I'm skeptical of any system that proves it can predict stock prices vaguely. I would be less skeptical if it was more humble, and specific in its constraints.

I'm skeptical of any system that proves it can predict stock prices vaguely. I would be less skeptical if it was more humble, and specific in its constraints. For example, could it predict the % probability a stock would up X% if it had N consecutive negative days? Or, how many minutes/hours after earnings are announced is the after-hours price usually indicative of what the price will be the next day? You can't predict everything, but you might be able to have a probabilistic prediction of specific scenarios. though even then I'll still be skeptical.. daily stock movements are totally random.