>> In isolation, human behavior is actually quite predictable; only in the aggregate does it get difficult to predict.
It's the opposite, isn't it? You can predict what a group of people will do on average, but you can't know whether a specific person picked out of that group will be average, or an outlier.