Analyzing Language Poll Data for my Indie Game
As part of the second stretch goal for the Diorama Break Kickstarter, I ran a poll for backers to see which language they’d like us to guarantee a localization of in the full game. To cut to the chase a little, the winner is…
¡Felicitaciones!
For those wondering about regional variants, our aim will be to provide a satisfactory experience for all Spanish speakers, but with almost twice as much support in the poll, Latin American Spanish (Español latinoamericano) will be our focus.
Full Results
We received over 260 responses to the poll, though as this was intended to be a backer-exclusive poll, I unfortunately had to reject about 15 respondents that provided invalid credentials. A few other credentials had typos (or provided a username instead of a backer number), but I was able to match them to valid credentials, so no worries there.
Question 1
“What additional language(s) would you like to see us guarantee a localization of in Diorama Break?”
Here’s the breakdown for the main question:
As you can see, Spanish was quite the overwhelming victor, having been selected by over 50% of respondents (as a reminder, the question used approval voting; respondents could select as many languages as they wanted). Here’s a breakdown of how many languages people chose:
Some other interesting takeaways:
I expected Portuguese to be a potential winner, or at least a strong runner-up, but it’s been surprisingly edged out by German! Still, many more Germans speak English than Brazilians/Portuguese, so it’ll be interesting to see how the breakdown changes when filtering by Q2 response.
Despite being at the top of the poll’s list due to having the largest Steam population, Russian didn’t receive too many votes, ending up beaten out by Polish and Esperanto.
Excluding the languages I expected to be heavy hitters (Russian, Spanish, German, Portuguese, and Korean), the conlangs (Esperanto and toki pona) ended up being the most popular!
4 people responded just to write in an answer to the effect of “N/A”. I guess they just really like responding to polls?
We received about a dozen write-ins for languages we’re already planning to translate to1. That’s just how dedicated we are to customer service though, delivering what you want before you’re even aware you’ll be receiving it.
For those wondering about Quebecois French specifically: We’re based in Quebec, and the translation will most likely be completed by people here in Quebec. For the demo’s translation, I deliberately tried to avoid overly colloquial language, so the one translation could be enjoyed by both French and Quebecois players2. That’ll be the philosophy going into the full game’s translation (for French and other languages), though with more time to plan and review we may be able to do things a bit more interestingly…
Question 2
“Why do you want to see the game translated to the language(s) you chose? Select the most important.”
This question was intended to help us determine exactly why people picked the language that they did. One thing I hadn’t anticipated when launching the poll was how many people would write in that they picked their language(s) because they believed it would get the game more sales/attention. I ended up adding that as a standard response, and basically all write-ins here ended up mapping cleanly to one of the standard responses (or, again, N/A). Here’s the breakdown:
A somewhat even split, though wanting to share the game with others was the most popular reason by a good margin. The more interesting takeaways come when you start filtering the Q1 responses by Q2 reason:
Pretty similar to the base breakdown, though this reason had the most write-ins by far.
Also pretty similar to the base, but German takes a big hit here. Perhaps German speakers feel like they get enough attention as it is.
Ah, now we have our answer. Clearly, German speakers think they have more attention to give than to receive.
Understandably, not many of the less popular languages show up here, but surprisingly: two Arabic write-ins. Arabic isn’t a supported interface language on Steam, so is perhaps underrepresented in surveys. It’s also a bit harder to translate to due to glyphs changing form based on their position3.
Surprisingly, the breakdown here is also similar to the base breakdown. You’d think the conlangs would have a bit of an advantage (although the Klingon write-in lands here, of course).
The only charts where Spanish got (barely) beaten out. I suppose it makes sense that there would be a lot of German speakers who can kind of speak English but would prefer German. For similar reasons, it would probably be relatively easy to translate to German. Might have to look into it…
It’s surprising though, that other languages are very underrepresented here. It seems that receiving more representation was more important to those speakers than their personal comfort (or, possibly, many non-speakers cast their votes for languages they wanted to see get more representation).
Questions 3 & 4
“If you picked Español/Português, which major variant would you prefer?”
Finally, let’s look at how respondents’ preferences for Latin American vs. European regional variants break down. From these graphs, I’m excluding those who did not pick the corresponding language:
As you can see, LATAM variants got overwhelming support, and those votes are definitely the reason Spanish topped the poll (though even without that support, Spanish would probably have gotten a similar amount of votes as Korean).
Conclusion
Thanks again to everyone who responded! This has definitely given me a bit to think about, and will certainly inform future localization decisions. For those worried that a poor showing on this poll means that there’s no chance for your language, don’t be. There are many factors that go into deciding what languages to translate to, including cost, complexity, potential market size, and more. Polling Kickstarter backers is also fairly biased, as a vast majority of users on that platform come from English-speaking countries4, and this is something I’m definitely keeping in mind when interpreting these results.
As always, you can subscribe with just your email to receive more dev stuff like this. I’ve been putting my Chris Z. hat on and doing some financial data analysis in the past few days, and I might have some very interesting stuff to share here in the near future. See you soon!
Though no one wrote in “English”, which would’ve been objectively the funniest write-in.
Also because, in both cases, such colloquialisms reference Christianity and blood a lot, which would clash with parts of the worldbuilding.
Also because it’s highly stratified; Modern Standard Arabic can work for stuff like menus, but can sound a bit stilted for natural dialogue. Based on some quick searching, it seems an Egyptian-sounding dialect is often used for localizations like this.
Americans in particular make up over 50% of Kickstarter backers. And, in fact, Germany is the biggest non-English country on Kickstarter, which is probably why German was so well-represented in this poll. (Source: https://grumpyspider.com/where-do-backers-come-from)