What was tried, and what each one did

The ones that worked share a property

Each of the three successful changes moved an activity rather than a time.

Exercise went to a different room. The interruptible job got a second surface. The washes went to a different hour, which looks like a time change and is really a change of which room the noise lands in, because the kitchen at seven in the evening already has noise in it.

The two that failed both tried to move when something happens. Dinner at eight rather than seven moved the collision from work-against- cooking to company-against-evening. Alternating the room moved the distribution and cost fifteen minutes of equipment-shifting at each end.

Activities in a dwelling can be moved. Times mostly cannot, because the times are what makes them the activities they are.

The cheapest change was the best

Moving the exercise cost half a minute and a coffee table, and removed four collisions.

It was available because exercise was the only one of the kitchen's six uses with no reason to be there. It had gone to the kitchen because that is where the clear floor was, which was true until a table moved.

That is the general form of the cheapest available improvement: find the activity in the most contested room that has the weakest claim to be there, and check whether its reason still holds.

The most expensive change was the fairest

Alternating the quiet room was the only change that addressed the distribution rather than the total, and it was abandoned after three days.

It worked. On those three days the collision split moved toward even and so did the gains split. It cost about fifteen minutes at each end of each move, for two monitors, a keyboard, a chair and papers, to free a room on the one day in ten when the interruptible job had a call of its own.

The household's conclusion is in the record: a nine square metre room cannot be time-shared between two jobs, because what has to move is the equipment rather than the person.

What none of them touched

The twenty-two acoustic collisions.

Five of the six changes were aimed at the kitchen and the living room, and sound is not a property of a room. A call and a kitchen remain audible to each other under every arrangement tried, and the one measure that would help is a door the household is not permitted to replace.

So the fortnight's changes removed about a quarter of the collisions and left the half that come from the building exactly where they were.

The order they were tried in, and why it matters

The two changes that worked best were made on the same day, at the start of week two, immediately after the first week's log was read back.

That is not a coincidence and it is the main argument for logging over noticing. Both changes address problems the household had lived with for six months without identifying: the exercise had never felt like a conflict, and the chain from an eleven o'clock displacement to a four o'clock one is not visible from inside.

Neither required insight. Both required a list.

What was not tried and should have been

Reviewing the fortnight, two obvious candidates were never attempted.

A second surface in bedroom one. Discussed and dropped because of the socket. An extension lead costs four pounds. On the chains evidence it would have provided a fourth destination and converted two early-ending afternoons into two more displacements, which may or may not be an improvement — but it was never tested, and the reason it was never tested is that the socket objection was raised once and nobody revisited it.

Asking the landlord about the door. Addressed at length elsewhere.

Both are cheap, both were available throughout, and both were closed off by a single early remark that nobody went back to. That pattern — an option eliminated by an offhand objection and never reopened — is the same mechanism that allocated bedroom two in five minutes, and it is the commonest thing this record found.