How to Teach Kids to Code: The First Game and What Comes Next

It's a weekday evening in late September. Your son has spent an hour playing someone else's game, and the leaflet about after-school clubs has been on the fridge since the start of term. You'd like him to try making a game himself for once. So you search for how to teach kids to code, even though you've never written a line of code yourself. Two worries come up straight away. You can't teach something you can't do yourself. And if the first lessons are dry theory, he'll give up within a week. You don't need to become his teacher. In this article we explain how to start at home and how to tell when it's time to find a course.

Start small. One project your child wants to finish and one evening are enough to see whether they enjoy making games. At home, your job is to spark their interest and support it. Choose a project your child really cares about. A game in which their own hero runs away from a monster works better than any video about what an algorithm is. The hero can be the family dog: your child draws it in five minutes and then makes it run away from the vacuum cleaner. Children keep going because they want to see their project come to life. A nine-year-old won't sit through an explanation of what a variable is, and doesn't need to. Variables can wait until the game needs one.

One evening with a small game will tell you more about your child's interest than any club leaflet. A good first project is a cat that follows the mouse pointer. It takes four blocks and about half an hour. Open a new project in Scratch: the cat is already on the stage. Drag in the block "when green flag clicked", attach a "forever" loop and put two blocks inside it: "point towards mouse-pointer" and "move 10 steps". Click the green flag, and the cat follows the mouse around the screen. Then let your child play with the numbers. If they change 10 to 50, the cat overshoots the pointer and starts to shake. With 1, it crawls along very slowly. The next day, watch what your child does. If they ask how to make the cat catch something, or want to add a coin and keep score, the interest is there. If they don't go back to the game, that's an answer too, and it cost you just one evening.

While your child builds the game, you are its first player. You play and say out loud what doesn't work. Your child does the fixing, and you don't need to understand the blocks for that. Say your child calls you over to look at the game. You play for a minute, and the cat walks straight through a wall. It's tempting to take the mouse and fix it yourself, but it's better not to. Just describe what you saw: "The cat walked through the wall. Is it supposed to do that?" Your child goes back to the blocks and starts looking for a fix. Your second job is to keep the interest alive. Ask your child to show the game to the family at dinner or to grandma on a video call. This part stays yours even when a teacher takes over the explaining.

Problems tend to start with the second or third game. To build it, a child needs something new, such as a variable or a condition. Scratch itself doesn't explain these, and at home there's often nobody who can say what to do next. It usually happens in one of two ways. In the first, your son wants a proper game and finds a video tutorial. He follows it block by block, and the game works. But when you ask him to make the enemy slower, he doesn't know which block to change. He copied the steps without understanding why they work. In the second, a ten-year-old wants a platform game with three levels and a boss. After two evenings he's stuck: the hero jumps but doesn't fall back down. Neither of you knows how to fix it. He gets upset and doesn't touch the laptop for a week. It isn't the child's fault. The task is simply harder than it looks: to make the hero jump properly, a child needs several topics they haven't learnt yet. On a course these topics come one at a time and in the right order, so by the time a child reaches a platform game, they're ready for it.

When a child looks for a mistake, they go through the same steps again and again: make a plan, try it, see that it doesn't work, find the place where it broke and fix it. Watching and planning your own thinking like this is called metacognition. In 2019, Scherer, Siddiq and Sánchez Viveros published a meta-analysis of 105 studies in the Journal of Educational Psychology. It found that learning to code has a moderate effect on other thinking skills. Metacognition was among the skills that benefited most, together with creative thinking, mathematical skills, spatial skills and reasoning. General school achievement and literacy benefited least. In a later paper the same authors added an important point: these skills don't develop by themselves. The result depends on how teaching is organised, for example whether new ideas are explained clearly and whether the child has enough time for each one. Here's how that looks in a lesson. A nine-year-old girl's cat won't stop at the wall, and the teacher doesn't fix it for her. He asks: "At what moment does the cat start doing the wrong thing?" She finds the missing block herself. Then the teacher explains what she has just used: a condition. In the next lesson the condition comes back in a bigger game, so she uses it again straight away.

A course solves both problems that stall coding at home. Your child gets a programme in which each lesson builds on the one before, and a teacher who helps when something doesn't work. Our Scratch course is for children aged 8 to 10. It runs online and has 40 lessons. We've spent ten years developing and improving our teaching method, so you don't have to work out what your child should learn next. Children start with a short animation in which their hero moves, then make a first mini-game with obstacles, a game with several levels and a chatbot that greets you and answers questions. Each new topic appears in the project that needs it. The course is built like a game with levels, so you shouldn't have to persuade your child to sit down at the computer. When something doesn't work, for example the hero won't land after a jump, the teacher patiently explains the problem and finishes the project together with your child. Our teachers are selected in three stages and assessed every three months. If your child misses a lesson, the teacher gives them a one-to-one catch-up session at no extra cost. Lessons run once a week: 90 minutes in a group of up to 10 children or 60 minutes one-to-one. In a group, children present their projects to each other and work in pairs. They also build one big game together, with each child responsible for their own part. Three times a year you receive a written report: what your child has covered, what they're good at and what to work on next. You can also come to an open lesson, watch the class and talk to the teacher. At the end of the course your child has a portfolio of real projects and an international-style certificate in English. For children of 12 and over who find blocks too easy, there's Python Start (ages 12 to 14), with real typed code and 2D games in PyGame.

The easiest way to find out whether the course suits your child is one lesson. Book a free 60-minute trial lesson: it's online and one-to-one. Your child tries coding for real, and you get a learning plan made for them.