Four farmers on one field, and the day that was boring

The owner’s order for the shift was a question nobody could answer by playing: mimic one, two, three and four players on the economy alone, and say whether the extra people are bored. The game’s co-op economy is already one shared pool, so the mimic is four bodies drawing on it, not four machines. Each extra hero is the player’s own character with a plain AI in the seat, and the autopilot brain runs on it unchanged; what had to be built was the manners, so four heroes never chop one tree, and the ledger, so every hero writes down at dawn how many seconds it chopped, carried, built, fought, watered, and how many it stood with nothing to do and why. Then the same ten seeds ran four times.

Boredom exists and it lasts two days. On the first morning the plot has twelve trees and four axes empty it in a minute and a half, so the third and fourth player stand idle for a sixth and a quarter of the day, every idle second “no tree”. On the second day it is under a tenth. From the third day on it is zero in every arm, because the forest the players plant back outgrows what four axes can fell: fifty trees stand at the tenth dawn whatever the head count. So the question he asked has a number, and the number says the first morning, not the game, is where the extra people wait.

The finding nobody asked for is the one that matters. With more hands the village starves. Solo, the village has grain and no hunger at the eighth dawn; every co-op arm has no grain and full hunger, one villager fewer and a level lower, while the logs pile up fifteen times higher and the guns stand at the cap. The mechanism is plain in the seed logs: one farmer on one field sows every thirty seconds in all four arms, but four pairs of hands raise houses faster, so the fourth mouth arrives on day four instead of day eight, and one field feeds about three. Solo hits the same wall on day ten. The test proxy has exactly one rule for farms, which is to build one, and no player with four friends and an empty larder would stop there. A second reader given the same evidence blind found the same two things and the same cause.

The instrument had two faults of its own, both caught by the numbers before they were believed. The claims that keep two axes off one trunk were cleared every tick, so the farm plan flipped between bodies frame by frame and the four-player village took four times as long as the solo one to raise its first field; the claims stick now. And a check that skipped grove spots off the floor for the larger groves of a lever also culled three trees from the default grove, and every solo number moved with it; the check now applies only to the enlarged grid, and the solo run is byte-identical again to the build before any of this. Four levers went in as switches, all off: more trees per player, more stone, a bigger wave, faster regrowth, and the proxy’s second farm for a hungry village. They are being measured against the corrected baseline on the same seeds as this is written. One design question went to the owner: the plot’s checklist counts only the host’s own logs, so three friends felling the nearest trees leave the host chopping the tutorial’s ten alone.