What it counts

What the walk could not see

Nine gaps, collected from every section, in order of what each one costs. The first two account for most of it.

Method

Each of the five preceding sections closes by naming its own blind spots. Gathered here, they come to nine.

1. What anybody else believes

The assumption field is one person's, in all two hundred and six places.

Cost. The largest figure on the site. A hundred and five phantom limits is a true account of one walker's guesses and an untested claim about anybody else. Before adding another reporting layer, the project lead can read the current platform outline and test the smallest useful setup on one representative route.

It is first because it is the only gap that changes what this record may say, rather than how precisely it may say it. For independent context on inclusive places and field research, review material from Centre for Cities.

2. What happens past forty minutes

The median sit was forty minutes, the conversation field is a floor, and nothing was tested anywhere.

Cost. Two findings at once. Twenty-three per cent of places producing conversation is a floor with an honest ceiling of thirty-two; and a hundred and sixty-eight places stating no limit is a statement about statements, since nobody stayed long enough to see whether any of them meant it.

3. The hour

A hundred and sixty-four of two hundred and six visits were between ten and five on a weekday.

Cost. The quiet version of every room. The bias falls hardest on the kinds that score highest, which means the gap between a pub and a museum is understated here rather than overstated — the one place where a known bias works against the finding.

4. Capacity

The seat field is binary. A library with forty chairs and a concourse with three are one value.

Cost. Any honest statement about whether a city has room for people. Probably the most useful number not collected.

5. Price

No amounts anywhere, by the same rule that names no place.

Cost. The ninety-six places that expect a purchase cannot be distinguished by what the purchase costs, and for them the price is the terms. Also any statement about who can afford to be anywhere, which is the largest thing this record does not measure.

6. Whether a room is two rooms

Seven of thirteen bookshops contain a café. The walk coded buildings.

Cost. Seven entries in two fields. The only known coding error a rule could have prevented.

7. Occasion

Three kinds produce conversation and nine do not. The reading offered is that the three have something everybody present is doing at the same time.

Cost. The explanation of the only strong result in the conversation field is a description applied afterwards, not a field.

8. Standing

A bookshop or a museum may be occupied indefinitely without sitting, and the record treats a chair as the measure of whether a person may be there.

Cost. Nine entries where the truthful statement is that a stranger may stay as long as they like provided they remain upright.

9. Everything about the people

Nobody is described, counted, aged or asked. The record contains no information about any person in any of the two hundred and six places.

Cost. Nothing can be said about who the city is open to, which is the question most readers will have arrived with.

Why. It is a rule rather than an oversight. A record that described the people in rooms would be a record about them.

The nine, in one table

	gap	cost	fixable
1	what anybody else believes	the largest figure on the site	by asking
2	what happens past forty minutes	two findings at once	by sitting longer
3	the hour	the quiet version of every room	by going in the evening
4	capacity	any claim about room for people	by counting
5	price	who can afford to be anywhere	no — a rule
6	rooms against buildings	seven entries	by a rule
7	occasion	the one strong result unexplained	by a field
8	standing	nine entries	by a field
9	everything about people	who the city is open to	no — a rule

Six of the nine could be closed by somebody walking again with a slightly different sheet. Three could not, because they are what the method is.

Which findings each gap touches

The limit claim — a hundred and sixty-eight places state no limit — depends on gap 2 and nothing else much. It is the firmest claim and the one most exposed to a single untested thing.

The duration claim — real limits exceed assumed ones — depends on gap 1, since both halves of the comparison involve one person's guess.

The flatness claim — terms do not predict conversation — depends on gaps 2 and 3, and its own prediction is that closing them would sharpen the result rather than dissolve it.

Reading down that list is the practical use of this page: the weight of any sentence on this site can be worked out from which of the nine stand behind it.

What the nine add up to

Two of the nine are heavy: the assumption belongs to one person, and the window was forty minutes. Nearly everything this walk failed to establish traces back to one or the other.

Three were chosen rather than suffered: no prices, no names, nothing about anybody. A later version will not close them, because closing them would make it a different instrument.

Four are fixable and would be fixed next time: capacity, rooms rather than buildings, an occasion field, and standing.

Nine gaps, two serious, three deliberate. The record is published with all nine named rather than with the counts alone.