A history of video games has to begin before anyone could play one, because the business that would eventually sell them was already fully formed. By the 1930s American manufacturers, most of them clustered around Chicago, had worked out how to make money from very short amusements. Build a machine cheap enough to place in a drugstore or a tavern, charge a coin for a game lasting a minute or two, and split the takings with whoever owns the floor space. Route operators moved the machines between locations as the novelty wore off. Nothing in that arrangement required electronics, and when screens arrived decades later they slid into a distribution system that was already waiting for them.
The same trade inherited a long argument about what kind of activity it was selling. Coin-operated machines sat uncomfortably close to gambling in the eyes of city governments, and several large American cities banned pinball outright, treating it as a rigged way of taking money from the young. Manufacturers answered by making the games more obviously a test of skill, and the arrival of player-controlled flippers in the late 1940s gave them their strongest argument yet. The distinction between chance and skill was never abstract. It determined whether a machine could legally be installed at all, and it set the terms of every later fight over whether games were harmless fun or a public problem.
The other half of the inheritance came out of the war. Radar, gunnery calculation and television research put a whole generation of engineers on intimate terms with the cathode-ray tube, and left behind both surplus components and settled habits of thought about tracking a moving object across a glowing screen. One result was an unusual patent filed in early 1947 for an amusement device that steered a spot at targets printed on a plastic overlay laid across the glass. It was never manufactured and it had no descendants at the time. It matters because it shows how naturally the technology of aiming and detecting suggested the idea of a game.
Computing in those same years was not a consumer activity, and nobody involved expected it to become one. The machines being assembled in universities, government laboratories and a handful of corporate research divisions were instruments, built to work out firing tables, process census returns or test theories about logic, and their running time was rationed accordingly. Yet a game is a compact way to demonstrate what a machine can do, which is why one of the first relay contraptions built purely to entertain a crowd turned up at a world's fair rather than in a laboratory report. Three separate traditions were now in place: a coin-operated trade, a mature display technology, and machines that could be programmed.