Focus

Localizing a whole new Western world

animated characters

Loreto Fueyo joined PTW in January 2019 as director of translation after 13+ years of industry experience. She holds a BA in English philology from the Universidad Complutense in Madrid, an MA in comparative literary studies from Kent University and an MSc in computational linguistics from King’s College London.

Loreto Fueyo headshot

Loreto Fueyo

Loreto Fueyo headshot

Loreto Fueyo

Loreto Fueyo joined PTW in January 2019 as director of translation after 13+ years of industry experience. She holds a BA in English philology from the Universidad Complutense in Madrid, an MA in comparative literary studies from Kent University and an MSc in computational linguistics from King’s College London.

V

ideo game localizations are unlike any other translation job. In addition to translation, there’s tone, story adaptation, context-dependant conversation and endless player interactions to consider, not to mention the sky-high expectations from seriously devoted fanbases. But localizing a game is also incredibly rewarding when you’re given the creative freedom to try new things, and especially when those things go right. And that’s just what happened when we localized LEVEL-5’s Yo-kai Watch 3 in 33 weeks.

Yo-Kai’s deep, deep history

To fully understand what a huge undertaking this localization was, it’s first important to understand Yo-kai Watch’s historical context.

For those unfamiliar, the first Yo-kai Watch game was released in Japan in 2013, setting the foundation for the rest of the series.

The Japanese sequel was released in 2014 with two different versions: Yo-Kai Watch 2 Honke (known as “Fleshy Souls” in English) and Yo-Kai Watch 2 Ganso (“Bony Spirits” in English). These two versions expanded the series considerably, not only with many new Yo-kai characters and locations, but a time travel element that multiplied its complexity.

In 2014 a third version, Yo-Kai Watch 2 Shin’uchi (“Psychic Specters” in English), was released to coincide with the first Yo-kai Watch feature film, Yo-kai Watch: Tanjō no Himitsu da Nyan! – a film that outperformed several international blockbusters at the Japanese box office.

What’s more, since its inception, Yo-kai Watch has featured in a range of top-grossing films, anime, manga and toys.

However, more than keeping consistent with the existing releases, we were also localizing Japanese folklore. You see, Yo-kai are a class of supernatural monsters, spirits and demons embedded in Japanese folklore that date back to the oldest known work of Japanese literature. And it’s this folklore that makes up the basis of their mythology, Kojiki: Record of Ancient Matters.

Given this strong heritage, the pressure to get even the smallest nuance right was significant.

Getting to know the characters

During Yo-kai Watch 3, players meet over 600 Yo-kai characters. As you can imagine, this makes managing and adhering to a glossary for all the Yo-kai — which contained not just their names but all their characterization information — a mammoth but necessary task.

This glossary kept track of every character’s details from across every piece of existing intellectual property. What’s more, it had to be constantly synced to maintain consistency across Yo-kai games. This was an extremely labor-intensive process that included cleaning up inevitable inconsistencies and updating existing terms based on new information and client feedback.

This level of research is paramount to any game’s success. With a clear catalogued understanding of the game, every team member can be on the same page. But to get this right, it’s important to establish a clear line of communication between all teams so when a change or inconsistency does arise everyone knows how to resolve it.

BBQing Yo-kai’s language

The previous Japanese versions of the Yo-kai Watch games took place in a Japan-inspired location called Japon. Half of Yo-kai 3’s story is centred around Nate (the protagonist) and his family, who move from Japon to an America-inspired country.

In the previous English versions of the games and anime, however, Nate and his family already lived in an American-inspired town called Springdale. So, the English version needed a new location for Nate and his family to move to: the made-up region of BBQ. Changing the location created more freedom for the linguistic choices that followed.

For example, in the Japanese version, Nate struggles with the Japanese to English language barrier, which causes several linguistic and cultural misunderstandings for players to navigate. The scenarios were often portrayed with confusing English phrases appearing in characters’ speech.

Of course, this wouldn’t make sense for the English version. Thus, the “BBQ accent” was born. So, instead of Nate being confronted with standard English, he was met with almost indecipherable speech patterns somewhat loosely drawn from caricatures of rural or Southern US accents. “Did you just move here?” was replaced with “Y’all jus’ a-move on ova hair?”

The BBQ accent was a creative solution different enough from the original standard English dialect to justify the existence of other elements in the game, too.

