>>11024184That's wrong I think. Consider a random walk in a vertical line where you go up and down depending if the coin is heads or tails. This random walk will get arbitrarily far from the starting position, unboundedly. After a certain point, it might never go down to bellow it's starting position again even.
Likewise, a sequence of coin tosses taken 100 by 100 where you have something like 51~56 heads and 44~49 tails. The heads:tails ratio will approach 1 very fast yet the difference will only grow and grow.
Gambler's fallacy is called a fallacy for a good reason
