@ chris campbell, post "lagging digits"."The implications of this on the upcoming digits in the sequence are rather interesting, as this would mean that zero is actually more likely to occur than the other numbers, something of a paradox considering the random nature of the decimal place string."
This is an improper usage of the law of averages. The law of averages does not take in to account events previously occurred. It merely states that as more trials occur, or the limit of trials approaches infinity, the experimental probability should converge on the theoretical probability. I agree with the idea that if you look from zero the proportion of any digit to total (experimental probability) should approach the theoretical probability; henceforth, the difference should occur more often to have the proportion equal out; however, it still an improper usage of determined fact.
-- (December 25, 2009) on Frequency of Each Digit of Pi