View Single Post
  #386  
Old July 16th 04, 08:07 AM
Regnirps
external usenet poster
 
Posts: n/a
Default

"Jeff Crowell" wrote:

I'll take issue with this, Charlie.


While each throw is statistically independent (assuming honest dice,
naturally), the fact that they are honest dice requires that the most
common throw be a seven. The more consecutive boxcars you
throw, the higher the probability that the next throw will NOT be
a 12. Boxcards is not a statistically likely event. Each throw **is**
an independent event, but the total population of throws is governed
by the overall statistical distribution.


You are trying to apply probability to a single event, which isn't valid. It
only covers ensembles, or groups of events. I should have used a single die in
my example or labeled them a and b so that all combinations are unique.

Take coin flipping instead. No mater how many times in a row you get heads, the
chances on the next toss are still 50/50. There is no reason not to get 100
heads in a row. The house plays the odds on thousands of players. It doesn't
work very well for the individual.

-- Charlie Springer