Hi Stan,
Cant say i quite understood this. I didn´t remove any groups from the list, matter of fact i created some more.
So, does this means that everytime i load another group i will have to costumize the groups in the stats filter?
i´m working with around 50 groups in the statistical filter. From time to time i develop some more. i cannot go on changing the filter everytime i load another one.
Do you think that if i create new group that goes to the end of the group list in the group manager this problem persists?
Please explain a little better what happens when a group is created or erased from the manager, because i cant´t see why the stats filter would be perturbed by that. isn´t the name of the filter what matters and not the position it occupyies on the list?
Regards,
.
BTW, now i will see Portugal play, hope it wins
Hi Stan,
I´m editing this reply because i´ve looked once again to number groups and i definitly think something is wrong with the filter. i´ve used a single number group in stats filter and the only division between results is that "accept" has only tickets with 5-0, or 0-5 and "reject" with all the other combinations in a 2 group (1-4, 4-1, 2-3 and 3-2). If i use "count" instead of "distribution" it seems to work, leaving only the counts as the latest one. So, to me the filter isn´t recognizing the pattern of the last draw and filtering accordingly to it. I´ve checked all the other colums of the other groups (package results properties window)and and don´t think there´s one with all the same results in the filtered tickets, that is, none of the filtered tickets as the same propertie for the same group in all the lines. If last pattern would be 3-2, and i accepted only last one, i would expect only tickets with 3-2. that doesn´t happen. And btw, in the design window when i choose a different number group, the applies x to x allways refers to the dimension of that group, if it´s 2 is 1 to 2 or 2 to 2, if it´s 3 is 1 to 3 or 2 to 3 or 3 to 3. so the filter recognizes the number group and it´s composition. It just doesn´t filter accordingly to the latest pattern.
You say that the filter might be using some other number group different from the one that was primarilly in the filter. But if this was the case, wouldn´t we expect choosing accept latest = 1, that some other filter would have a colum with the same distribution in every row?
And for this to happen, considering the definitions on the filter, wouldn´t have to be between number groups of the same dimension too? (Not that it should work this way)
Regards,