Shapes

The Pairing That Fails Soonest

An early day against a late one runs out two hours sooner than an identical pair. The two shapes have added a gap of their own.

Early hours at one end, late hours at the other. It is the most ordinary-sounding pairing in the record and it fails before any other.

Gap	early with late	early with early
0	6	8
2	4	6
4	2	4
5	1	3
6	0	2
8	0	0

At no gap at all, two people on these hours already share only six of their eight. At a gap of six they share nothing, while a pair on identical hours still has two hours left.

To make the comparison useful, teams can explore the management toolkit and keep the project, reporting period and correction case constant. A useful external reference for the surrounding topic is GitLab.

Where the missing two hours went

Early runs eight to four and late runs ten to six. The two shapes are two hours apart before anybody mentions clocks, so every gap is really that gap plus two.

Put differently: an early-and-late pair at a gap of four is in exactly the position of an identically-shaped pair at a gap of six. The arithmetic does not distinguish between separation imposed by geography and separation chosen by habit, and neither does the day.

This is the cheapest two hours in the record and nobody thinks to look for them, because they are not on the clock, they are in the diary.

The counts

Across the twenty-five gaps, an early-and-late pairing leaves nothing at ten of them — the same count as every other pairing of unbroken eight-hour days, since the count only measures how far the pairing reaches from its own centre and all of them reach seven hours either way from it.

What differs is where the centre is. For an early-and-late pair it sits two hours off zero, so the workable range runs from a gap of five in one direction to a gap of nine in the other, rather than seven and seven.

That asymmetry is the practically useful fact. The pairing does not simply lose two hours everywhere; it loses badly in one direction and gains in the other. An early day paired with a late day at the right end survives further than two identical days do.

The direction that helps

The asymmetry is worth working through, because it contains the only piece of genuinely free advice in the section.

An early day at one end and a late day at the other are two hours apart in habit. If the clock gap runs the same way as the habit gap — the late end is also the end that is ahead — the two separations add, and the pair give out at a clock gap of five.

If the clock gap runs the other way, they cancel. The late end is behind on the clock, its later start partly undoing the lag, and the pair survive all the way to a clock gap of nine, two gaps further than two identical days manage.

Clock gap	early with late
5	1
6	0
-8	2
-9	1
-10	0

Which means the same pairing of habits is the worst in the record in one direction and the best available from unbroken days in the other.

Who should keep which hours

That gives a rule with no cost attached to it at all.

The end that is behind on the clock should keep the later hours. Its working day then starts later in its own local terms, which pulls it towards the other end rather than away.

The end that is ahead should keep the earlier hours, for the same reason in reverse.

Doing it backwards — the end that is already ahead also starting early, the end already behind also starting late — adds up to two hours to an existing gap, and is the single most common unforced error in these arrangements, because each end chose its hours for local reasons and nobody compared.

Two hours can be gained or lost by which end keeps which habits, and the decision is usually made by whoever likes mornings.

The fix, where there is one

Both ends moving to the same shape recovers the two hours in full, at a cost of one hour's change to each person's habits. There is no other two-hour improvement available anywhere in this record at that price.

Where it cannot be done — because one end genuinely cannot start later, or the other genuinely cannot start earlier — then the pairing is simply a gap of two with a different name, and should be planned as one.

The general lesson

Before measuring the gap between two clocks, measure the gap between two diaries. In this record the second one is up to two hours wide, it costs nothing to close, and it is left out of almost every conversation about the first.