送交者: steven 于 July 18, 2001 14:19:39:
回答: yes, like you said, 由 008 于 July 18, 2001 11:34:36:
I am not sure about your point here, Seems to me, you are trying to draw the conclusion based on whether
my guess of next tossing outcome is correct or not. As for 2037 toss example, what I was doing was to
point out a small set of experiment would not change the probability. In fact, any particular set of
experiment, the outcome can be arbitary. Probability experiment deals with large set of test. Coin
tossing has been test for so many years and for many different kind of coins, we have not yet see a
different result.Simply base on the outcome of 9 toss, and jump into conclusion saying the probability
is different is a bit lacking of mature thinking.
As far as the 9 tosses problem, the original question is whether random variable can be use to determine
the probability of next tosses, and the answer is not. Random variable does not work that way, and since
coin tossing is not a system with feedback, previous outcome does not affact the future outcome. That is
why the probability remains the same. You argued that same outcome occure consecutively, that means the
probability cannot be 1/2. The point is, your test sample is too small to draw any meaningful
conclusion. If you do not agree, try to send a paper to Probability Theory and convince the world that
based on your 9 coin tossing outcome, the probability of head or tail on tossing that coin is not 1/2.
On the otherhand, if you toss 9000 times, and the statistic shows the head and tail are significantly
not even, say different magnitude, then, I would agree for that coin, the probability of head or tail
are not equal, one is bigger than the other.