Case 2 · inference from limited data · Core Maths
Who's actually playing?
1Ask
Is MaffsGames used in classrooms, or at home? It matters: a teacher needs games that work on a projector in 15 minutes; a student at home needs ones that explain themselves. We never know who you are, so we have to work it out from clues.
2Collect
We record what happened, never who: which game, right or wrong, and when. Google Analytics adds roughly which country. No names, no accounts, no school. That's all the evidence we have.
3Picture it
Games started by hour, UK time, last 4 weeks
Games started each day, last 30 days
4You decide
What does the evidence say? Before you answer, think about:
- If this were classrooms, what would the weekday hours look like? What would half-term do to it?
- What else would produce the same pattern? (Homework set in class. A teacher testing games at lunch. Us.)
- One class of 30 can start 30 games in a minute. One student can start 30 games in an evening. Can our data tell those apart?
- What one extra piece of data would settle it, and could we collect it without knowing who you are?
What this data can't tell us
- Starts aren't people.
- Countries come from Google and are rough.
- Our own testing is in the numbers.
- School holidays differ between England, Wales, Scotland and Northern Ireland.
- Figures below 10 are shown as <10: too small to mean anything.