, Britain
A baker in Bradford works backwards from the moment the shop opens. Bread is wanted in the morning, so it is baked before the morning, so the dough is mixed before that, and somebody is in the bakehouse in the dark every day of the week because that is the only way the loaves are warm when the door opens.
From 1 January 1974 that was possible on three days out of seven. The ovens were electric, the electricity was rationed to three consecutive days a week for commercial users, and bread does not keep.
Most accounts of the three day week are about factories, where the problem was output and the answer was compression. A bakery had a harder version of the same problem, because the thing a bakery makes is worthless four days later.
Three consecutive days is a worse constraint than three days
The restriction that ran from 1 January to 7 March 1974 gave commercial premises three consecutive days of electricity. Consecutive is the word that does the damage.
Given Monday, Wednesday and Friday, a baker could have put fresh bread into the shop every other day and covered the week tolerably. Nobody was given that. You got a block of three and then a gap of four, and the gap is the problem, because the fourth day of a gap is a day when the shop has nothing fresh to sell and the customer has to go somewhere that does.
So the choice came down to where you put the block, and there is no right answer, only a choice about which half of your week you are prepared to lose.
Take the first half of the week and you bake Monday, Tuesday and Wednesday. The shop is properly supplied when trade is quietest, and the end of the week, which is when people actually buy, is served with whatever survived. Take the second half and you bake Thursday, Friday and Saturday, you are strongest into the busiest days, and the shop opens on Monday with nothing worth selling and a customer base that has spent the weekend getting used to buying bread elsewhere.
A baker supplying other shops rather than selling over the counter had a third version of the problem, and a worse one, because their customers had their own three days, and there was no reason those three days were the same three days.
Compressing the week runs into the oven, not the effort
The instinct is to bake a week of bread in three days. On the powered days you could run right through the night, because the supply was there, and plenty of bakeries did exactly that.
It only goes so far, and what stops it is not willingness. An oven holds what it holds and takes as long as it takes. A week of output through three days of oven means roughly doubling throughput on each of those days, which means somewhere to prove twice as much dough, twice as many tins, twice as much space to cool what comes out, and enough hands to keep all of it moving. The bakehouse was built for a normal week and none of those things expand because the calendar has changed.
And the output has to go somewhere. Bread baked on Wednesday for sale on Saturday is not bread anybody wants to buy on Saturday. Plain bread is past its best within a day and effectively unsellable at full price by the second. The goods that keep are the enriched, fatty and sugared ones, fruit cake and shortbread and the like, so one genuine answer was to shift the mix away from what the bakery was for and towards what would survive the gap. That is a real option and it changes the business. It also means competing on a different shelf with everybody else who had the same idea.
Refrigeration was not the way out either. A freezer full of dough needs power on the four days there is none, and the industrial answer that arrived in later decades, part-baked bread finished off wherever it is sold, was not yet ordinary.
The wage bill did not shrink to fit
The one cost that took no notice of the restriction was labour.
Bakery work is skilled. A baker who knows your ovens, your flour and your timings is not replaceable at short notice, and the restriction was known to be temporary. Laying people off for ten weeks meant looking for them again in March, when every other bakery was looking too. Keeping them meant paying for days on which no work could be done, out of a business that was taking less. The choice was mostly settled by how much cash sat behind the business rather than by what the owner would have preferred.
The other pressure was on price, and it went the wrong way. Costs were climbing hard through 1974, flour among them, but a baker could not simply reprice to cover a shortened week. Price control was in force, and it worked on costs rather than on scarcity. A higher flour bill could be passed on. The four days a week the ovens were cold could not, because nobody had charged you for them. The shortened week was uncompensated by definition.
The shop and the bakehouse were two different businesses
The useful thing to notice, and the thing that separated the bakeries that came through from the ones that did not, is that only half of the operation was rationed.
The bakehouse needed electricity. The shop did not. A shop can trade in daylight, take cash, and sell anything on its shelves, and plenty of shops did exactly that through those ten weeks with the lights off. So the constraint was on production, not on selling, and a baker who thought of the shop as a place to sell goods rather than as the front end of an oven had options: bought-in lines that kept, goods that did not need baking, and a reason for the customer to come in on the days there was no fresh bread.
Those customers were the whole asset. Ten weeks is long enough for a household to change where it buys its bread, and long enough for that change to stick.
Where this shows up in the game
Three Days a Week is a historical start at hard difficulty. You run Horton Road Bakery in Bradford from January 1974, as a sole trader, for twelve months. The objective is simply to be still trading at the end. Survival alone scores, and finishing without having made anybody redundant scores higher, because the jobs you held are counted in the epilogue.
For the first three months, capacity is capped at three days in seven and cannot be raised by hiring, which is the whole point: this is a constraint you cannot work your way out of. Anything baked on a powered day and not sold spoils, so overbaking to fill the gap has a direct cost. Wages are still owed for the unpowered days unless you lay staff off, and laying them off costs you reputation. Those three rules pull against each other on purpose, and the year continues for nine months after the restriction lifts, which is where the rest of 1974 does its own damage. The failure condition is the bakery closing.
The seed is fixed at 600003, so everybody attempting it meets the same winter and the same customers, and the results are comparable. It runs under ironman rules: one save, and what happens happens. The Three Days a Week page lists the start position, the objective and the rules.