Given a large string of Base-N digits (in other words, not necessarily binary), is there software or an algorithm that can output whether the data is reasonably pseudorandom or reveals a pattern?
In context, if one were to record and string together hundreds of thousands of results from a casino game, each taking a value of 0 through 5, what can be done to discover whether the results are pseudorandom, or intelligently designed to produce some unknown psychological effect on the gambler? As such design hypotheses take almost any form one can imagine, a general solution or a solution testing a wide range of hypotheses would be required, as opposed to testing hypotheses one at a time.