What's wrong with connecting dots? We project the past into the future routinely and reliably. Why can't it run backwards? Epidemiologists certainly do that. Consider today's cinema ... audience sizes, production cost, production values, censorship, graphic content, crew size, number of awards. It scarcely matters which parameters are picked as long as they can be measured. Next, examine those parameters for the 1990s, 1970s, and 1950s. From those data points, you can infer trends. From those trends, you can probably make a decent guess as to when motion photography was invented.