Half-finished games

The first ten minutes are the ones you cannot get back

A period is not 45 minutes of teaching. Take off settling in, devices out of the trolley, the two logins that will not work and a recap of last week, and you are teaching for about thirty. An empty project spends the first ten of those on questions that are not the learning objective: what kind of game this is, what it looks like, where to start. In a project week those are the point. In a lesson whose objective is “pupils can use a conditional to change what happens on screen”, they are overhead.

Watch a mixed-ability class of twenty-five open that empty project and it splits within three minutes. Three or four are away with something far too ambitious to finish. Roughly half make a competent, slow start. Then there is the group a teacher actually worries about, and they are not doing nothing: renaming the project, scrolling the character library, asking a neighbour what they are making. That looks like engagement and it is stalling, for a rational reason: nothing on screen to react to, no obviously correct first move, nothing yet to be wrong about.

So you spend the period going desk to desk restarting them one at a time, which is triage rather than teaching, and the class ends the lesson further apart than it started. The blank project did that.

Modification puts the difficulty in the right place

Give the same class a game that already runs and the lesson changes shape. Everyone has something on screen by minute five, and the first instruction can be “play it, then break it”, which nobody can fail to begin. The work is modification rather than creation, and modification has a property blank creation does not: before you can change something, you have to find it.

That is the assessment argument, and it holds with or without a pilot behind it. Ask a pupil to make the enemy move faster and they have to find the value that controls speed among several numbers, change it and check the result against their prediction. Ask them to end the game when the score reaches ten and they have to find where the score is counted and decide where a condition belongs. You see the understanding, or its absence, because they must point at something before they can touch it. A blank start is a noisier signal: a pupil with nothing on screen at minute twenty might not understand conditionals, or might simply not have chosen a game.

This is ordinary teaching, not a shortcut

Worked examples are old practice: a learner studies a complete solution before producing one of their own, because for a novice that is more efficient than floundering through the whole problem. Scaffolding is the same idea across time, support deliberately built in and, the part that gets skipped, deliberately taken away again. A half-finished game is a worked example you can run, take apart and put back together wrongly.

Plan the removal when you plan the support: scaffolding that never comes off is a ceiling. A sequence in which each template arrives with less already done in it, ending with a task that starts from nothing, gets a class somewhere the blank canvas never would. Two kinds of open-endedness are also worth separating. A task can be open in its outcome, so that twenty-five pupils hand in twenty-five different games, or open at its starting point, so that each of them has to invent a way in. The first is what you want; the second is what a blank project gives you, and inside one period most of the class ends up in the same unfinished place.

What the Spanish pilot did with this

The Andalusian pilot got there from the practical end. Its team built pre-coded templates for adventure, action, quiz and puzzle games, mechanics already written, linked to academic competences by subject and ability level, so that pupils could start from an almost finished game. The published evaluation reports that the templates saved more than 40% of coding time and made it easier for pupils to bring images and sensors into their games, and its authors credit them forCreate@School’s clear lead overPocket Code on practicality in teachers’ ratings. Their closing recommendation was to build more templates.

Take that for what it is: a finding about coding time and about teacher ratings in two Andalusian schools, gathered with acceptance surveys, not evidence that scaffolded tasks teach better. The same evaluation has pupils rating Create@School as neutral overall. What it settles is the practical half of our argument; how it was rated has the scores component by component.

• Template types Adventure, action, quiz, puzzle
• Reported saving More than 40% of coding time
• Where Two Andalusian schools; 115 pupils rated the experience
• What was measured Acceptance and user experience, not learning outcomes

The risk: a template handed back untouched

This is the honest objection, and the answer is yes, it happens: a pupil recolours the character, changes the title screen and hands in something that looks like finished work. A recoloured sprite photographs well and demonstrates nothing. Set the task badly and a template hides a pupil who has understood nothing, where a blank project at least fails visibly. The difference is made in the wording of the task and in thirty seconds per pupil at the end of the lesson.

• Specify the change, and specify one that cannot be reached from the surface. “It must end when the player has collected five coins” has a right answer you can check by playing it, and no amount of recolouring gets a pupil there.
• Break the template on purpose. One comparison the wrong way round means the game cannot come back working unless somebody found it.
• Have the pupil show you the block they changed and say what it does. That separates the ones who found the logic from the ones who changed things until the error stopped.

What we would still want to know

The 40% figure, the template types and the teachers’ ratings come from the published evaluation. Everything about the first ten minutes and about catching an untouched template is professional judgement with no study behind it. What nobody has tested, as far as we can tell, is whether pupils who work from templates all term can still start from nothing once the scaffolding comes off, which is the question that decides whether this is teaching or a trick. More on what survives a real period is on the classroom hub.