Candy Crush

What if we could make language learning as addictive as Candy Crush?


Candy Crush: a model for language learning?

On a two hour train journey at the weekend, I spotted seven people playing Candy Crush. One of them was me. I couldn’t put it down. Why is Candy Crush SO addictive? And how could we apply the qualities that make it so compulsive to language learning?

There’s a really nice article here, which talks about what it does to your brain, and how it gets you hooked. It’s worth a read, but the highlights are:

  • it’s very simple to complete the early levels, so you get a sense of satisfaction (and a hit of dopamine) really easily
  • as the game progresses, it gets harder to win and earn rewards, and the reward schedule becomes unpredictable. This is what makes fruit machines so much fun, too – knowing that it might just be you who hits that jackpot…
  • the game gives you an illusion of control. You can earn boosters as you go, and use them whenever you want, which gives you the feeling that you’re able to affect the outcome of the current game. This feeling of control is really important in keeping you engaged in the gaming process. Without it, too many failures could frustrate you into giving up
  • you are limited in how long you can play for – which means you can never overdo it. You only get 5 lives, and once they are gone, you have to wait for new lives to spawn, or buy more through in-app purchases. If you wait it out, playing the game again once you have new lives becomes even more rewarding
  • and finally – it’s candy. Chocolate! Sweet, shiny, sticky deliciousness! That’s no coincidence – the team at King are playing on our love of all things sweet to draw us in, for sure.

I find all of this fascinating, and as a player I can recognise these effects on me when I’m matching candies on my smartphone. I’ve caught myself thinking – just one more go! – and then been really disappointed to discover that I have no lives left, and need to wait 20+ minutes until I can play again. I haven’t stooped so low as to purchase more lives, but nonetheless, it’s bad.

Might it be good too, though? Surely, the more we understand about the way our brains work, the better we can apply this knowledge to arguably more useful things, like learning another language? Let’s take each of the factors that make this game so irritatingly unputdownable, and play with how they might be applied in language learning.


Make the early levels simple

This one has to be a no-brainer. Let’s say you’ve signed up to learn Turkish with me. You walk into the room and sit down. I smile at you, you smile at me. You’re excited, everything feels good. Then I launch into Turkish and talk at you for two minutes about everything we’re going to be learning together. At best, you’re going to dissolve into hysterical laughter. At worst, you’re going to be upset and completely demotivated. Of course, this is how total immersion works – the class is managed from start to finish in the target (‘foreign’) language – but the vocabulary used needs to be very carefully controlled, and the learners need to be brought along step by step.

In the context of language learning, I think simple in the early stages also means:

  1. not too much vocabulary. I love grammar, and through learning and teaching languages it’s clear (to me, at any rate) that once you have the basic structures in place, you can slot vocab in and out of the spaces for ever more. So, to start with, focus on some simple structures and the minimum vocabulary required to show how they work.
  2. nothing that sounds too similar to anything else. In Mandarin Chinese, for example, the word for ‘buy’ is ‘mǎi’, and the word for ‘sell’ is ‘mài’. The only difference between these two words is the tone (read my post here for more about tones in Mandarin), so to an English speaker’s ear, they sound almost identical. In his excellent book, The Learning Revolution, Dr. Jonathan Solity describes the learning methods of Michel Thomas (if you’re not familiar with his work, check it out; I’ll be doing a review soon). One of Thomas’ rules for learning is to separate similar items, because teaching very similar things together is a recipe for confusion every time.
  3. no explicit grammar – Yes! Really. Believe me, I LOVE grammar (I have an ‘apostrophe police‘ t-shirt), but we need to get way better at teaching it – or not teaching it. Watch for another post on this soon.

This list is just a starter for ten, but you get the idea. Keep it simple; avoid confusion and overload.


Make rewards unpredictable

This is an interesting one, and it highlights the difference between gaming for pure pleasure, and games for learning. There’s a lot of research out there showing that gamification (making learning into games) helps massively in creating and maintaining a motivating learning environment. This article from mashable explains how dopamine (remember?) aids cognition – in other words, learning through games is addictive and helps us to understand and remember better.

The key difference between a game like Candy Crush and a game for language learning, like Duolingo, for example, is that language learning requires input, feedback and correction. Crushing candy doesn’t. You either match the candies and crush ’em all, or you lose a life. Simples.

