Playing Around With Various Running Averages

[Bret Cahill]

https://data.intrade.com/graphing/jsp/closingPricesForm.jsp?contractId=743474&tradeURL=https://www.intrade.com

Is there any average, i.e., 30 day, 90 day, 180 day, triangular,
exponential, etc. that does _not_ extrapolate out to above 50% in 3
months? Assume the extrapolation is just a straight line.

Is there any numerical technique that provides the criteria for the
lowest possible result?


Bret Cahill

 

[Clay]

On Aug 9, 11:22 am, Bret Cahill <Bret_E_Cah...@yahoo.com> wrote:

https://data.intrade.com/graphing/jsp/closingPricesForm.jsp?contractI...

Is there any average, i.e., 30 day, 90 day, 180 day, triangular,
exponential, etc. that does _not_ extrapolate out to above 50% in 3
months?  Assume the extrapolation is just a straight line.

Is there any numerical technique that provides the criteria for the
lowest possible result?

Bret Cahill

A digital crystal ball ;-)

 

[Fred Marshall]

On 8/9/2012 8:22 AM, Bret Cahill wrote:

https://data.intrade.com/graphing/jsp/closingPricesForm.jsp?contractId=743474&tradeURL=https://www.intrade.com

Is there any average, i.e., 30 day, 90 day, 180 day, triangular,
exponential, etc. that does _not_ extrapolate out to above 50% in 3
months? Assume the extrapolation is just a straight line.

Is there any numerical technique that provides the criteria for the
lowest possible result?


Bret Cahill


Bret,

The question makes little sense to me. Maybe you can refine it some?

An average cannot "extrapolate" as it is but one number. You can put a
straight line through a single point with *any* slope you want. But
that's not much of an extrapoloation.

To do a straight line extrapolation, you need a straight line fit.
There are all kinds of those. Depending on the data, and the method
somewhat, you will either get a positive slope or a negative slope.

Months, furlongs per fortnight, whatever.... don't mean much as you can
assign whatever independent variable you like to a data set. It's just
scaling.

As far as "50%" as an absolute value:
- 50% of what?
The stated quest is to get yet a third point on the straight line isn't
it?

The other stated quest is to yield a low point ... another third point
on the straight line?

So, I guess you realy want a negative slope. Good luck.

Fred