Yeah I tend to agree with Alan on this one. With such a wide range of possibilities in VP you will invariably run into what looks like amazing patterns within randomness. I won't go so far as to say Rob is completely wrong, because there are real stories of gaffed machines. I just wouldn't lose much sleep over it.

Here's an interesting story to follow up on that:

A math professor once told his students to do a homework assignment flipping a coin 200 times and to record the results. When the students turned in their work he could tell right away when they cheated with fake results without actually flipping the coin. How could he tell? Because the odds are astronomically against anyone not flipping a coin on the same side six times in a row during a 200 flip trial. Most students didn't know this and the cheaters could easily be flunked.

Six flips of a coin on the same side looks like a freakish occurrence (1/64 chance), but it's to be expected within a set of 200 flips.