Free Republic
Browse · Search
News/Activism
Topics · Post Article

To: Doctor Stochastic
Wrong, there's no contradiction. You are exhibiting a typical misunderstanding of complexity

No. I used your definitions. You stated random mechanisms always yield complexity. A random mechanism yielded pi which you said was not complex.

The sequence of crossings (or non-crossings) in a large number of throws may be complex.

That "probability" does not equal one then.

Only a randomly generated sequence can yield complexity.

How many times do I have to flip a coin..... Oh never mind the "probability" is still not one.

1,149 posted on 05/26/2005 10:07:29 PM PDT by AndrewC
[ Post Reply | Private Reply | To 1146 | View Replies ]


To: All; Doctor Stochastic; PatrickHenry
[A confused person wrote while dropping red herrings:] No. I used your definitions.

No you didn't.

You stated random mechanisms always yield complexity.

Clue for the confused #1: An Introduction to Kolmogorov Complexity and Its Applications

A random mechanism yielded pi which you said was not complex.

Clue for the confused #2: This random process does *not* yield pi (nor did Doctor Stochastic say that it did). It yields a random rational number of rather high complexity which is statistically constrained to likely be in the close neighborhood of pi, because the probability of a "hit" is related to pi itself. But the result itself is guaranteed to *not* actually be the non-complex constant "pi" (because pi is irrational, whereas the result of the Buffon needle-throwing experiment will be rational since it is the quotient of two integers).

Clue for the confused #3: Tallying up the "hits" and dividing by the total number of drops is a deterministic conversion which drastically reduces the amount of complexity in the original sequence of drop results.

Clue for the confused #4: Consider the similar case of flipping a fair coin 1023 times in a row. The resulting sequence has a Kolmogorov Complexity near 1023 bits. Now do the Buffon's Needle method in order to approximate the probability of a "hit" (a head) on any given flip, which is exactly 0.5 -- divide the number of heads (0-1023) by the number of flips (constant 1023). Due to the "lumping" effect of extracting only the total number of heads (while ignoring their sequence), the numeric result of (Nhits/1023) has a Kolmogorov Complexity of only around 10 bits (fewer, actually, since the distribution of the number of heads is not uniform across 0-1023), resulting in over a 99% reduction in complexity.

[The sequence of crossings (or non-crossings) in a large number of throws may be complex.]
That "probability" does not equal one then.

Clue for the confused #5: "May" in this context is not used in the sense of "may or may not". It is used in the sense of, "you may have a brain, but you're not using it." Doctor Stochastic was saying, "Although the sequence of throws may be complex, a deterministic description of throws is not".

[Only a randomly generated sequence can yield complexity.]
How many times do I have to flip a coin.....

How much complexity do you want?

Oh never mind the "probability" is still not one.

Sure it is.

1,163 posted on 05/27/2005 12:22:40 AM PDT by Ichneumon
[ Post Reply | Private Reply | To 1149 | View Replies ]

To: AndrewC

Again you are making common mistakes in probability. Just because a random mechanism yields pi does not make pi complex. By your reasoning, 1/2 would be complex as there are random mechanisms which yield 1/2.

Pi is not complex because there are simple mechanisms, which I already posted to you, that yield pi.

Likewise you are mistaken in the meaning of probability 1. These matters are covered in elementary courses in probability theory.


1,224 posted on 05/27/2005 6:10:23 AM PDT by Doctor Stochastic (Vegetabilisch = chaotisch is der Charakter der Modernen. - Friedrich Schlegel)
[ Post Reply | Private Reply | To 1149 | View Replies ]

Free Republic
Browse · Search
News/Activism
Topics · Post Article


FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson