11th February 2013, 06:58 AM
|
|
Member
|
|
Join Date: Sep 2011
Posts: 1,534
|
|
Quote:
|
Originally Posted by michaelg
As there is doubt about the efficiency of the filter I now hope to look at the three days from Friday 8 to yesterday and see how the deleted (T, D and C) selections would have fared. When known, I will detail the result here.
|
michael,
You will need at least 1000 (or even better 2000) selections to make any sort of assumptions to these filters for this system. Your strike rate is about 98% so basing filters on a few days results is not the best idea. |