Crowdsourcing is the new buzzword these days. Apparently people shouldn't be paid when their job is fun, and therefore both game translators and game developers are expected to work for free. Do you like your work? Then you will no longer be paid. Hah, take that!

The idea of crowdsourcing is indeed a very attractive one. Localization can be expensive if carried out the wrong way, and by gathering an enthusiastic crowd that would love to translate your game for free, you can save quite some money. To make crowdsourcing even easier, several crowdsourcing platforms have been launched: you register, you upload your app and off you go.

Now, there are several reasons why crowdsourcing can fail before you have even started. First, you need to have an interesting product that is actually fun to translate, so you can forget about your End User License Agreements.

Second, you need to have a large following, as else you won't be able to gather enough people willing to translate your product. And a large following is exactly what non-established game developers don't have.

Third, most developers have come up with a unique concept they don't want their competitors to know about... yet. And the more people know about the inner workings of your product, the bigger the chance that your unique concept becomes public domain before the game has even been launched. Professional developers have their translators sign special NDA's that heavily penalize them if they ever talk about your product to anyone else. In that light, throwing your entire code in the cloud doesn't seem like a very good idea.

Fourth, console manufacturers like Nintendo, Microsoft and Sony want you to use very specific platform terminology in each language you translate to. Professional game translators are trained to use these terms, which are updated monthly by the console manufacturers. If the translators get even one term wrong, you may need to resubmit the entire game. And when you realize the cost of that, using even the most expensive translation agency in the world suddenly becomes a very attractive option.

Fifth, if you hand a text to thousands of translators, thousands of them will use different terms for everything, so that every single item in your game will end up having dozens of different translations. "Please bring me the staff. Ah, I see you have the rod already. How about me enchanting the stick?" Only to refer to it as stave, wand, cane, pole or club a few lines later. If you crowdsource, say goodbye to consistency, unless you crowdsource consistency checks too. Do you?

Sixth, having the word American stamped in your passport does not make you a great English translator. Read the English of the users on your forum and you will notice that many of them can't even spell their own language correctly, let alone translate to it. But!, I hear you say, the translation won't be carried out by just one person... it will be carried out by the crowd: people can upvote and downvote translations and in the end only the best translations will survive.

Good point. So let's see what actually goes on behind the screens of one of the best known crowdsourcing platforms out there, which we'll call Matata.

I registered myself as a translator on Matata... all they wanted to know was my target language (Dutch), my name (Loek) and my e-mail address (see my contact details). Apparently that's all you need these days to become a reliable and professional translator. Nothing was asked about my educational background, and I could have specified Russian and Swahili as two other target languages I translate to (I could actually try, and the results would be interesting to say the least).

After the registration was completed, I could take a look at all the apps and other texts "developers" had submitted for free translation by the crowd. See the table below and have fun (you can click on the images to enlarge them)!

	Most of the projects behind the screens are not texts about apps, but merely one-liners submitted by curious individuals who want to approach their new love targets in random countries. Which may explain why your app will literally drown in hundreds of other messages from people looking for a freebie. If you ever wonder why crowd-translating your app takes several months, this explains why: you're competing with Mr. Casanova!
	Here's another reason why people may never see your apps. Matata's site is bug-ridden and often translations don't even show up. But who cares about technology... as long as the translations are reliable, right?
	A drowning app fighting for its life on page 48 of the site.
	I just started clicking randomly on the translations that had been submitted by people claiming to be Dutch. This particular user made at least 4 spelling errors in just two sentences, an accomplishment in itself. This despite the fact that his "reliability" score is 551, meaning that he belongs to the 40 best translators in the translator pool (and the best Dutch translators on the whole site).
	This user is even better. His reliability score is 875, making him the crème de la crème of this site. Only 30 people are better than him (other languages than Dutch included!), and look how brilliantly he translated the line “Waves of coming in enemies” What? That's not what you meant? Translators come to these platforms because they have too much time on their hands, and they have too much time on their hands because they don't have enough work to keep them busy, and they don't have enough work to keep them busy because clients think their translations suck. Yes, these are the people who are going to translate your blood, sweat and tears. Because they suck.
	Another app submitted by Matata's target group. It contains only this line, and it's meant for the newest dating app to be launched in the States and China. Or maybe it's for someone watching online peep shows. We'll never know.
	This translator managed to even spell the word "I" (the personal pronoun) wrong. It's "Ik" in Dutch, but then again, getting like 50% of the translation for 0% of the price is still a great deal, of course.
	The same translator again, and he seems to be on a roll. One blatant error in the grammar here, and one word left in English. Then again, just two errors in one segment is not annojing ad all.
	Some translators are very considerate. They put multiple suggestions in the same translation, so that you can pick the best suggestion yourself. But how are you going to pick the right suggestion if you don't speak the target language? How are you even going to spot suggestions in the first place? I'm trying to picture how that will look on your splash screen: "Press or push space to start or begin the game or adventure!"
	Here we see three sentences, and all three of them are grammatically incorrect. As you can see, none of them got any upvotes or downvotes, leaving the developer (you) with 3 absolutely ridiculous translations. The last line is from a user called RKAT, actually the 3rd best translator on the whole site. RKAT has 4785 points, which makes him almost godlike. However, I count four incredible and very grave errors in this very single line. It looks more like Russian than Dutch. Or maybe it is Russian? No matter what, with friends like RKAT, you don't need enemies. Crowdsourcing, Google Translate and all the other hypes are not the magic wand they pretend to be. I'm sure your mother taught you that there simply is no such thing as a free lunch. If this is indeed how you are planning to localize your game, maybe you shouldn't localize it at all.