How Big Things Get Done

Flyvbjerg spent decades assembling a database of large projects and their outcomes, and with Gardner turns that evidence into a short set of rules. The central finding is that big projects fail because planning is rushed and delivery is slow, when the reliable pattern is the reverse. The book explains reference class forecasting, which anchors an estimate to the record of comparable projects rather than to the plan in front of you, and makes the case for modular design as the way to build large things predictably. Examples run from the Sydney Opera House and nuclear plants to film production and kitchen renovations. Useful for anyone who has to defend a number for work that has never been done here before.