Three Teams. Two Countries. Four Weeks. No Budget.

​Twelve young people in Tanzania and Pakistan spent four weeks this summer running a community experiment they chose themselves. No budget, no institutional permission, no payment in advance.

​Their work is marked on 18 September by external reviewers with licence to attack it, against a scoring guide published before any of them started. Half the marks go to how honestly a team handles its own evidence rather than to what the evidence says. I do not score my own programme.

​This is the session where all of it becomes public.

​What you will see. The films. The evidence, including the parts that did not work. All three teams in their own words. One team's headline measure got smaller every time their method got better: plus 1.14, then plus 0.43, then minus 0.17 once the largest single mover is removed. The review found the last of those. The team published it in full and kept the participants whose scores had gone down.

​What we are not claiming. Community impact. Two of the three teams took no baseline and cannot now produce one. What we can show is capability: twelve young people who got measurably better at handling evidence.

​Who it is for. Anyone who designs programmes, runs transformations, teaches, or signs off on evidence somebody else produced.

​One hour, online, recorded. Come with a hard question, it is the most useful thing you can bring.