From: Iain Strachan <igd.strachan@gmail.com>

Date: Sat May 09 2009 - 13:26:40 EDT

Date: Sat May 09 2009 - 13:26:40 EDT

On Sat, May 9, 2009 at 5:03 PM, Dave Wallace <wmdavid.wallace@gmail.com>wrote:

*> Iain Strachan wrote:
*

*>
*

*>>
*

*>>
*

*>>
*

*>> I remember we discussed this quite a bit on the list when Lawrence first
*

*>> posted the link to Koonin's paper, with the astronomical 10^(-1018)
*

*>> probability calculation.
*

*>>
*

*>> The problem I have with Koonin's approach is the same as the problem I
*

*>> have with ID.
*

*>>
*

*>> The difficulty is that it isn't really an explanation. To invoke
*

*>> something (be it Multiverse, or an Intelligent Designer) that can explain
*

*>> ANYTHING is no explanation at all. It is as if I had a data set of, say 100
*

*>> measurements and stated that given a sufficiently complex mathematical
*

*>> function y(x) I can produce an exact fit to my dataset. But this applies
*

*>> even if the data is just random noise.
*

*>>
*

*> And the resulting polynomial is close to useless for interpolation or
*

*> extrapolation.
*

Just to be pedantic and geeky, I should add that even if a low order

polynomial were used to get a reasonable fit to the data, it would still be

useless for _extrapolation_ because the polynomial grows without limit

outside the domain of the data fitting. For example, function approximation

using Chebyshev polynomials is only valid within a fixed range (I think

normalised to the range -1 to +1.

However your point about _interpolation_ is exactly what I meant.

Regards,

Iain

To unsubscribe, send a message to majordomo@calvin.edu with

"unsubscribe asa" (no quotes) as the body of the message.

Received on Sat May 9 13:27:13 2009

*
This archive was generated by hypermail 2.1.8
: Sat May 09 2009 - 13:27:13 EDT
*