[samatman]

This is the perfect illustration of the gambler's it-isn't-necessarily-a-fallacy!

Despite a uniform distribution of tetrominoes, if you haven't seen that all important I in awhile, the run is "cold" and you are in fact "due".

For the gambler's fallacy to apply, events have to be both random and uncorrelated, where most interesting real-life situations (i.e. not games) the degree of correlation between events isn't zero: that's one way to observe a normal distribution, but not even the usual one.