---
id: 2026-01-25T22:59:00-0500
title: 2026-01-25 22:59:??
tags:
  - topic/estimating
daily: "[[2026-01-25]]"
---
# 2026-01-25 22:59:??

## _Judgmental Decomposition_

[[macgregor_1994_judgmental-decomposition]]

I really hate this study.

I may be out of my league,
but it seems wrong to draw conclusions about decomposition as a method
when the work was done by the researchers,
especially when some of their decompositions really suck.

> ##### Circumference of 50¢ coin
>
> * Diameter in inches of a 50¢ coin
> * Number of pieces of string the length of the diameter needed to wrap around circumference

That's not a decomposition.
If you don't remember $\pi$
that's a harder problem than it was before.

> ##### Bushels of wheat
>
> * Population of the world
> * Number of bushels of wheat consumed per person per year
> * Proportion of wheat wasted per year

I might have included _how much a bushel is_,[^1]
there's an order of magnitude right there.

[^1]: $1~\text{US bushel} \equiv 9\frac{3571}{11550}~\text{US Gallons} \approx 9.3~\text{US Gallons}$

The study suggests decomposition has no effect on estimate _confidence_
---which I'm tempted to believe because its funny
and it tracks with my anecdotal experience---
but I wonder if subjects using their own decompositions would present similarly.