Which Shift Pattern Should You Actually Use?

Do you actually need a named pattern?

If you are small, probably not, and that is worth saying before the rest of this.

Most small operators schedule against demand and against the availability their team has handed in, and never adopt a named pattern at all. That is not a lack of rigour. Named patterns solve one specific problem: filling the same post continuously, with a fixed set of crews, fairly, on repeat, for years, without anybody re-deciding it every month. If your operation closes overnight, or your team is small enough that you can hold everyone's availability in your head, the pattern is answering a question you have not asked.

They earn their keep at scale and at continuity. A plant that never stops, a hospital unit, a control room, a security contract. Somewhere the same post has to be filled every hour of every week and the fairness has to survive being repeated a hundred times.

So read what follows as a set of solved problems you can borrow from rather than a menu you have to order from. Even if you never adopt one, the mechanisms inside them are worth knowing: how they rotate weekends, how they handle nights, and where they put the overtime. Those problems arrive whether or not you have named your pattern.

What actually decides the pattern?

Three things, in this order.

How many hours the post has to be staffed. A position covered around the clock is 168 hours a week. A position covered two shifts a day, seven days, is 112. A position covered during business hours only is more like 50. These are completely different problems and they do not share solutions.

How many people you have. This is the constraint that quietly eliminates most of the options. Four people on a continuously staffed post means twelve hour shifts, because there is no eight hour arrangement that covers 168 hours with four people without enormous overtime. Six or more and eight hour shifts open up. If you have not worked out the headcount yet, start with how many employees you need for 24/7 coverage, because the answer narrows this decision considerably.

What your staff will actually accept. A pattern people hate is a pattern that generates turnover, and turnover makes the headcount problem worse. This sounds soft and it is the most expensive of the three to get wrong.

Notice what is not on the list: which pattern sounds most efficient. Every pattern below covers the same hours. None of them creates labour out of nothing.

One boundary before the detail. This article is about rotation patterns for covering a shift-based operation. It is not about employment arrangements like flextime, job sharing or remote work, which are a different question covered in work schedule types.

Which patterns cover a 24/7 operation?

These are the patterns built for continuous coverage. All of them use twelve hour shifts, because two twelve hour shifts fill a day exactly and three eight hour shifts need half again as many people.

The 2-2-3, usually called the Pitman schedule. Two days on, two off, three on, then the pattern inverts the following week. It runs on a fourteen day cycle with four crews, and its defining feature is that it gives every crew a full weekend off every other cycle. That single property is why it is popular with the people who work it, and it is the reason it turns up so often in emergency services, hospitals and plants. The full mechanics are in the Pitman schedule guide.

The DuPont schedule. A four week cycle of twelve hour shifts, rotating between days and nights, ending in seven consecutive days off. The seven day block is the feature crews name first when asked why they like it. It asks more of people in the middle of the cycle in exchange for that stretch at the end. Details in the DuPont schedule guide.

Four on, four off. Four twelve hour shifts followed by four days off, on a rolling eight day cycle. It is the simplest of the three to explain and to build, and because the cycle is eight days rather than seven it moves through the week, so weekends fall differently for everyone over time. That is either its best feature or its worst, depending on whether your staff want predictable weekends.

All three average around 42 hours a week. That is not an accident and it is not a flaw. Four crews covering 168 hours is 42 hours each by definition. The two hours above 40 are overtime by design, paid predictably, in exchange for needing four people per post instead of five.

Which patterns are not for continuous coverage?

These get grouped with the ones above and should not be. They solve a different problem: shortening the working week for an operation that closes.

The 4/10. Four ten hour days, one extra day off every week. It suits operations with a long single shift and a natural close, and the trade is a longer day for a three day weekend. See the 4/10 schedule guide.

The 9/80. Eighty hours across nine days, with every other Friday off. The mechanism that makes it work is an eight hour crossover day, split down the middle by the workweek boundary so that each week lands on exactly 40 hours. Get the boundary wrong and you have built permanent overtime into every other week, which is the single most common way this pattern goes wrong. The Department of Labor's overtime fact sheet sets out the workweek definition the whole arrangement depends on, and the 9/80 guide walks through the crossover day. The federal guidance on compressed work schedules is a useful reference for both of these.

