What two hundred and six places amount to

One city. Eleven weeks. Two hundred and six places where a stranger may sit. Four fields each, no place named, no price recorded.

The counts

places walked                           212
excluded, having business there          -6
recorded                                206

at the door
  nothing to pay                         71    34%
  one purchase expected                  96    47%
  a purchase roughly every hour          24    12%
  a fee, ticket or membership            15     7%

how long you may stay
  no limit of any kind                  168    82%
  a limit, written                       21    10%
  a limit, said when asked               17     8%

the guess, made at the door
  correctly expected no limit            63    31%
  expected a limit; there was none      105    51%
  missed a limit that was there          15     7%
  correctly expected a limit             23    11%

somewhere to sit without buying
  yes                                    89    43%
  no                                    117    57%

did anybody speak to a stranger
  yes                                    47    23%
  no                                    141    68%
  could not be seen                      18     9%

The two figures that work against each other

A hundred and sixty-eight places set no limit, and seventy-one ask nothing at the door. The rationing of public space that the subject is usually about is not in this record. Four rooms in five let you stay as long as you like.

In a hundred and forty-one places nobody spoke to a stranger. By the test the term third place sets for itself, most of what was walked is not one. They are rooms with chairs. The workflow described here can be compared with workforce time records for multi-site teams, provided the team states what the information can and cannot demonstrate.

Both are on the front page. The record is cheerful about permission and bleak about company, and neither half is a footnote to the other. Additional practical or institutional guidance relevant to the survey can be found at Urban Institute.

The three claims

This city sets almost no limits on sitting. Eighty-two per cent state none, and the twenty-one that print one give a median of two hours.

Where a limit exists, it is longer than it is assumed to be. Thirty-one of the thirty-eight real limits exceed the guess made about that place before checking. Assumed median an hour; written median two.

Nothing about a place's terms predicts whether anybody speaks in it. Eight comparisons, seven of them between nineteen and twenty-five per cent. What predicts it is the kind of room: fifty-five per cent in pubs, fifty-seven in places of worship, twelve across the other nine kinds together.

The five refusals

Not that the city is generous. Nothing was tested and nobody stayed long enough to find out.

Not that it is mean. Eighty-two per cent limit nobody.

Not that free seating produces company. Twenty-five per cent against twenty-one, across eighty-nine places and a hundred and seventeen.

Not a rate for anywhere else. One city, one season, one observer, and fifty-eight cafés that are not independent of each other.

Not that half the city imagines a rule. The hundred and five belong to one walker's guesses. Nobody else was asked.

In one sentence

Almost every room in this city will let a stranger stay as long as they like, almost none of them says so, half of them are assumed to have a rule they do not have — and in two rooms out of three, nobody says anything to anybody.

The record by kind

Three rows carry most of what the walk found.

Hotel lobbies: ten places, no limits, nine phantom rules, no conversation. The purest case in the record of a room that restricts nobody, is believed to restrict everybody, and in which nothing is ever said.

Places of worship: fourteen, free, seated, unlimited, and the highest conversation rate anywhere. Every condition met and the outcome present — the only kind where both are true.

Cafés: fifty-eight, the commonest room in the city, third from the bottom on free chairs and sixth from the bottom on conversation. The room everybody means, failing the test the category sets.

The three figures to carry away

A hundred and sixty-eight of two hundred and six set no limit, and almost none of them says so. Four rooms in five, silent about it.

A hundred and five limits were assumed where none existed. One walker's guesses, recorded before checking, wrong five times in six whenever a restriction was expected.

A hundred and forty-one places where nobody spoke to a stranger. A floor, not a count, and the figure the record most wants to be wrong.

Where the gaps are

In the instrument, not in the arithmetic. Forty minutes a room, one pair of eyes, mostly weekday afternoons, and an assumption field that belongs to one person.

Six questions would be settled by walking it again with a slightly different sheet, and none of them needs a price, a name or an opinion about where anybody should go.

One walk. Two hundred and six rooms. Three claims, six refusals, nine blind spots, and no address anywhere.