|
PureBytes Links
Trading Reference Links
|
greene@xxxxxxxxxxxxxxx wrote:
> Regarding the issue of data...
>
> He found it
> hard to believe that people would invest or trade money on the basis of tested
> systems if they weren't absolutely 100% sure that their data was as pure as the
> driven snow. After examining the data I had then (which was full of errors) -
> I thought his point was well taken.
I agree here. I take this data with a grain of salt. This is the reason why I
keep my time frames constant - I used to try to optimize them, but I felt this was
adding a degree of potential bias to the equation. On the other hand, you've got
to trade or invest on something, and no matter what criteria you use, there is no
sure way to know whether the data is pure or not. I'll admit, though, that
optimization does tend to make things worse, which is one of the reasons I use it
sparingly.
Regards,
A.J.
|