The use of filters and the quantified cover

Talk about anything Wheel Generator related which doesn't fit in the other forums
Post Reply
Frank
Regular
Regular
Posts: 34
Joined: Sat Nov 22, 2008 1:55 pm

The use of filters and the quantified cover

Post by Frank » Sat Nov 22, 2008 4:20 pm

Hi LA,

I think that Wheel Generator quantifies the wheel cover, no matter what filters we use.

I made a 8,6,5if6=4 wheel with cover =100% in Wheel Generator 1.5 Demo, using no filters; my numbers have been : 01-02-03-04-05-06-07-08

After clearing, I made one common numbers filter : 01,02,03 at least 3 of them; I cannot get a 8,6,5if6=4 wheel 100% cover now.

As you can check, two lines are needed to fullfill that wheel criteria:

01,02,03,04,05,06
01,02,03,04,07,08


Is that I'm not working well with Wheel Generator? or is that it doesn't make the cover calculations only in the set of the filtered lines ("good moves") but with all the 28?

Frank

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: The use of filters and the quantified cover

Post by lottoarchitect » Sat Nov 22, 2008 4:49 pm

Hi Frank,

It depends on what you try to achieve.

Let's start with the basic aspect. If you examine an 8,6,5,6 = 4 (100% 5if6) wheel and you request from this to include any 3 numbers (e.g. 1,2,3) to appear in all its tickets, you'll see that it is impossible to have this requirement fulfilled 100%, no matter which 3 different numbers you'll use.

01 02 03 04 05 07
01 02 03 04 05 08
01 03 04 06 07 08
02 03 05 06 07 08

Thus, you'll either have this wheel constructed as 100% and not fulfilling one ticket to the requirement of 3 numbers, or you'll have all the tickets fulfilling the requirement of 3 numbers but not the coverage. It is impossible to have both coverage and common filter fulfilled 100%
Now, the proportional mode you have used, because we have only 4 blocks and this is translated as improving from 3 passing tickets to 4 (25% improvement) is a better improvement than fulfilling the coverage requirement (from 3 to 0 -> 3/28 = 10.7% improvement), when having equal priorities. You either have to assign a higher priority for the coverage engine (i.e. set it to 100% and the common filter to e.g. 20%) so to affect the proportional improvements evaluation or switch to absolute mode (this is not really recommended when working with filters but will do the work in this particular case). WG correctly decided to leave the coverage to missing 3 and fulfilling the common requirement completely. It is a matter of settings and what you asked from the engine to do.

cheers
lottoarchitect

Frank
Regular
Regular
Posts: 34
Joined: Sat Nov 22, 2008 1:55 pm

Re: The use of filters and the quantified cover

Post by Frank » Sat Nov 22, 2008 6:42 pm

lottoarchitect wrote:Hi Frank,

It depends on what you try to achieve.
I want to achieve 5 hits when 6 of my 8 number set hit, and 3 of them are 01,02,03; let's suppose these will be the drawn numbers: 01,02,03,05,06,07

How many lines from the possible 28, cover the 5if6 prize? The answer is: two lines, for example 01,02,03,04,05,06 + 01,02,03,04,07,08

These two lines cover 100% the 10 lines that have 01,02,03 in common.

I suppose that WG is calculating the overall cover (how many 6 of 8 are covered?); when applying filters, WG could calculate the cover of only filtered lines.

Do you think this can happen in future?

Let's start with the basic aspect. If you examine an 8,6,5,6 = 4 (100% 5if6) wheel and you request from this to include any 3 numbers (e.g. 1,2,3) to appear in all its tickets, you'll see that it is impossible to have this requirement fulfilled 100%, no matter which 3 different numbers you'll use.

01 02 03 04 05 07
01 02 03 04 05 08
01 03 04 06 07 08
02 03 05 06 07 08

Thus, you'll either have this wheel constructed as 100% and not fulfilling one ticket to the requirement of 3 numbers, or you'll have all the tickets fulfilling the requirement of 3 numbers but not the coverage. It is impossible to have both coverage and common filter fulfilled 100%
Now, the proportional mode you have used, because we have only 4 blocks and this is translated as improving from 3 passing tickets to 4 (25% improvement) is a better improvement than fulfilling the coverage requirement (from 3 to 0 -> 3/28 = 10.7% improvement), when having equal priorities. You either have to assign a higher priority for the coverage engine (i.e. set it to 100% and the common filter to e.g. 20%) so to affect the proportional improvements evaluation or switch to ablolute mode (this is not really recommended when working with filters but will do the work in this particular case). WG correctly decided to leave the coverage to missing 3 and fulfilling the common requirement completely. It is a matter of settings and what you asked from the engine to do.

cheers
lottoarchitect
A good trick that I didn't know; now I can get a 8,6,5if6=4 100% with my filter; however in the 4 lines there are two lines that hit the 5if6 when 6 numbers of 8 are drawn and 3 of them are 01,02,03 and two more lines that are not needed to guarantee the prize I want, while doubling the probability of my winnings; in this wheel I play the double of lines it's needed to guarantee my aim!

I'm afraid that doesn't help players so sure of their filters.

