Updated 25 July 2026

What Changed

Planning Rice Run 5 from a nutrition target instead of a shopping list told us what Rice Run 4 had actually delivered. It was less than we published.

The Number

Three Figures, One Problem

76

children, counted by name

Roster sheets photographed at both houses, with ages.

1,676

kcal per child, per day

What a child here needs, weighted by age and sex. Thai DRI 2563.

952

kcal per child, per day

What Rice Run 5 delivers — 57% of the plate. The houses cover the rest.

For four runs we planned the same way. Take a headcount, work out what the trucks can carry, fill them, deliver. It gets food to children and it has never once failed to. What it does not do is tell you how the load compares to what those particular children actually need.

We are not the only thing on the plate — there is school lunch, greens from the garden, an egg or a piece of fish. But we had never once written down what share of the plate we were. For Rice Run 5 we did, and it changed what goes on the truck: more rice, six times the cooking oil, and mung beans for the first time, for protein that rice alone cannot supply.

How We Missed It

We planned from a headcount and a shopping list, never from a nutrition target, so nobody ever checked. You buy as much rice as the money buys, you photograph the receipt, you drive it there. Every number we published was true about the money. None of them were about the food.

The meal count is the clearest example. We divided rice by 0.3kg and called each slice a meal. 0.3kg of dry rice is around 1,080 kcal — closer to three days of what a child on that run actually received than to one meal. So the Rice Run 4 recap claimed 5,000 meals on the same page that said 80 children for 180 days, which is 14,400 child-days. Both numbers were published. They are out by a factor of about 2.9. Nobody caught it, us included, for five months.

The tonnage went the same way. 1,500kg was the number on the social post and it stuck. The receipt says ฿31,986 of rice at ฿342 for a 15kg bag, which is 93.5 bags, which is 1,403kg.

Rice Run 5

Four Things We Do Differently

01

A roster, not a headcount

We used to plan for "about 90 kids". Both houses wrote out their sheets: 76 children, with ages. Bilay recorded theirs; Samson’s are inferred from school grade, which probably runs young, so the real requirement may be a little higher than we model.

02

A calorie target, not a shopping list

Every previous run started with a cart and ended with a receipt. This one starts with 1,676 kcal, 40g of protein and 52g of fat per child per day, and sizes the order to close the gap that school lunch, the garden and an egg or a fish don’t already cover.

03

Standard rice, not premium

Same grain, same 15kg bag, and the kids can’t tell. The ฿31,986 we spent on premium rice in February would have bought about 480kg more as standard — nutritionally identical. From Rice Run 5 we buy standard and put the difference into food.

04

Mung beans, for protein

Rice protein is incomplete on its own; it is short on lysine. Beans supply exactly that. 243kg of mung beans go on the truck this run. Neither kitchen has cooked them at this volume before, so this one we watch closely.

The Corrections

What We Restated

Every figure below was live on this site. The old ones are still here so you can check our working.

ClaimAs publishedRestatedWhy
Rice delivered, Rice Run 41,500 kg1,403 kg฿31,986 of rice at ฿342 a 15kg bag is 93.5 bags.
Meals fed, Rice Run 4~5,000 mealsWithdrawnWe counted a meal as 0.3kg of dry rice — about 1,080 kcal, or three days of what a child actually got. The same page said 80 children × 180 days = 14,400 child-days.
Rice delivered, Rice Run 31,167 kg1,200 kg80 bags of 15kg on the Makro receipts. The old figure was back-derived from the meal count.
Rice, all four runs3,796 kg3,803 kgRe-derived run by run from the receipts: 450 + 750 + 1,200 + 1,403.
Cost per unit€0.22 per meal€0.12 · $0.14 · ฿4.53 per child, per day฿65,239 spent, 80 children, 180 days. A day is a real unit; that meal wasn’t.
Reserves8.6 months~4.5 monthsPriced at Rice Run 5’s ration rather than Rice Run 4’s thinner one.
Children supported9076Counted off the house rosters instead of estimated.
Food share of spend, Rice Run 449%54.4%49% is rice alone. On the basket we actually measure — rice, oil, protein — it was 54.4%. We had been quoting the smaller number against the bigger target.

The Rice Run 4 recap keeps its photo captions as shot — the collage really does say 1,500KG, and painting over a photograph is not a correction. Read the recap.

What Didn't Change

€2,457 · $2,902 · ฿92,133 raised. 30 donors. 7 countries. 80 children. Every receipt still photographed, every purchase still published. The money went where we said it went — we were just measuring the food with the wrong ruler.