Wheels analyzer

Wheels analyzer

Postby Gatsby » Tue Sep 25, 2012 7:47 pm

Hi,

Last week i purposed a new feature to Stan concerning the analysis of Wheels. This new analyzer would be able to deal with variable length of lines to be checked. as Stan questioned about the interest of other users in this feature i´m presenting this request in here in order to assert that and at the same time help Stan to devise about the priority of such feature.

>
> The ideia is EL to be able to accept and load wheels of different
> sizes (CSV or Txt) and and have a analyzer window where we could check
> draw by draw where winnings have happen. Maybe this could help
> choosing, inside the wheel, groups that have never been draw or considering
> their history (i.e according to the number of hits in each wheel
> line), less or more probable to contains more than 2 numbers at least.
>
> The analyzer would compare each line of the loaded wheel with each
> winning combination and present the results in a window configured for
> that.
>
> For example, a wheel like (50, 17, 5) has 921 blocks of 17 numbers
> that cover all 5 combinations. So, what i pretend is to know for each
> of winning numbers combinations what is the respective distribution
> among the 921 blocks of 17 numbers.
>
> The result could be presented in a matrix with the lines of the wheels
> and all the history draws in column (a designation as Draw1, etc,
> would be sufficient, though if one could hover over the draw and the
> corresponding numbers would appear would be nice), the intersection of
> the lines with the columns would then give how many numbers have been
> hit in each of the wheel line for each draw (0, 1, 2, 3, 5 ?). Maybe
> we could have the number of hits and in parenthesis the actual
> numbers. The number of back draws could be also selected, like in all,
> last 10, etc.

As Stan questioned that new table purposed would be quite difficult to read - whwn most of the table cells would be empty because there would be no hits, i realized that his concerns make sense, but only with very big wheels. that way what we would be
> looking for here is "coverings" with more numbers and lower bounds. I
> want to play from 11 to 20 numbers with some certainty to get x prizes
> when I "guess" n numbers from a set of v. . As I can see EL, namely
> Package winnings, doesn´t allow me to load lines of 17 numbers, and
> even if it did it does not permit to visualize a pattern of picks
> distribution among those lines.
>
> So, in an example for a covering design with 17 numbers, you have a covering design that covers all
> combinations from 50 numbers in 921 lines of 17 numbers. You can say
> you have 921 combos or even 921 different number groups. What i´m
> aiming is to track the pattern of hits among this 921 number groups,
> but this is not achievable with number groups manager, unless you´re
> capable of creating (and el5 supporting it obviously) all that groups.
> There as to be a concise way to check this, and this is the reason I
> proposed a distinct analyzer and a matrix. If you could have a files
> manager capable of loading this covering designs and checking their
> pattern,or make previsions on the base of some pattern that would be even better.
>
> The covering designs i´m addressing can go from dozens to many
> thousands, but a limitation could be imposed on this. Designs with
> thousand of combinations are not viable due to the fact that all too
> many cells would be empty. This is intended for designs with hundreds of lines, difficultly more.
>
> Imagine you have a line with 17 numbers in that covering that had in the last 5 draws
> always
> 3 hits when normally it varies between 0 and 3. On the base of that
> you could discard all combinations of 5 numbers from that set of 17.
> Or, with the help from other filters, choose some line of 17 numbers
> that have more numbers of those you think are more probable in the
> next draw. Maybe the combination of these two perspectives helps to make a better option.

Well, this being said, some feedback on this subject would be very welcome for, as said before, permit Stan to devise the necessity and priority of such proposal.

thanks,
Gatsby
 
Posts: 305
Joined: Tue May 01, 2007 3:24 pm

Return to Comments, suggestions, feature requests

Who is online

Users browsing this forum: No registered users and 11 guests