Friday, March 28, 2014

Impressions from the Localization Summit at the Games Developer Conference

I recently attended the GDC conference and were impressed to see how advanced the tools and thinking on translation and localization was amongst the game developer community. This is a new frontier for the professional translation industry and it is amazing to see how quickly games developers are learning about localization issues and developing tools and methodologies to rapidly make their products viable in international markets. It is commonplace to see development tools that are truly both multi-platform capable and multilingual capable in active use throughout the community. Their comfort level with multimedia (video, voice) data, multi-platform development (mobile, desktop and TV) issues and willingness to consider technology was quite refreshing, and we think it is quite possible that this community will drive translation and localization technology advances much more rapidly than the traditional software and documentation community where things change much more slowly. While the use of MT is still very limited, the exploration of the viability is much more pragmatic and informed than the torturous trail we have seen in the traditional documentation translation business. This is perhaps because things move much more rapidly in the games business and most products have short and intense life cycles and behooves the developers to  quickly capitalize on international opportunities while their games are hot. We predict that MT will be more frequently used there in future as developers become more informed about the limitations and possibilities of MT.

This is a guest post by Lauren Scanlan that provides a quick summary of some of the localization related presentations at GDC.Lauren Scanlan is currently a Freelance Localization Professional and manga proofreader looking to dive into the wide, fun world of game localization. She's based in Denver, where she spends her time baking, reading, and wrangling the odd dinosaur. You can find her online at @lsscanlan,, or connect with her on LinkedIn at

On March 18th I attended the Localization Summit at Game Developer's Conference (GDC) 2014, which, in truth, was the whole reason I came to GDC in the first place. I've been working in the localization industry for three years, and as an avid fan of video games, I thought that combining the two might be a good fit for me – and I wanted to see what the state of localization is within the gaming world.

The Summit, organized by Fabio Minazzi (Localization Summit Advisory Board Member and Account Manager at game localization service provider Binari Sonori) had a variety of talks – from emerging markets (Localizing Games for Spanish Speaking America, Emerging Communities: A Snapshot of the Brazilian Indie Game Development Scene), to culturalization (Journey to the West: A Chinese Game Localization Primer) to how localization can be improved, either through communication with an LSP (Indie Games Localization: Is It Worth It?) or improving tools and processes within your own localization department (The Future of Localization Testing, LAMS: Building a Localization Tool for Everyone – both talks given by Sony Computer Entertainment Europe senior employees). 

There was also a series of five mini-talks called Localization Microtalks: Globetrotting in the Fast Lane, which covered indie game localization, an open-source localization frameworks, localized advertising, mo-cap dubbing technology and processes, and using app description localization as a tool to “test the waters” for localization. All of these talks should be available as part of the GDC vault, and are well worth a watch for those with Vault access!

The two final talks of the day were uncomfortable for some of the people in the room – Crowdsourcing the Localization of Gone Home and What is the Place of Machine Translation in Today's Gaming Industry? Both took a look at methods that do not necessarily rely only on trained linguists to complete translations, and have the potential to make localization accessible to others without the budget (more so in the crowdsourcing talk than the machine translation panel), even if it is not 100% perfect or within total control of the creators.

It seems to me that, while things seem to be going well in game localization, there are areas for improvement. The first concentration seemed to be on how to get indie games, and especially indie mobile games, out to international markets, including emerging markets, which may not yet have a framework for distribution which works well for the consumer. The issues that face most indie games are cost management for the studio and accessibility for the gamers. Belén Agulló García, Language Production Manager at Pink Noise and Jonas Waever, Creative Director at Logic Artists (Indie Games Localization) mentioned that, as localization is a part of marketing, it made more sense to use the marketing budget for their localization. I thought this was a fantastic idea, since (as all indie studios that participated in the Summit would attest) they received sizable returns on their localization investment, which could then be invested back into marketing. 

