Inside the classroom, IF seems to help to stimulate memory, language work, discussion and co-operation within small groups. It also helps to foster a logical approach to problem solving, and is a good way of giving practice in basic IT keyboard skills as there is little reliance on the mouse. Within the UK National Curriculum for English, such games help to develop oracy (both speaking and listening with a partner or small group), spelling and reading comprehension skills. Games written for children in the 7-14 age range may also be helpful for IF beginners and useful in work with foreign English students.

Writing games for use in schools is, in many ways, the same as writing for an adult audience. You need to take the same care with spelling and grammar, and the game needs at least as good a range of verb/noun alternatives to avoid the 'guess the right word' problems.

There are also few things extra that you need to think about which are specific to the school situation. One, fairly obviously, is to avoid any type of bad language, racism or sexual references. Kissing Sleeping Beauty or a frog is just about okay, but nothing more. It's one thing for a child to come across such things for themselves in everyday life, quite another if a school gives it to them. You also need to be very careful with any religious or cultural messages and references in the game, both explicit and implicit. These can deeply upset some children and parents, though it's fine for a game to have a religious theme if that's made clear.

When writing a game for a school, remember not to make it too long. Ideally, the game should be finishable by most children within one afternoon (or two at the most). Most schools in the UK now have an ICT suite, but time here is often very limited and the game and its interpreter will need to have been installed on every machine. Typically, a game will be used on the one or two computers within the classroom of an interested teacher, and it takes a long time to work through the whole class playing the game in groups of two, three or four.

Games also need very good documentation, so the teacher can find what help to give where without having to find the time to sit down and work through the game themselves. Given the pressure on teacher's time, this will almost never happen unless he/she has a particular interest in IF and computers. The documentation should include a room-by-room list of puzzles and objects as well as a full walkthrough, and a schematic game map is also useful.

To reduce teacher involvement (remember, the other 26 or so in the class need to be taught at the same time) it's a good idea to build in some sort of hint system to cover any puzzle harder than examining something to find an item. Also vital (if you're writing your own parser) is a save/restore function. Make sure any undo functions are enabled and mention their existence somewhere obvious in the game. Few children remember to save regularly unless the teacher constantly reminds them. It's also a good idea to implement a 'resurrection' function if the player can be killed ('undo' may not always be enough to recover from a lost game situation). There should also be some way of rescuing vital items if they can be eaten, given away to dwarves or dropped into chasms. Many children seem to automatically eat any food items found in the game regardless, and having to restart leads to much disappointment. Try to avoid making the game unwinnable by any player actions - for instance, if a player steps into the dark, he/she should be able to recover by moving straight back in the reverse direction. Make the game as kind and forgiving of player mistakes as possible.

Children love pictures in games, and it makes the game much more accessible. It seems to encourage children to read the text, rather than the reverse. If you do produce a game with pictures (using Inform Glulx or TADS 3 or some such), remember that children will always go by the picture rather than the text. If the picture shows the dwarf with a red coat and the text says he has a green coat, the green coat will be 'wrong'. Don't put subtle clues into pictures - they will almost certainly be missed. You also need to do quite a bit of work with the scenery - children will often try to examine/take things they see in the picture rather than the text, and will complain bitterly about the standard Inform 'you don't see that here' response if something's visible in the picture. You don't need to put everything mentioned in the text into the picture - sometimes things are best left to the imagination (especially if you're not a very gifted artist!). Some NPCs and creatures such as fearsome dragons fall into this category. (Sometimes children ask why something mentioned in the text isn't in the picture - they always seem quite happy when I say "Because it's behind you.") The reward of finding new pictures or sounds seems to be a great motivation for making progress in the game. This also provides motivation for taking shortcuts, if they are at all possible. It's a good idea to disable 'take all', otherwise a few children can discover this and type it automatically for every location visited, without bothering to read the text. Similarly, carefully check any rescue, resurrect or teleport routine doesn't give the chance to shortcut any parts of the game (eg. being killed in the dark will then resurrect you with a source of light).

Children are terrible at making maps as they go through the game (unless they have an IF-loving parent who's taught them) - if reminded to do this by the teacher at the start of the game, they may manage to plot a couple of rooms or so before forgetting. I've never known a child to manage more than this in class. If you can, consider implementing an on-screen map that can be called up as needed. This can be arranged to show just the rooms visited, so as not to give away much of the game. Failing that, think about including a black-and-white A4 GIF/JPG sketch map of the playing area that can be printed out and photocopied by the teacher (as children always seem to write or draw on them). You need something that stops the teacher being asked questions like 'how do you get from the cave to the haystack?' every few minutes.