The Baylor plan. Two long weekend shifts, historically two sixteen hour shifts, paid as a full week. It exists to solve weekend staffing specifically, and it is mostly a healthcare arrangement. It is a weekend coverage tool rather than a whole-operation pattern, and the arithmetic on a sixteen hour shift deserves its own look in the Baylor shifts guide.

How do the patterns compare?

What does each pattern cost in fatigue?

This is where the patterns genuinely differ, and it is worth more attention than the headcount arithmetic because it is the part that shows up as errors, injuries and resignations rather than as a number in a spreadsheet.

Two variables matter more than the rest.

Shift length. OSHA's guidance on extended and unusual work shifts treats anything beyond the standard eight hour day as carrying additional risk that has to be managed rather than absorbed. NIOSH's review of long work hours links extended shifts to reduced performance and higher injury rates. This does not make twelve hour shifts wrong, and plenty of crews prefer them. It means a twelve hour pattern needs breaks, handovers and overtime limits that an eight hour pattern can get away with neglecting.

Rotation speed and direction. Moving somebody from days to nights is harder on them than either days or nights alone, and how often you do it matters. Patterns that flip people between days and nights inside a single week are the hardest to live with. Forward rotation, moving days to evenings to nights, is generally easier to adapt to than backward rotation. The rotating shifts guide covers the mechanics, and the effects of working night shifts covers what you are actually asking of people.

A pattern that looks identical on a spreadsheet can be humane or brutal depending on these two, which is why "we run 2-2-3" tells you much less than it sounds like it does.

So which one should you pick?

Work down this list and stop at the first match.

Does the operation close at any point in the week? If yes, you are choosing between a standard five day week, the 4/10 and the 9/80. None of the twelve hour rotation patterns apply. Pick the 4/10 if a longer day is acceptable and people want a weekly three day weekend. Pick the 9/80 if ten hour days are too long but a fortnightly Friday is attractive, and only if you are confident about the workweek boundary.

Is the operation continuous, with four or five people per post? Twelve hour shifts, and you are choosing between 2-2-3, DuPont and four on four off. Pick 2-2-3 if predictable alternating weekends matter most. Pick DuPont if people would rather bank their time off into one long block. Pick four on four off if simplicity matters more than weekend predictability.

Is the operation continuous with six or more people per post? Eight hour shifts become workable, which reduces fatigue risk and gives more scheduling flexibility, at the cost of more handovers. In practice most teams at this size never formalise it into a named pattern. They build the rota from availability each cycle, and for a single site that is a perfectly good answer.

Is the problem specifically weekends? Look at a dedicated weekend arrangement like the Baylor plan before reshaping the whole operation around it.

What to do before you change anyone's pattern

Changing a shift pattern is a bigger intervention than it looks, because people have built childcare, second jobs, studies and commutes around the current one.

One thing worth naming first, because it is usually the real story. Most pattern changes over the last few years have not been optimisation projects. They have been a response to running lighter than the business would like. A pattern change made because you are short-staffed is a different thing from one made because the pattern is wrong, and it is worth being honest with yourself about which one you are doing. A rota redesign will not fix a headcount problem, it will only move the shortfall somewhere less visible, and the people being asked to absorb it will know the difference immediately.

Three things worth doing first.

Model the hours before you announce anything. Work out the average weekly hours the new pattern produces per person, and where it crosses 40. A pattern that averages 42 hours is fine if you have decided to pay for it and a problem if you discover it in the first payroll run.

Ask the people who will work it. Not as a vote, but because they will spot the practical problems you cannot see from the schedule. The crew who currently swap two shifts a month to make school pickup work will tell you immediately whether the new pattern kills that.

Give real notice. Depending on where you operate, you may have a legal minimum for schedule changes, and in several cities changing a published schedule carries a premium. That is covered in predictive scheduling laws. Even where no rule applies, a pattern change announced two weeks out is a pattern change people will resent.

The pattern that works is almost never the one that looks most elegant on paper. It is the one that covers the hours, keeps the overtime where you decided to put it, and that your team can plan a life around.