You can apply this idea of unpredictable rewards, for example, by varying difficulty in end-of-unit tests or knowledge checks; but this needs to be done with care. A learning tool has a duty of care to its users to treat them fairly, to encourage them constantly, and to allow them to feel successful often so that they remain motivated and feel a continuous sense of progression. A variable schedule of reward can work well though. The guys over at Real Deal Spanish do this, introducing new language at the start of the lesson in a chunk, leaving you thinking – What the heck? I’ll never get that! Then they break it down into bitesize pieces, and build it back up again, so that by the end of the lesson you are chatting away along with the native speaker. Another example: Pimsleur, in their Mandarin Chinese 1 audio lessons, teach new language, get you to practice it a while, and then just as you’re thinking you’ll never remember how to say this new phrase, they ask you to say something really simple from a previous lesson. The rush of relief when these easy words pop up is huge. It’s reassuring – it reminds you that you really HAVE learned something – and it takes the pressure off having to recall the new language in that moment. It also gives you a really rewarding sense of progression; a word or phrase which might previously have been new and apparently difficult to pronounce or remember, suddenly turns up like an old friend, all familiar and easy-going, and makes you feel warm and smug and I-can-speak-Chinese-after-all-glowy.


Give players/learners a sense of control

Candy Crush gives you control by letting you choose, at the start of each round, from a small selection of booster candies. These boosters are things like exploding spot-balls (gobstoppers?) and jelly fish, which pop the candies they land on. These generally have a random effect on the outcome (in my experience – although I am jaded) but they do help keep you motivated, because they let you think you can win more easily.

Our language learning parallels might be:

  1. Allowing learners to choose the order in which they learn different lessons (a ‘sandbox’ approach). There is a caveat here, because the syllabus would need to be organised so that it doesn’t matter which lessons you pick and in which order. I think a sandbox set-up is completely possible with topics like vocabulary building. Teaching grammar might pose more of a head-scratcher. It seems obvious that with grammatical lessons, there is a natural order in which to learn. For example, you wouldn’t try to teach someone the perfect tense in French before you’d taught them about pronouns and endings and the present tense (at least of avoir and être) too. In this article, Katharine Nielson, Chief Education Officer at Voxy, states categorically that “…autonomous language-learning doesn’t work and that people don’t know how to pick the right things.” Voxy check people’s proficiency and their reasons for learning, and then present what they think will be of most use.
  2. Giving people a ‘joker’ to allow them to move on to another lesson without quite knowing enough. I can imagine toes curling at this idea. But it’s not all bad. I will use myself as an example again. I’m a voracious learner and I want to feel that I am making progress every day, even on days when my brain is on under-drive and nothing seems to go in. So if I don’t get full marks on my test because I STILL can’t remember that ‘tak’ is Polish for ‘yes’ and not for ‘Thanks’, I might nonetheless want to move on to something new. Care is needed of course – some things build on others. But I am certain we could design a tool to take this into account and yet still allow people to move forward with small gaps in their knowledge. That’s life, right?


Give people limited lives

When you play CC, you start with 5 lives. When you fail at a level, you lose a life. Lives regenerate themselves, but it’s at the rate of about one per 20 minutes, so if you have a candy binge and lose all of your lives at once, you are required (unless you are prepared to buy more, which I’m not) to put it down and go and do something more interesting instead for a while. The clever thing about this is that you can’t overdose on the game and get bored with it – at least, not as quickly as you can with other games. How could we apply this idea to language learning?

  1. In real life, unless you are REALLY unlucky, you don’t lose a ‘life’ if you get something wrong. Either we manage to communicate our meaning, or we don’t. Hopefully, our interlocutor will support us in getting to our goal and will let us know if they need more help to understand us. (On a side note: if you’re not familiar with them, Grice’s maxims of conversation, or more broadly of communication, are really interesting. Grice explained the basics of how we naturally – almost subconsciously – cooperate with one another to reach mutual understanding when we talk to each other. This is a good place to start.) Still, a language-learning tool exists in a different space, and I think adding another dimension by giving people lives and creating jeopardy is fun. Most online tools and courses use scoring already to a degree – and it’s just one step on from there. What do you do with that score? Why not create some added pressure to get things right? And perhaps if I run out of lives, send me to the refresher page on the language point I’m trying to grasp, before I can go on?
  2. Specifically, the idea of limited access here – of exclusivity – is really useful. Effective language learning and language practice needs to be a sustained exercise. I’ve found building language learning into my day (for example, doing Polish in the car, or listening to Spanish radio while I bake) most effective, and it means I do it consistently. The neat idea here though is that when we limit things, we enjoy them more. We get hooked on them and we never completely satisfy the craving. So, bite-size learning here is key, I think. Little, interesting, engaging blobs of content, which give a sense of achievement and progression, and which always leave you fired up for the next bit. That is the way to success.


It’s candy

I don’t know of any language learning tools that use candy as part of the paradigm. But that’s not really the point. Candy Crush is GORGEOUS. It’s bright and colourful. It’s well-designed and beautifully animated. It’s simple to use, needs no user guides or help buttons and it’s a pleasure to play. There’s no reason why the tools we make for language learning shouldn’t aspire to these heady heights of design quality and addictiveness too.

I look forward to seeing what happens next. Sweet.