# Chapter 25: separate solutions

## Question 1

Why not release steady by default?

It was not the development-selected candidate; choosing it after guard inspection contaminates the guard role.

## Question 2

What changed gain allows flashy to pass?

One repaired outcome over 4 pairs gives 0.25>=0.2.

## Question 3

Does contamination erase observed gain?

No; it invalidates the clean acceptance claim, while descriptive gain remains.