and gamers alike. And they are often sorely missing when a translator receives translation files.
Related post: How Not to Make a Game Localization-Friendly
Reason #5: Lack of context
This is the biggest issue of all. I don't know how many games I had to translate blindly, with no build being available to play the game, even on projects that took many months. How many times I had to ask questions like "Is this character in cell 321 the same character as in cell 345?" or "Is that single word "Plant" in cell 5012 meant as a verb or as a noun? "Who is speaking here, and to whom?" The sad truth is: We often don't know what we are translating.
Developers do not always see the translator's perspective and why we don't understand a short, seemingly simple piece of text. They may not speak fluent English either.
It is not rare to receive this simple reply to a plead for more context:
"There is no context."
I kid you not.
Reasons #1, 2, and 7 often make it impossible to spend more time on investigation.
Reason #6: Too many cooks
A 2 million word MMO will have to be translated by several translators if it is to be launched within this century. And such a project will do best with a translation agency.
However, a small project of some ten thousand words might be better off being translated by only one translator. If the developer hires a freelance translator, they know who will be responsible for the whole translation. If they choose to work with an agency, said agency will assign the project to whoever is currently available. And they will split the work if a translator cannot handle it alone within the proposed deadline. If a translator only translates a fraction of a game, they might feel less attached to it than if they took care of a complete localization. It's not their baby.
If the developer sends files in batches or as regular updates to an agency, chances are the previous translator won't be available and the project will be assigned to someone else. This leads to inconsistency in translation choices and style.
To top it off: The agency might offer the project to someone who is not specialized in video game localization—and if we're unlucky, that translator will accept.
Reason #7: Low translation rates
The job market is currently exploding for bilingual people who try to make it as translators. Nothing against people earning something on the side with their secondary skills, but not all of them can call themselves professional translators. And most of them do not know (or care) how to produce a great game localization. Since translation agencies have to make a certain margin in order to sustain themselves, they might choose to work with the cheaper translators which often are the less experienced ones.
This vast pool of translators, pseudo-translators, and translation agencies underpricing each other leads to all of them having to work twice as much to make a living. And working twice as much means having half the time to do things properly.
Reason #8: No localization testing
Nintendo, as my very first employer, spoiled me deeply. Nintendo games are known for their great localization—and for good reason. They hire excellent translators, pay them fairly, and give them time to get familiar with the game before they translate it. That's where it usually ends for most game translators, if we are extremely lucky.
But Nintendo doesn't stop there. They also have amazing localization testing teams who know the game inside out and are given an ample amount of time to check it thoroughly. All text is seen in context, meaning on-screen, by several ambitious localization testers who make it their mission to polish the game as best as possible. At least that's how it used to be.
I cannot stress the necessity of a proper localization testing period enough. Yes, it costs extra time and extra money—but its results are so vast that it should not be omitted. The best proofreader won't do as good a job as a dedicated tester who knows the game well and is trained to question any text in context with a critical eye.
Related post: Bulletproof Proofreading Tips—How to Catch Those Bugs
Reason #9: The developer made a mistake
A big title of over 2 million words I have recently worked on together with an awesome, experienced team has received some unexpected criticism. Fans noted that some of the (German) in-game text appeared in French. Some of it was English. They immediately blamed the translators for forgetting to translate those parts or for translating it into the wrong language.
Seriously: Do you think I would accidentally translate a random sentence into the wrong language? Does that ever happen to anyone? Please. Obviously something has gone wrong on the developer's side.
Reason #10: The developer hasn't gotten around to implement the text yet
Another part of the criticism of that very same fan community was that a lot of the text they got to see on the German launch date was full of all kinds of bugs: Mistranslations, text overflows... you name it. People mentioned that the developer has been robbed by amateur translators who tried to make a quick buck by delivering a machine-translated game.
I almost got a heart attack looking at those screenshots. They were right! What were we thinking?!? I immediately looked at my textfiles and realized that this was text from months ago. It has long been polished—and not one of those bugs that were shown as examples of a job poorly done were existent in what we delivered.
Reason #11: The developer had a tiny update and took care of it on their own
Sometimes the developer might realize that there is a tiny part in the game they forgot to send to the translator. Or maybe they added a new line of text that they wanted to implement right away. From an economical point of view, it might have seemed to be the best decision to machine-translate that text. This can lead to unintended funny results since such translation tools are not good with putting things into context yet.
This might suck for players. But it's the developer's right to do to their game as they please—and we translators cannot guarantee for what is being done with the text we deliver.
Reason #12: The translation is excellent and you don't know it
How many times I read gamers' complaints online about a translation choice not being "correct"... Strictly speaking, games are supposed to be localized, not merely translated. Localization is not about being accurate. It is about creating an engaging text that will help gamers of a specific region immerse into the game universe. A word-by-word translation does not make for a fun gaming experience. Don't ask yourself whether "bläh" would be more correct than "blah". Ask yourself whether you're having a good time playing this game.