Martina Santoro from Okam Game Studio and Alejandro Gonzales, CEO and Studio Director at Brainz (Localizing Games for Spanish Speaking Latin America) mentioned that some mobile games, especially freemium games, were hard to bring into the Latin American market because of the limited payment methods. Prepaid cards and vouchers seem to be popular, but this is certainly an issue that needs to be addressed as new markets start opening up. For entirely free-to-play games, or games on platforms that are already taken care of, this may not be an issue – and may make the market more enticing. Accessibility is going to be a growing issue for both localization service providers and distribution platforms, and I think a collaborative effort to grow in these new markets and find solutions would be great.

The next area for improvement seems to be in translation and culturalization. For example, contrary to what I've done in general localization, it seems that since the app store only lists two kinds of Spanish (Mexican Spanish and Spanish), it's better to localize an app into neutral Spanish to make it easily understandable for the LATAM region. This illustrates well that distribution platforms play a large role in how effective (or not!) localization can be – though I was also pleasantly surprised to learn that the App Store and Google Play will spotlight and promote localized games in the relevant market, which has meant huge success for those games overall. Shaun Newcomer, Vice President of Reality Squared Games (Journey to the West), gave a great talk on some of the culturalization issues of bringing Chinese games to the US market. For example, games in China usually have a short life cycle and high monetization, and the storylines can be of middling quality, since the games are not expected to last very long. However, US/Western consumers dislike high monetization in the interest of fairness, and look for a higher quality storyline, leading Newcomer to amend both of these things in-house before releasing games to the US market. I had never before considered aspects like monetization as a part of the localization process, but it makes sense to find out which markets lean more toward monetization and which shy away from it. Crid Yu, Vice President and Managing Director of North America for InMobi (Localization Microtalks) mentioned that it was also important to look at alternative marketing methods for each market, citing as an example using subway advertisements for mobile games in Seoul, South Korea. I thought this was a great idea – having been to Seoul myself and seeing the amount of uber-connected Seoulites with their phones out on the subway, always looking for the next new thing.

Finally, the last issue is how much Machine Translation should play a part in game localization. I've been working in Linguistic QA for three years now, and when I first started, I was trained on how to use SDL tools like Trados and Studio as part and parcel of the localization process. Since we usually leverage it only on material that needs to be the same year after year (employee satisfaction surveys) or technical specs (medical devices) that generally don't change, it makes the localizations far more cost-effective for our clients. However, games are full of mostly creative, non-repetitive text, which makes this method dicey, as the quality can be lessened and the text can sound stale if it's always repeated. One of the best uses that came up within the panel (What is the Place of Machine Translation in Today's Gaming Industry?) was for live chat – it would be great to use an automated system to instantly replace the most common words in a sentence, or the most common sentences – that way, players on European servers (for example) could easily get the main points across to each other while not interrupting the gameplay. Another good application, I think, would be to keep a TM of UI terms that won't often (if ever) change, such as “Play,” “Continue,” “Game Over,” etc. Also, depending on the intended quality/lifecycle/churn of the game, it may be better for some companies to use machine translation and go with an imperfect translation if it gets the game through faster... though, as a Linguistic QA-er, I am certainly not advocating lower-quality games!! 

 I also thought Christopher Burgess, Senior Programmer at SCEE (LAMS) did a wonderful job talking us through the struggles with building the perfect project management/localization software, showing that even customizable tools can be a challenge to work with to get the results you want.  Even when using Machine Translation, the users need to be well-trained and always should think critically about what they're using the MT software for. 

Overall, I learned quite a bit through the Localization Summit, and I think it's a wonderful addition to the GDC programming. It was great to meet so many people who were just starting out in game development, or just about to localize, in the audience! I'm glad that non-localization people are taking the time to educate themselves about localization and their options, and I only hope that others join them next year. I'll certainly be there!

1 comment:

  1. It's an interesting post about localization and I would say that it's better to localize an app into neutral Spanish to make it easily understandable for the particular region as this illustrates well that distribution platforms play a large role in how effective localization can be.