Cyclical Luck

[Zedd X]

Lets say that someone says that they have a foolproof system to playing roulette (a game in which we all know cannot be beat in the longrun). This system involves betting red or black based on previous rolls of the roulette ball. For example, if 5 reds hit in a row, there is a higher probability that black will be next and should be bet. What would be your reaction to this??? I'm sure the posters on this board would unanimously agree that this is complete garbage.

So why is it that some poker literature seems to hint at exactly this strategy?

Here is the most profound example that comes to mind:

Excerpt from Zen and The Art of Poker by Larry W. Phillips:
pg75-76

If your cards are below average, but you've been winning with anything and everything, you should play more hands. Conversely, if you've been getting fairly good hands, but you've lost with them, you might want to fold some more...Mathematicians tell us that each hand takes place independently of all others. This is good advice to ignore...Longtime experienced card players believe in this bunching of luck.

I have read similar advice in other wellknown poker literature and somehow I am inclined to believe this; although it doesnt seem to line up with the concept of random probability. Could this be due to human error in the shuffling of the deck (i.e. clumping of cards)? Or could it possibly be a non-random shuffling algorithm online??? These could contribute slightly to the factor, but I beleive these "non-random flukes" so to speak are so minute it can be ignored in the calculation of probable outcomes. Could it be a function of our own psychological involvment in the game? Or could this advice just be plain inacurate (although I am lead to believe otherwise).

Just something to think about.