Post date: Oct 25, 2017 1:28:31 PM
Last weekend we got a visit from an old friend and now successful game developer, to whom I haven't spoken in years. He got hold of our company over LinkedIN and he wanted to know if we could do some 3.000 words for the next day. Considering we never worked in his product and he didn't even have any reference kit prepared, I asked for 3 days. One to get familiar with it and prepare a proper kit, one to translate and a morning to get it review. He was stunned. The game already had some 60.000 words of content, with monthly 3 to 5k updates, usually done by his current localization vendor in 24h. And he had never received any other reply from his vendor than "OK!".
The conversation developed into him sending the update to the usual vendor, while Locsmiths would do some spot checking and check the current content database. While the linguistics were OK-ish, the text was a total chaos and mayhem regarding terminology, capitalization, etc. Clearly had been made by several different translators along the time and without any references or glossary support. He was stunned again. He had never heard of "shotgun approach loc" (when vendors send the request to the database and first to reply gets the job). He had never heard that a good baseline for EN>PT production is around 2500 words per day. He had never heard that you can create Termbases in CAT tools and have the project consistent term-wise if you spend time making a decent project, for instance on memoQ. He instantly assumed the vendor would care so much about his product as he does. Saying OK to everything is NOT OK! If your vendor always says OK, be very afraid.
So, now he knows how things work and is stuck with a 60.000 pile of hog-wash that he can't afford to update, as the game is about to be discontinued. Sure, he was happy to send 5.000 words on a Friday and get them back on Monday morning. He never actually needed that fast, he just thought it was the norm since no one told him otherwise and he had read it somewhere online "up to 10.000 per day". But I bet the players who played his game weren't happy trying to find Gear (cogwheels as it was translated) when in fact Gear in the game just refers to pieces of equipment. Sad story ends here, Locsmiths is already handling the new game for him.
I never cease to be amazed about the "fear" or "yes, sir" policies that are jammed between developer/producer and the localization process. "Accept everything", "Follow the instructions blindlessly", "Never question", etc, etc. Does this really help the developers/producers? Is this really what you want from your localization partner?
Does this mean localization has to be a tug-of-war between provider and requester? Of course not! But also doesn't mean the developer/producer is always right, always sure about what they want/need, or even confident it will all work out as they think. So, to discuss, is to care. Is to care about your work, is to care about the original product, is to care about the localized product. We just want to be able to do our job the best we can and the biggest gainer on that equation is you, Mr. Developer/Publisher. They say "don't trust a bald barber", we say "don't trust a Yes Man vendor".
Bruno Dias