To: Greg S. who wrote (850 ) 1/27/2000 1:39:00 AM From: Sun Tzu Respond to
As you pointed out, the interdependencies of the win or loss are hard to calculate. So let's focus just on card distribution. Assuming a random distribution of cards, you would expect that in a large enough sample, one third of the cards are high and another third are low cards (and a third in between), as evidenced by playing the whole set. On the other hand, as the subset shrinks, predicting the percentage of the cards that will be in each group becomes impossible (how can I know what the next card will be right off the bat?) [Random Walkers pay attention] . BUT as high cards pile up, the sample size increases. Which in turn increases the odds that the next card will not be a high card.The percentage of stocks that are 2 standard deviations above their 200 day moving average should be ~4% (almost by definition). The market can withstand the overbought situation until this number reaches ~18%. Prior to October '87 this number was ~27%, in the mini-crash of '90 it was ~22% and in the April of '98 it hit the all time high of 35%. Read your stat books and try to calculate how unlikely it is for 35% of your samples to be 2 standard deviations apart from the mean.