Regarding the game itself, avoid mazes at all costs. Children have no idea how to deal with them, and will move at random for a minute or two before giving up and asking if they can do something else. Also, try to avoid move counters leading to things like lamps running out - they can cause tremendous frustration. Try to avoid illogical puzzles - children love fantasy, but expect everything in it to make clear sense and follow rules. Most have come across the more well-known fairytales such as Cinderella and The Three Bears, and will enjoy recognizing puzzles based around elements of known stories and tales. Don't worry about including versions of old classic adventure game puzzles - the children will almost certainly be meeting them for the first time.

You need to think about the vocabulary and the level of reading required in the game. You need to make the text reasonably straightforward, but not too simplified unless the game is specifically aimed at the poorer readers. Don't completely avoid more unusual words - children are used to dealing with new words all the time in their everyday reading. Just don't give them too many all at once, and don't over-complicate the writing. Try to make sure the meaning of anything uncommon can be gathered easily from the context. There's a required basic reading competency for the simplest of games that gives a cut-off point somewhere around 6-7 years - much below that and only a few children will have the reading and writing skills needed to progress in a game, and some won't be able to relate to a game model at all.

One of the hardest things to do is to match the interest content of the game to the age range you're aiming at. You need to avoid making the children feel patronized by over-simplification, over-sentimentality or cutesiness. By the age of 6-7 (around the lowest age you can write for), children are becoming increasingly independent and daring - games based around the safe world of the nursery and cuddly toys (that might be fine for 4 and 5 year-olds) often don't go down very well. To get a feel for the sort of subjects that work well with a particular age group (and the text level that's appropriate), it's a good idea to try reading a few books popular with that age group by good children's authors, such as Roald Dahl.

If possible, try to have a reasonably-sized landscape the children can move in before they have to solve the first puzzle - this gives them a chance to get a feel for the game and how it operates. As with adult games, try to make the first puzzle or two fairly easy to draw players into the game.

Most children seem to be very sensitive about the gender of the player character - girls don't really like playing as a male character (particularly if they have to kiss Sleeping Beauty to wake her), and boys especially hate playing as a female character (the Lara Croft effect doesn't seem to work here). (It's interesting to get a class of 10-year olds to write a story where the girls write as a boy character and vice versa - it reveals some strange preconceptions.) If possible try to use wording so the player character can be either male or female, unless this is a particular feature of the game. Try to be aware of gender stereotyping - avoid things like ordering female NPCs to bake a cake or sweep the floor.

It's usually a good idea to avoid violence in the game. Many parents and teachers hold strong anti-violence views, and we want to discourage the idea that violence is a useful answer to solving real-life problems. However, many children themselves have no qualms about violence directed against evil ogres and dragons in a storybook context - many of the traditional fairytales include it (Jack in the Beanstalk, Little Red Riding Hood), and children recognize this clearly as fantasy at this age. Some children, though, will feel uncomfortable having to fight or kill giants and dragons in an IF game, and overall it should be possible to complete a game in a non-violent way if the player so chooses. Random acts of violence and destruction in the game should be discouraged (by the death of the player character, loss of points or some such - at least by a pithy game response).

References to the supernatural should also be considered. Mention of storybook-type spells, magic and witches seems generally acceptable in most schools (although a few fundamentalist parents, teachers and schools may object), but you certainly don't want to give any children nightmares by scaring them. Avoid anything genuinely creepy, or anything that could be classed as horror - zombies, corpses, ghosts and the like, that some parents might object to.

Mostly, children in UK schools seem unbothered by US spellings once this has been pointed out to them, and the same probably applies in reverse. It may be a good idea to include the alternative spellings in any noun or verb synonyms, and both British and US noun alternatives where they exist (such as pavement/sidewalk) if you think your game might be used in other countries. If it's really an issue, consider compiling British and US versions of the game.

Try and make the game good fun to play. When testing the game, if possible try to watch a group of children playing through it and listen to their conversation. See which parts (if any!) they enjoy. Watch what they try to do, and if possible build appropriate responses into the game. Children enjoy a good humorous response as much as adults.

Good IF games used in the classroom seem to catch the imagination of a lot of children, many of whom enjoy the relaxed pace, intellectual challenge and collaboration with their friends. They often want to use the same IF game at home and ask where they can get it from. Several interested teachers can introduce IF to hundreds, if not thousands of children (and often their families) over a few years. Writing a game for children is certainly worth a try.

William Stott, April 2004   (wstott@bigfoot.com)