# Arrival-Wave Room-Blocking Methodology

**Version:** 2.0
**Date:** 2026-08-20
**Supersedes:** v1.0 (2026-06-15), which described the method with no day attached.
**Type:** Method, with the arrival day it was built for

---

## Why this version has a date on it

Version 1.0 described three arrival waves and told you to clean in wave order.
That is correct and it is what every operations manual says. What it could not
tell you is what a bad arrival day actually looks like, which is the only thing
that makes the sequencing matter.

This version is built on one real day at **Westmark Fairbanks**, a 400-room
property with 356 rooms sellable on floors 2–8. The day is Friday 17 July 2026.
No guest information appears here; the pre-arrival sheet this is drawn from was
built so that no guest name is ever stored or displayed, and that constraint is
part of the method rather than a redaction applied afterwards.

---

## 1. The day the method is for

| | |
|---|---:|
| Parties arriving | **285** |
| Guests | **564** |
| Rooms needed, one per party | **285** of 356 sellable |
| Headroom | **71 rooms** |
| Two-guest parties | **244** |
| Parties of three or more | 15 |

Two numbers on that table are the whole problem.

**285 of 356.** The property is 80% committed before a single walk-in, and every
one of those rooms has to be clean and correct on the same afternoon.

**180 guests arrive in a single minute, at 17:45.** That is the number that breaks
the standard advice. Coaches do not trickle. When a tour lands, the desk does not
face a queue that forms over an hour — it faces most of an evening's arrivals
standing in the lobby at once. A room that becomes ready at 18:00 is late for a
guest who arrived at 17:45, no matter how good the day's overall completion rate
looks on the board.

**This is why wave order beats room order.** Cleaning the floors in numerical
sequence produces the same number of clean rooms by end of shift and the wrong
ones by 17:45.

## 2. The three waves

- **Wave 1 — guaranteed-early.** Guests who requested or paid for early check-in,
  plus segments that historically show up early.
- **Wave 2 — the standard afternoon.** The bulk of an ordinary day.
- **Wave 3 — the coach drop.** Not "late arrivals" in the usual sense. A single
  timestamped event with a known passenger count, hours of notice, and no
  variance to speak of.

On a property that takes tour business, wave 3 is not the tail of the day. It is
the largest single block on it, and it is the most predictable thing on the sheet
— which makes it the easiest to plan for and the most expensive to be wrong about.

## 3. The bed decision, which is upstream of blocking

Blocking by room *type* has to be settled before any specific unit is assigned.
On this day the type question is sharp, because 244 of 285 parties are two guests
and the property census is **218 King · 181 Queen · 1 Full**.

Two guests can mean one bed or two, and the sheet does not know which:

- If every two-guest party is a couple → **244 Kings needed, short 26.**
- If every two-guest party is two people sharing → **244 two-bed rooms needed,
  short 63.**

Neither extreme happens, but the exposure is not symmetric, and that asymmetry
decides the rule:

> **Default to King when unsure. It is the cheaper error.**

Being short a King means some couples get a two-bed room, which is a mild
disappointment. Being short a two-bed room means two unrelated people are asked
to share one bed, which is a re-room at the desk during the exact minute 180
people are standing in the lobby.

**The general principle: when a blocking decision is uncertain, compute the cost
of each direction of being wrong, not the probability of each outcome.** The
probabilities here are unknowable from the sheet. The costs are not.

## 4. The method, six steps

1. **Pull the arrivals list and group into waves,** using requested arrival time,
   segment history, and any transport schedule that fixes a time.
2. **Block by room TYPE first,** so every reservation's required type is reserved
   before specific units are assigned. Apply the cheaper-error rule to the
   ambiguous ones.
3. **Sequence housekeeping in wave order,** not floor order. The board is ordered
   by when a room is needed.
4. **Assign specific units late.** A unit assigned in the morning is a constraint
   all day; a type held in the morning is a decision deferred cheaply.
5. **Hold a flex buffer** out of the headroom for walk-ins, maintenance failures
   and re-rooms. On this day, 71 rooms of headroom is not 71 rooms of slack.
6. **Reconcile midday and re-sequence.** The morning plan is a hypothesis about a
   day that has not happened yet.

## 5. Where this method has no answer

**Parties of three or more are handled by hand.** There were 15 of them on this
day. No blocking rule covers them, and pretending otherwise would push the work
onto whoever is at the desk without telling them.

**Not every coach joins cleanly.** On this day one coach appeared on the dispatch
with a passenger count and no matching party detail — the record existed but the
join failed. The sheet says so explicitly and instructs the desk to work it from
the driver manifest on arrival. **A pre-arrival sheet that silently drops the rows
it cannot match is more dangerous than one that shows a gap**, because the desk
then discovers the gap at 17:45 rather than the night before.

**No-ETA reservations remain the standing weakness.** They cannot be waved, so
they are blocked by type and cleared into whatever readiness exists when they
appear.

## 6. What transfers, and what does not

The waves, the block-by-type-first rule and the cheaper-error rule are general.
They depend on nothing about this building.

What does not transfer is the shape of the day. A 400-room property taking coach
business has a wave 3 that dominates the afternoon. A 79-room property without
tour blocks may have no wave 3 at all, and running this sequencing there would be
ceremony. **Measure the arrival distribution before adopting the waves** — the
method is a response to a shape, and if the shape is not there the method is not
warranted.

The one thing worth carrying to any property regardless: **build the pre-arrival
sheet so it cannot hold a guest name.** Not scrubbed afterwards — unable to. It
costs nothing at design time and it removes an entire class of exposure from a
document that gets printed, carried and left on a desk.

---

*Version 2.0. Superseded versions stay listed rather than disappearing. Figures
are operational counts from one arrival day. No guest information, and no term
from any commercial agreement, appears in this document.*