Frank

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: The use of filters and the quantified cover

Post by lottoarchitect » Sat Nov 22, 2008 7:16 pm

Hi Frank,

first of all, the 8,6,5,6 requires minimum 4 lines to have this 100% guarantee. In your example of the two lines 01,02,03,04,05,06 + 01,02,03,04,07,08, if the actual draw is 03 04 05 06 07 08 then you do not have 5if6. WG calculates the overall cover of the wheel since we have to play the whole set of produced tickets to obtain the provided guarantee. Of course, if the actual draw drawn is 01,02,03,05,06,07, you'll have 5 correct numbers but this does not mean playing only those two tickets we have 5if6 100%. Actually with only two tickets, you get a c(8,6,5,6) 78.57% maximum for 5if6.
Apart from the above, I'm not sure I understand what you are after. The initial assumption you make of c(8,6,5,6) = 2 is invalid to make any comments on the following.
A good trick that I didn't know; now I can get a 8,6,5if6=4 100% with my filter; however in the 4 lines there are two lines that hit the 5if6 when 6 numbers of 8 are drawn and 3 of them are 01,02,03 and two more lines that are not needed to guarantee the prize I want, while doubling the probability of my winnings; in this wheel I play the double of lines it's needed to guarantee my aim!
since you need minimum 4 lines to ensure the guarantee, thus all 4 tickets are needed.
Perhaps you could provide a detailed example of what you want to achieve?

cheers
lottoarchitect

Frank
Regular
Regular
Posts: 34
Joined: Sat Nov 22, 2008 1:55 pm

Re: The use of filters and the quantified cover

Post by Frank » Sat Nov 22, 2008 8:43 pm

LA,

the example I gave is good enough.

There are 28 possible lines in C(8,6) as you know; but only 10 of them have 01,02,03 inside:

01,02,03,04,05,06 (6)
01,02,03,04,05,07 (5)
01,02,03,04,05,08 (5)
01,02,03,04,06,07 (5)
01,02,03,04,06,08 (5)
01,02,03,04,07,08 (6)
01,02,03,05,06,07 (5)
01,02,03,05,06,08 (5)
01,02,03,05,07,08 (5)
01,02,03,06,07,08 (5)


That means, these are the "good moves", the lines that fullfill the filter criteria : at least 3 of the group 01,02,03; in other words, regarding this filter, they are the lines that must be covered 100%, 5if6.

If you check now, you'll see that 01,02,03,04,05,06 + 01,02,03,04,07,08 hit at least 5 hits in 8 of those 10 lines and 6 hits in the remaining 2.

I'm not saying that they cover every 28 lines of C(8,6); they cover 5if6 when, in the 6 drawn numbers, 3 of them are 01,02,03 and the other 3 are inside 04,05,06,07,08.

Please, let me know if I'm not explaining myself well.

Frank

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: The use of filters and the quantified cover

Post by lottoarchitect » Sun Nov 23, 2008 12:20 am

Hi Frank,

this example clearly describes what you are after. However, the term 'bad move' is equally applied to the coverage as well. This cannot be done in this version because WG picks from the whole set of 28 combinations, not only those 10 defined by the common filter. The reason is simple: If we use only these 10 combinations, there will be a big missing gap for the coverage aspect and since WG aims to improve both, it will pick from the remaining 28-10=18 combinations as well, since the improvement criteria suggest that this would be beneficial for the overall covering both for the common filter and the coverage of the wheel. In short, you want to use the common filter as a constructor for the initial list of combinations the engine will be allowed to pick from and not as it is designed to operate within the whole WG's idea. It is similar to obtaining a set of combinations an letting the system pick among them (in this example your 10 combinations) and try to find those that offer the best improvement to any other filters/coverage you have set. I have plans adding such a feature.

cheers
lottoarchitect

Frank
Regular
Regular
Posts: 34
Joined: Sat Nov 22, 2008 1:55 pm

Re: The use of filters and the quantified cover

Post by Frank » Sun Nov 23, 2008 2:53 pm

:D
For those players that use filters, that feature will be useful.

WG can be the first to make that, regarding 3if3, 4if4, 4if5 and 5if5.

Till now, I know some softwares that allow filters and calculate the specific cover, but only for 3if6, 4if6 and 5if6.

:D

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: The use of filters and the quantified cover

Post by lottoarchitect » Sat Nov 29, 2008 4:48 pm

Hi Frank,

can you post examples of real achievable coverage using your suggestion with the software you use to do this? Please don't forget to mention the filter and its parameters you have used as a guide. I have added this feature in WG and I want to test its efficiency in coverage achievement.

cheers
lottoarchitect

Frank
Regular
Regular
Posts: 34
Joined: Sat Nov 22, 2008 1:55 pm

Re: The use of filters and the quantified cover

Post by Frank » Sat Nov 29, 2008 7:43 pm

Yes I can.

Right now I'm trying the 18,6,4,6 .

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: The use of filters and the quantified cover

Post by lottoarchitect » Sat Nov 29, 2008 7:49 pm

Ok, I'm waiting for your results. Also I'd like other wheels too as reference. Pick any sort of wheels you want. The more real data I have the better.

cheers
lottoarchitect

Post Reply

Who is online

Users browsing this forum: No registered users and 9 guests