Confidence Intervals

On a bumpy bus through Croatia to the coast.

There I sat, bumping along, hot, uncomfortable, thinking about structural stability.  Coming to probable conclusions.  Windows are less stable than roofs.  Roofs are less stable than walls.  People are less stable than architecture.

Statistics requires a certain quorum of samples for confidence, but every third house was bombed out and unrebuilt.  Plenty of confidence.  Statistical confidence.  I´m told that there are other kinds of confidence, but these appear to be harder to achieve.