How long does it take to translate a video game?
People usually think translation can be done fast because "the text is already written", but we must remember that localization is a creative endeavor, so it takes a lot of time and energy to make something worthwhile.
Usually, in Brazilian Portuguese, we consider that one translator takes closer to 8 hours to translate 3,000 words. Obviously, this isn't a set limit, but an approximation, because the actual volume someone can do depends on a lot of factors, including the text quality, how easy (or hard) the text is, how much reference the translator has, the context (dialogues are usually faster to work with than menus, for example), and so on. And, of course, it also depends on the languages themselves.
For a review of the translation, as long as the text is well written, we consider it takes 1/3 of the translation time, so a reviewer can do something like 9,000 words in 8 hours.
Now, that all seems fine, but deadlines in video game localization are usually tight, so more people need to work on the same project at the same time. However, the more people a project has, the more uneven the tone can become, and the bigger the chance of having inconsistencies. No matter how well the project is documented, the tone and style are defined, and how much reference everyone has, each person has their own writing voice and it takes a lot of energy to adapt that and follow external guidelines. I'm not saying it's impossible, but the more people involved in a project, the higher the chances of something going amiss.
You know the phrase "a woman can make a baby in 9 months, but 9 women can't make a baby in 1 month"? Creative translation works kinda the same way: although we can make a 9-month job fit into a 1-month deadline, just... don't expect the same level of quality. That's why proper planning is important.
Most of the time, the localization process is ongoing while the game is in development. This means we translate content as it is written, and we edit the translation as the content is updated. In an ideal world, the best approach would be to have the text finished before sending it for translation, but in a lot of cases, the writing is only finished a few months before release, which wouldn't leave enough time for a proper job.
Given the realities of game development, it's hard to keep a good schedule, so it's important to have good communication with the localization team. This means warning them beforehand when a new batch is being prepared for translation, or sharing the expected scheduling so they themselves can schedule translators and reviewers, trying to always keep the same people on the project — helping to keep a consistent tone and style.
The reason I'm saying this is because, more often than not, translation batches arrive without prior communication, so we scramble to schedule people on short notice, and with tight deadlines that aren't ideal for this kind of job. I mean, just because a translator can work 3,000 words a day, doesn't mean that they can receive this volume today and have it translated by tomorrow (especially when things like time zones are factored in). It is very common to receive batches of 500 words, sometimes even 1,000, with a same-day deadline, which is absurd. Despite some agencies loving to do stuff like this, the quality always suffers. Translators are humans and need some time to acclimate to the text and devote energy to them.
So, all things considered, here's some math about the ideal time it would take to translate something, and then how it usually goes.
Let's consider a game with around 100,000 words. It seems like a lot, but believe me, it's not that much for a game with around 20 hours of gameplay and plenty of dialogue. So, 100,000 words, divided by 3,000 words per day means it's a 33-day job for one translator. Taking into consideration weekends, and adding a bit of leeway for unexpected things that might happen (like going slower when translating menus and items, or having questions and waiting for answers, or having holidays along the way), it's safe to assume it's a 2-month job for one person to translate. Let's add 2 or 3 weeks for a reviewer to proofread the work (remember, if the translation has good quality, the reviewer only needs 1/3 of the translation time). Rounding a bit up, it would take around 3 months to have 100,000 words completely translated and reviewed.
This scenario considers having the 100,000 words ready for translation while the rest of the game is still being developed, that the 3 months it would take for the job to be done wouldn't have any major incidents, and that the translation would be ready before the game launched. It's a pretty optimistic scenario.
What usually happens is:
A batch of 20,000 words arrives with a one-and-a-half-week deadline (to be translated and proofread). Then another 5,000 pops up 3 days later with a 4-day deadline. Since none of this was previously communicated, the translators are already working on other projects, so more people have to be added. Sometimes we can negotiate deadline extensions, sometimes not, because it all depends on the studio's or publisher's milestones or internal planning.
Then there's a 2-week period without a peep, then suddenly a drop of 30,000 words with a 1-month deadline. One week later, 10,000 more words for the same deadline. 1 month later, after everything is finished, there's a month of silence, then small batches here and there ranging from 1,000 words to 10,000 words.
While all of this happens, some of the text got updated, be it to update the story, accommodate something that wasn't initially planned, or remove stuff that wasn't working. After 3 months, the translators might've worked on 120,000 words, but the game still has 30,000 words to translate that haven't been edited in the original language yet.
Bear in mind that there's also the time spent checking references, asking questions and waiting for answers, or something went wrong with one of the batches and it needs to be corrected for 5 or 6 languages at the same time without breaking any of the work already done, and so on. Meanwhile, a team originally planned to have 2 or 3 people had grown to 5 or 6 to deal with the deadlines, so every translator has less work to do by themselves, meaning they also have to accept other jobs to make up for it, making it more difficult to schedule batches that suddenly appear without a warning.
It's understandable that, given how game development is full of challenges and unexpected events (in and out of production), it's hard to make a schedule and follow it through. That's why the more planning and communication between the parties involved, the better, as it allows them to quickly deal with unforeseen issues while trying to keep up with the pace.
That's also why I recommend avoiding things like "urgent text for translation", that so many companies do. Even if it's 100 words — something that can be dealt with within minutes — asking it to be delivered on the same day or the next morning is like letting a speck of dust get inside a Swiss mechanical watch: it might not seem like a big deal but, in the long run, it will accumulate issues, especially if it happens frequently.
The takeaway here is that the longer someone can work on something, the better that something will be, and translation, especially when talking about localizing a media as complex as video games, is no exception.
So, ideally, translating a video game takes as long as the developers can wait for that translation to be finished.