Re: Chex Wat: Pi is "random" and "not predictable"?

On Wed, 01 Aug 2007 12:37:44 -0700, Seanpit wrote:

On Jul 31, 4:56 pm, Shane <remarcsdNOS...@xxxxxxxxxxxxxx> wrote:

The sequence of numbers in pi is most certainly predictable.

So predict the next five digits, from this 30 digit sequence.
724587006606315588174881520920.....

962829254091715364367892590360 011330530548820466521384146951
941511609433057270365759591953 092186117381932611793105118548
074462379962749567351885752724 . . .

Pretty good aren't I?

Nope, you didn't show your work.

Also, the string I posted could be from a different part of the
sequence than what you assumed, in which case your calculation and
your "prediction" are wrong.

I doubt very much that you took your sequence from anywhere else in
Pi.

What are the odds? ; )

It was a certainty that you didn't predict it.

Sure I did. All "predictions" are based on calculations.

Agreed, but they are not calculations themselves, that is why we have
the two different words. You do not write 2+2 on the board and ask
students to predict the answer, you ask them to calulate it.

That is how
one can determine the odds that the prediction is likely to be
correct. What other definition of "prediction" do you have? It
almost sounds like you expect a prediction to be some sort of mystical
process.

No, its just that you compared the string with a calculation for pi,
whereas the challenge was to predicet the next 5 digits using only the
30 digit sequence iteself. But hey, don't sweat it, I know full well
why you could not acheive that.

Oh and you are not allowed to use a calculation for pi, merely use the
given sequence. If required, I can make the sequence longer, but I
suspect that will not help.

The "calculation" or various algorithms for calculating pi is the
means by which pi is "predicted" and known to be non-random.

Then it is not *predicted*, it is calculated.

The prediction is based on a known calculation.

So you agree it is not predicted but calculated?

Just because the
prediction could have been wrong, doesn't mean it isn't useful or a
valid scientific hypothesis.

???

A
"calculation" or algorithm that produces pi every time is the means by
which one can indeed prove that pi is in fact predictable and
therefore not random.

Come on Sean, surely you can do better than that? You have just
reinforced that pi is, as Howard posted, calculable.

Of course pi is calculable. That's what makes it non-random. No
calculable number is actually "random" or "non-predictable" - as long
as you know the algorithm.

It's certainly non predictable, as has been pointed out elsethread,
and as you demonstrated yourself by requiring a calculation of pi
itself, rather than a calculation on just the sequence, to arrive at
your "prediction".

It seems you have a different definition of "predictable". Please do
define what it means to be "predictable" or "non-random" . . .

As above, I asked you to predict the next 5 digits of a 30 digit
string taken from pi, using only the 30 digit string as a basis for
you prediction. It seems you could not do this, so fell back on a
calculation for pi and found the numbers. IOW you calculated them, you
used no prediction at all.

You have not
demonstrated that it is predictable
based on the previous sequence, no
matter how long that sequence is.

I'm not sure what you are driving at . . . ? If you are suggesting
that any finite sequence can theoretically exist anywhere in pi an
infinite amount of times, that's almost certainly true. This does not
make pi "random" or "unpredictable". You're trying to confuse two
different concepts.

Actually I am trying to distinguish beytween them, you seem to be the
one confusing prediction with calculation.

What is your definition of "prediction"?

Different from yours, as it does not include pure calculation.

Your suggestion that given a finite sequence without any obvious
repeating pattern, the subsequent sequences could not be predicted, is
both true and not true. It is true that the subsequent sequences
could not be predicted with perfection. However, it is not true that
the subsequent sequences could not be predicted with better than even
odds of success.

For example, lets say scientists happened to detect a radiosignal
coming from outer space containing the first 100 digits of pi repeated
over and over again 1,000 times in a row (a total of 100,000 digits.
While it is certainly possible that such a sequence could have been
the result of purely random generation and could exist anywhere an
infinite number of times within the infinity of pi, what are the odds
that such a sequence would be the result of a random vs. a biased
process of generation? Which hypothesis would be right more often
than not? - The hypothesis of random generation or biased generation?
I can guarantee you that such a signal, if actually detected, would be
front-page news across the world.

Non sequitur. The subjecdt was the facility to predict the subsequent
digits of a string, not detect a recurring pattern.

You miss the point. The point is that the above mentioned "recurring
pattern" could actually be found in pi. Therefore, it is possible
that is was in fact the result of truly random generation and is not
therefore "biased" in origin and therefore does not represent some
ETI. This is always possible. It is just that, at some point, it
isn't the most likely possibility.

Non sequitur. The subject was the facility to predict the subsequent
digits of a string, not detect a recurring pattern. And where did the
ETI bit come from?

I'm not any form of mathematician,
and yet I can see the glaring distinction
between what Howard claimed
about pi and what you say he claimed.

Can you now?

Yep. He claimed that pi could be calculated--which you agree with--
but not predictable based on the sequence--which you disagreed with at
the time, but have subsequently agreed with.

Pi is predictable based on the sequence itself, by itself,

So show how you arrived at your prediction, not calculation, of the
above 5 digits.

and becomes
more and more predictable with each additional success

Additional success??? there has yet to be one success.

[...]

.

Relevant Pages

  • Re: Chex Wat: Pi is "random" and "not predictable"?
    ... I doubt very much that you took your sequence from anywhere else in ... one can determine the odds that the prediction is likely to be ... The prediction is based on a known calculation. ... - The hypothesis of random generation or biased generation? ...
    (talk.origins)
  • Re: Glenns prediction fails.
    ... at least *partially* right - perfect prediction of pi would not be ... does not mean that the prediction based on a finite sequence would ... prediction of "the next n digits of pi" would be correct. ... And, if our conjecture about the structure of pi is correct, then yes, ...
    (talk.origins)
  • Re: Chex Wat: Pi is "random" and "not predictable"?
    ... Predict the next two digits from knowing these six: ... successful prediction over time, the more predictive power it will ... given sequence. ... - The hypothesis of random generation or biased generation? ...
    (talk.origins)
  • Re: Chex Wat: Pi is "random" and "not predictable"?
    ... Predict the next two digits from knowing these six: ... So for your prediction you need more than the sequence and that it ... - The hypothesis of random generation or biased generation? ...
    (talk.origins)
  • Re: Chex Wat: Pi is "random" and "not predictable"?
    ... do muck it up with your first term, the size of total sequence space, ... total sequence space minus some value involving hd. ... calculation of the odds of some randomly chosen sequence matching a ... one confusing prediction with calculation. ...
    (talk.origins)