In this blog I'm going to talk more master game entrepreneur Luis von Ahn about his new game Duolingo, which he's using to help translate the web, and his top 10 lessons in game design.
Luis von Ahn is a master "micro-game" designer, whose last major success reCAPTCHA is now is digitizing 100 million words a day -- about two million books a year. His newest game, called Duolingo, is intended to help translate the World Wide Web.
"Duolingo really started when I asked the question of how can we translate the Web into every major language," Luis told me. "First of all, more than 50 percent of the Web is written in English and less than 50 percent of the world's population speaks English, so we'd like to really translate the Web into every major language," he said.
Luis knew that computers could do a bit of the translation work, but not all. "Sure, you can use computers for certain brief snippets," he told me, "but really they're not good enough yet to translate the Web. But if you want to translate the whole Web, you can't do it with 50 people or 100 people, so we started wondering how we might get, say, 100 million people, helping us translate the Web into every major language. That's kind of how Duolingo started."
"One challenge is the lack of bilinguals. I don't even know if there are 100 million people out there who are bilingual enough who use the Web to help me translate," said Luis. Another problem was this: "How are we going to motivate people to do this for free? I can't pay 100 million people."
Then Luis and his colleagues realized they could get work done by transforming translation into something that millions of people already want to do, learn another language. "There are about 1.2 billion people out there trying to learn a foreign language," Luis said. "Stuff needs to be translated, so why can't we get these people who are learning a foreign language to translate this stuff for us? That, in essence, is Duolingo, a language-learning website where, as people are learning, they're helping us to translate the web."
The gamification part of Duolingo comes from what Luis describes as a skill tree. "It really works out well," he said. "Basically the whole game is laid out as a tree and you unlock parts of the game." As a user completes a tree, he or she becomes more proficient in that language.
It's been super-successful, Luis said, with some 300,000 active users. "Of course, our goal is 100 million, but we're on it."
Next, Luis outlined for me his "Gamification 101" lecture, which he offers to entrepreneurs who want to utilize gamification to build their business or solve problems. Here are his top 10 pieces of advice:
- Gamification can solve a lot of problems with data interpretation. If you can figure out how to turn the problem into a game, you can have the crowd solve it, Luis said.
- Playing a game should be easy, but companies should realize that creating one is extremely difficult. "It's a lot harder than you would imagine," Luis said. "Simply slapping points in there doesn't usually work that well."
- Games themselves don't keep people engaged very long, unless they have a social component. Games "keep people engaged for the first week or the first month, but not for years," Luis said. "I think it's other people that keep people engaged for years. So, I think if you're thinking about gamifying, think about gamifying with other people."
- Games must be social. "The experience of the game has to be with other people and there has to be a way for the other people to be creative as well," Luis said. "What keeps people coming back to Facebook is that there are other people there who are being creative, who are expressing their thoughts, who are putting up pictures." (You can sign into Duolingo, for example, through Facebook, and keep your friends informed of your progress.) "That keeps people coming back for years," he said. "Things like Tetris don't keep people coming back for years. You really need to have other people involved to create loyalty to a game."
- The problem you're solving has to be really well integrated into the game mechanics. "It's deep integration to the game mechanics as opposed to shallow 'let's put some points and a leaderboard,'" Luis said. With Duolingo, learning the language is tied in to recognizing the digital version of a book in that language: you learn and contribute at the same time.
- Playing has to mean something to the player: A player can't win all the time, a player has to be able to fail to keep it interesting. "Basically, in all the most fun games in the world you can actually fail," Luis said. "It's a tricky thing, though, of how often you let people fail: How hard do you make the game? I think it has to be hard so that it's not super easy to win, but if you fail too much you get frustrated."
- Leaderboards are good, but only for the best players in your game. Be careful not to alienate entry-level players. "Leaderboards are awesome for the people near the top of the leaderboard and not for anybody else," Luis said. "Leaderboards tend to motivate the top players. The other 95 percent are not. If a leaderboard only helps motivate the top 5 percent, you've got to do something else," he said. One solution is to have regional leaderboards, so people don't feel overwhelmed.
- Be cautious when using monetary rewards or awards. "I've never seen the use of prizes," Luis said. "We tried prizes for a little while. It didn't do anything. I've never been able to make monetary awards work for me. Usually, it just makes people cheat at the game to get the monetary award. But I don't know if it makes it that much more fun. Gambling works for a reason so I assume it can be made to work. But we just never really made it work."
- Add random rewards, however. "Randomness is quite powerful," said Luis. "The fact that there is some chance that you will win gives you hope," he said. "That's why card games are so fun in some sense. You don't have to be the best poker player on earth, but you still have a chance of winning. In different games we have added things like a random bonus. It's completely random and now you get a bonus round or something and it really motivates people."
- Play the game yourself offline. It avoids mistakes later. "If you're going to make a computer game, the first thing you should do is play it without the computer. Somehow figure out a way to play the game just with pencil and paper," Luis said. "That really tells you a lot. About 90 percent of the things that I come up with are just not super but you can catch the majority of them early. So, if you're coming up with a game, you should try it on pencil and paper beforehand. You will spend two hours on that and this will save you eight weeks of coding later."
NOTE: Over the next year, I'm embarking on a BOLD mission -- to speak to top CEOs and entrepreneurs to find out their secrets to success. My last book Abundance, which hit No. 1 on Amazon, No. 2 on the New York Times and was at the top of Bill Gates' personal reading list, shows us the technologies that empower us to create a world of Abundance over the next 20 to 30 years. BOLD, my next book, will provide you with tools you can use to make your dreams come true and help you solve the world's grand challenges to create a world of Abundance. I'm going to write this book and share it with you every week through a series of blog posts. Each step of the way, I'll ask for your input and feedback. Top contributors will be credited within the book as a special "thank you," and all contributors will be recognized on the forthcoming BOLD book website. To ensure you never miss a message, sign up for my newsletter here.