Week 4 Field session: Counting servers

A walk around four queues in the same building to work out how many servers each one really has.

17 March 2027

Monday will have told you that pooling wins. This session is a walk, and the question we ask at every single stop is simply how many servers there are, which turns out to be a great deal harder than it sounds.

What happens#

We walk together to four queues on or near campus, which are the library desk, the campus pharmacy, the largest supermarket within ten minutes and the bus interchange. At each one you get ten minutes of looking followed by ten minutes of arguing about what you saw.

Three questions at every stop.

1. How many servers does this queue have. Not how many counters, how many servers. A counter with a trainee being supervised is a fraction of a server. A self-service machine that fails one time in eight is another fraction, and the fraction is not obvious.
2. Is it pooled, and if not, which of the five reasons applies.
3. If you pooled it tomorrow, what would break.

The supermarket#

We spend longest here because it is the best case against pooling and students usually arrive convinced it is simply badly designed. Look for the space, the impulse shelving, the express lane threshold, and the fact that the tills are not interchangeable.

Weather#

We walk in anything short of a storm. If it is cancelled you will know by 09:00.

Bring#

• Shoes you can walk forty minutes in.
• A notebook. No stopwatch needed today; this session is about looking.
• Your five reasons not to pool, from Monday, written down.

What leaves the room#

One page per queue with a server count you can defend, a pooled or not-pooled classification, and the reason. Two of these will contradict somebody else’s and that argument is where the marks in The Redesign come from.