Almost every game they invented was competitive, physical and fast. Even the seated games demanded intense concentration. Anything with long or complicated rules was abandoned within minutes.
Round one · 17–20 September
Four days with 21 strangers
Round one wasn't about building a nutrition game. It was about finding what these teenagers think is fun — so we could hide the nutrition inside it later.
Day 01 · The theory of fun
Finding the gamer persona
Objective
Identify the gamer persona of Ethiopian adolescents, and define what "fun" actually means to them.
We opened with an icebreaker, not a lecture. Name, grade, age — then: why are you here? Their answers set expectations, and signalled we weren't teachers.
Team formation was physical: draw an animal at random, then find your teammates using only its sound or gesture. It broke up friend clusters and let us read each student's energy before a single design task began.
Then we asked what games they loved. Groups listed everything they had ever enjoyed, narrowed to a top five, and the whole group voted.
Most played favourite games — cohort vote
Football13
ቅልብልቦሽ / ወደም13
Tic-tac-toe12
Bingo11
ብይ (marbles)10
Our football is different from the others because it isn't just 22 people fighting for one ball; it is an experience that actively engages the audience.Participant, Day 1
That single reframing was worth more than the vote counts. For this demographic, the crowd around a game is part of the game.
Next: cardboard, cups, rope, chalk, Jenga blocks and table-tennis balls — thirty minutes to build something playable. The same familiar materials used in professional prototyping.
Evidence chain 01Their native gaming language
Observed
Evidence
Five student-built prototypes, all "I win, you lose". Consistent drop-off in engagement whenever a game took more than a sentence to explain.
Design decision
Gursha cannot be slow, quiet or rule-heavy. Friendly competition, simple rules, and something for the hands or the mind to be doing at all times.
In the afternoon, five commercial games — Jenga, UNO, Spot It!, Dominoes and the Ethiopian word game ምን ልበል — each teaching a different mechanic: tension, speed, luck, verbal pressure. Every team played all five, then voted.
Commercial game rotation — fun rating vote
Jenga18
UNO10
Spot It!10
ምን ልበል8
Dominoeslast
Evidence chain 02Tension beats strategy
Observed
Jenga won by a landslide. UNO and Spot It! tied behind it. Dominoes — abstract, slow, strategic — came last.
Evidence
18 votes against 10, 10 and 8. Physical consequence and speed outranked planning in every group.
Design decision
Build around tension and speed, not turn-based strategy. A quiet board game would have lost this cohort by round two.
I understand that we can create games.Day 1 reflection circle
Day 02 · Games for impact
From fun to purpose
Objective
Connect games to nutrition, culture and behaviour — without losing the previous day's energy.
Instead of recapping Day 1, we asked what had survived the night. Only the active experiences — though a few had kept the vocabulary (challenge, reward), a sign some wanted the theory too.
We taught mechanics through football — the shared language. Ball, dribbling, no hands. Then: delete the no-hands rule? It becomes running, and boring. Add a second ball? Chaos.
Then we mapped the food environment. Each student drew one food onto a wall-sized buffet — shiro and tibs beside pizza, burgers and cake. What do you eat in a day? And what would you eat if money were no object?
Evidence chain 03The awareness gap
Observed
Asked to vote healthy or unhealthy on their own food lists, students judged almost entirely by taste. Foods they liked they defended; foods they disliked they called unhealthy.
Evidence
Whole-class voting session. "Healthy" functioned as a synonym for desirable, not as a biological property.
Design decision
Teach food groups and proportions as concrete categories rather than moralising about good and bad food — which this group will simply argue with.
After lunch, nutritionist Dawit Alemayehu delivered the guidelines as a lecture. The framing was smart — a balanced diet is shiro, injera and a vegetable, food already on their tables, not expensive imports. The framing landed. The format did not.
Those two games were the missing link. The students had fun (Day 1) and facts (Day 2), but no model for combining them. After Weze, they saw that an educational game doesn't have to be a quiz.
Evidence chain 04The lecture dip
Observed
When the day shifted to lecture format, the room went quiet and visibly tired — the only sustained drop in engagement observed across the four days.
Evidence
Confirmed independently by the research partners in the debrief. Messages were also distorted in transmission: several students left believing more than one glass of milk is unhealthy.
Design decision
Remove the lecture entirely in Round 2 and replace it with a competitive format. Back it with visual handouts.
Day 03 · Ideation and game making
Forty ideas, then four
Objective
Fuse game design principles and nutrition knowledge into something playable by the end of the day.
We asked each team for five nutrition game ideas in fifteen minutes. The room froze — they were trying so hard to be correct that most couldn't reach two.
So we flipped it. Here's an overworked single mother, guilty about her kids' screen time — how could you make her life absolutely miserable? The ideas came instantly, because ruining something is fun and fixing it is frightening.
The time pressure was the point. With no time to judge, the later ideas came out more original — the obvious options were already spent.
Stress makes you creative.Participant, on the Crazy 8s sprint
Each team came back from lunch with around forty raw ideas. Then combine and compete: same problem? pit them and keep the stronger. Both good but different? merge. Forty to eight to four, each round on a strict timer, each defended to the room.
Evidence chain 05They compete, they don't combine
Observed
We expected teams to preserve everyone's ideas by merging them. Instead they were ruthless — boring or confusing ideas were discarded outright rather than repaired.
Evidence
Both filtering rounds ran fast and decisively. Ideas were defended as "funny" or "different"; games that were fun but confusing lost to games that were simple and active.
Design decision
Design for a cohort with high standards and no patience for weak mechanics. Humour and physical action are the currency; complexity is not.
Twenty minutes of prototyping followed. Scoring systems that looked elegant on flip chart paper turned out untrackable in play, and teams simplified them on the spot. Then a blind round-robin: one player from each team tried the other stations while the creators watched in silence.
Evidence chain 06Three enemies of the game
Observed
Peer feedback was blunt: symbols were confusing, some players held a role that gave them nothing to do, and decks were stuffed with unhealthy foods.
Evidence
"The rules were a bit confusing" was the most common comment. Runs of consecutive penalty cards stalled play entirely.
Design decision
Name the three failure modes — confusion, stagnation, passive roles — and design against each: standard iconography, a deck weighted toward positive nutrition, and an action attached to every role.
Day 04 · Iterations
The regrouping experiment
Objective
Synthesise the strongest mechanics into a minimum viable product, and validate it through playtesting.
Overnight we dissolved every team and rebuilt four groups strictly by participation level — the most active students together, and so on. The loudest could no longer dominate the quiet ones, and the quiet ones had nobody to hide behind.
Evidence chain 07Innovation needs a decision-maker
Observed
Three of the four restructured groups produced a playable game. The fourth produced nothing — it spent the entire session in a power struggle.
Evidence
Groups with a clear informal hierarchy moved fastest, though they showed ownership bias: they polished the dominant member's Day 3 game rather than building a genuine hybrid.
Design decision
Future club sessions need assigned roles. Too many equal strong voices produces gridlock, not innovation.
The final games made this cohort's preferences unmistakable: party games, not strategy. Laughter and quick turns over deep planning; physical agility and luck over arithmetic and resource management.
The reliance on luck is worth sitting with. In a strategy game the cleverest student wins every time; in a game of luck or speed, anyone at the table can win — and for this group, that is fairness.
- PhysicalityThe final game must include a physical action — which is where the gursha feeding mechanic comes from.
- SimplicityNo heavy text, no arithmetic. If it needs a paragraph to explain, it is already lost.
- Social riskReal stakes in every round, because they told us repeatedly that a game without risk is boring.