What followed was a considerable amount of research to create a detailed and consistent accent that had to be clearly established before the localization could start.

In order to keep our fabricated Southern-inspired accent consistent, we created reference files to break down the BBQ accent in detail. Here’s an abridged guide to speaking BBQ:

1. Shed syllables. In almost any three-syllable word, you can contract out any vowel from the middle syllable to make a two-syllable word.

Examples: Italy = It’ly, Florida = Flor’da, Johnathan = John’than, Melody = Mel’dy.

2. Forget the “G”s. Never, ever pronounce the “g” in words ending in “ing.”

Examples: Fishin’, Cookin’, Readin’ and Writin’.

3. “L”s are optional. Ignore “L”s following vowels in the middle of words or simply replace them with “W”s

Examples: Light Bub (Bulb), Code (Cold) and Caw (Call).

Then all the linguists tasked with the English localization received a full induction on consist rules for BBQ speech.

Throughout the project, communication between translators and revisers was constant and essential to create a fully developed and consistent new world. Often, team members would raise specific questions in our internal chat system on how to BBQ a sentence or how a certain scenario should be dealt with. These internal chat channels were a very useful tool for the teams, with each language having their own channel that spanned the translation and the localization quality assurance (LQA) teams.

English-language version of a video game
Foreign-language version of a video game

In the English-language version, a dialect was created (top) to replicate a sense of foreign-language confusion (bottom).

BBQ’s legacy

The BBQ style even spread to other parts of the game. By befriending a certain Yo-kai with interpretation powers (the Lionguist), Nate can understand the foreign language being used by the locals. For the Japanese version, this meant that most of the game could be continued in Japanese.

In English, however, even with the Yo-kai interpreter, we had the characters continue to speak with a moderate accent. So, although there wasn’t a difference in the Japanese version, for the sake of creating an interesting and fully fleshed-out world, the localization team made the decision to go the extra mile and have all the BBQ characters talk with an accent, albeit diluted from full-on BBQ.

FIGS was a different story

For the French, Italian, German and Spanish (FIGS) localization teams, these location and dialect issues weren’t such a problem because each game was originally based in their respective countries. For example, in the Spanish version of previous games, the main character lived in a Spain-inspired country and everyone spoke Spanish.

So, the FIGS versions used regular English rather than BBQ English to recreate the same sense of bafflement in a new country.

While the FIGS translators and revisers weren’t challenged with creating a dialect for their respective versions in place of BBQ English, they were faced with translating from the BBQ dialect into standard English, which at times proved particularly difficult.

Communication between teams pays off

Communication between all of the language teams, including English, was essential to localizing each scenario in the best way possible.

Communication was necessary not just for translating into or from BBQ, but also for creating dialogue that was both entertaining and relevant.

A beautiful example of this can be found in the scene where another character, Hailey, befriends Treetter, a tree Yo-kai that can only communicate through social media. Here, the localization team made use of the situation to introduce a number of jokes that were not present in the Japanese version.

The scale of the localization

Yo-Kai Watch 3 had over 1.6 million Japanese characters and kanji that needed translating into English, and more than 600,000 words to then translate into FIGS.

The Japanese to English work comprised of familiarization, a full localization cycle, script preparation and LQA. The entire process took 33 weeks, with LQA starting before the translation was complete.

Similarly, LQA for FIGS commenced before the translation phase had finished. This meant that, even though we started the FIGS translation five weeks after English, the LQA for both projects were completed at the same time.

It’s this type of concurrent work that allowed us to compress delivery time without unduly increasing team sizes and thus affecting quality.

The keys to a successful game localization

There were four key elements that were vital to this project’s success. I’d put them in this order:

1. Research and prepare. No amount of time during or after the work will make up for lack of preparation. It really pays off, solves problems down the line and sets the project up for success.

2. Dare to try. Embrace the chance to be creative and go for it. Give your teams the freedom to play. But keep in mind, this isn’t possible without having already done the above or having the following…

3. Create a forum for discussion, not just within the team but also with the client. Open and transparent communication throughout the whole process will pave the way to triumph.

4. Link to open and transparent discussion and schedule through partnership, not just with your client but also with internal teams. When you start a FIGS translation long before the English one is complete, or LQA while the game is still being localized, working together as one unit is the only way to make a successful project happen.