A game that works in the room where it was invented has proven nothing. So we tested twice — first with the co-creators, then with teenagers who'd never heard of the project — before building the machine to deliver it.
Internal validation · 28–29 October 2025
Both original cohorts returned to play the first real prototypes. We measured two things: did the games teach the guidelines, and were they actually fun?
The nutrition content integrated cleanly — students engaged with the guidelines naturally during play, without it feeling like a lesson. The main thing being tested, and it cleared.
The mechanics were the problem. Students repeated their taste for competition and challenge, and were blunt: one game ran far too long, with unclear rules and confusing scoring.
Two honest caveats, logged rather than smoothed over: peer pressure visibly shaped the favourite-game vote, and the sessions ran after school, when students were already tired.
External validation · November 2025
We revised the rulebooks on the internal feedback, then a completely fresh cohort — no context, no memory of designing anything — played all three prototypes. The only way to tell whether rules are genuinely learnable rather than simply remembered.
Fast card play — grab cards and sort them into food groups at speed, so the pressure falls on recognition and reflex rather than deliberation.
The clear favourite of the blind cohort. Some food items were unfamiliar at first, but the pace kept energy high throughout.
Learning outcome: a clear grasp of the food groups, with a marked improvement in identifying grains and white roots.
Cleared for rollout
Builds the physical act of feeding another player into the board game itself — something every tester said they had never encountered in a board game before.
Rules felt confusing in the first minutes and became intuitive as soon as play started. Testers reported high replay value and said repeat plays would teach them more.
Learning outcome: hit its target exactly — students could state the recommended amounts and proportions from each food group.
Cleared for rollout
The most ambitious of the three. Players navigate the food environment and make choices based on where food is sourced.
Testers liked the idea and struggled with the execution: too long to play, and the core concept, rules and scoring were all hard to grasp. They asked explicitly for a clearer explanation of how to play and how to win.
Learning outcome: the message still landed — students could identify where to access healthier versus unhealthier food.
Sent back for simplificationAcross both cohorts, the FBDG messaging embedded successfully in every prototype. Prototype 3 consistently ran long and confused players on scoring.
Blind testers ranked Prototype 1 first on engagement, reported Prototype 2 intuitive after a slow start, and asked for a rewritten explanation of Prototype 3.
Take 1 and 2 forward. Simplify 3's rules, scoring and playtime, and give 2 a collaborative narrative wrapper — a school bell, a lunch break, a balanced meal to assemble together.
Strategic review
With both test rounds done, we checked readiness for the long intervention. Three problems surfaced — none about whether the games were good.
Before the study began, the team returned to School 7 to hand the finished games back to the students who invented them — showing exactly how their complaints about rules, scoring and length had been fixed.
Every co-creator's name is printed on the back of each rulebook. Physical sets went to the school and its student clubs — turning a research instrument into part of school life.
Phase 3 readiness · January – February 2026
Ethical clearance arrived on 22 January 2026 and unlocked the school-based study. Rather than wait for the research team's baseline assessments (the COM-B and DQQ instruments), the delivery team built capacity in parallel.
Mornings: play the games until the rules are instinctive — you can't run a room around a game you've only read about. Afternoons: the Ethiopian guidelines and the food groups in the deck, so a surprise question mid-session gets a real answer, not a deflection.
Candidates ran sessions while peers role-played the adolescents. Not everyone can hold a room of energetic teenagers — the exercises found that out before any student was involved. Four criteria decided it: patience and empathy under pressure, friendliness rather than authority, flawless rule comprehension, and quick fair judgement when an argument breaks out mid-game.
Twelve schools: six control, six intervention, balanced across three government and three private. We visited every intervention site before launch to agree schedules and venues with each school's focal person.
Every school returned the same answer: sessions could only run during the students' lunch break. Pulling students from class or keeping them after school was not feasible anywhere.
Lunch is the most valuable free time an adolescent has. Forcing a choice between eating with friends and playing an educational game is a serious drop-out risk across a months-long study.
Design the logistics around the meal, not against it. Games set up in advance, session times strictly respected, students able to eat before or while they play.
Before the first session, the team ran an onboarding day at each of the six schools. Deliberately not framed as a research study — being a research subject is an awkward thing to be — but as a long-running competition.
It answered the only question that matters to a fifteen-year-old giving up a lunch break: what's in it for me. The team credits it directly for the strong week-one attendance.
