3 possible bugs in last development build

[Gatsby]

Hi Stan,

Working around with filters related to number movement and highest match, i´ve found the situations in the file attached that seems to be bugs.

Regards,

Three possible bugs.docx

Filipe.

Three possible bugs2.doc

[stan]

1. what 'highest match' combination are you referring to?

2. if you choose 'latest 1' then the filter is looking for tickets that have movement pattern '+++--' when compared with the latest draw '15-25-38-40-41'. but there is no such combination of numbers to fit that movement pattern.

3. same as above. if the latest pattern is '+++--' for latest draw '15-25-38-40-41' then there's no possible combination of fit this pattern. no last digit can be lower than zero

[Gatsby]

Hi Stan,

1) See file attached

2) True, this is a awkward combination to give as an example, because applying the movement would give for instance: 16-26-39-39-40 ...

3) As for "Last digit ..." A combination like 26-27-28-39-40 should pass. In this case 9 is lower than 0. It´s a circular pattern. Don´t you think it should filter this combination?

Then again, maybe this ( 16-26-39-39-40 ) could be solved by originating the combination "16-26-29-39-40". A combination that matches the pattern of movement. This way there will allways be combinations with the same pattern to remove.

[stan]

Hi Stan,

1) See file attached

winning match definitions are part of the lottery setup. if the rules of euro millions lottery have changed i need to adjust the lottery settings then...

3) As for "Last digit ..." A combination like 26-27-28-39-40 should pass. In this case 9 is lower than 0. It´s a circular pattern. Don´t you think it should filter this combination?

[/quote]
my implementation of this filter and stats doesn't apply any circularity. if it was a circular pattern than 8 (7,6,5,4,3,2,1) would be lower than 0 as well...

[Gatsby]

Hi Stan,

I think that in this case it shoul allow circularity. Can´t you implement a sub rotine that allows that combinations that arise in a non organized way (like 49-25-23-12-1) to be organized (1-12-23-25-49) and then filtered?

[stan]

Hi Stan,

I think that in this case it shoul allow circularity.

anyone else agrees this should the expected behavior?

[Jerzy]

Probably you can find a kind of circularity in the full package tickets, but there is no circularity at all in lotto drawings results and number movements.

The idea of circularity of something in the filter was not presented in a clear and convincing way and I do not know what it is about. I do not know if the filter with circularity will be worth using or not.

[Gatsby]

Hi Stan and Jerzy.

for sometime i´ve been advocating the use of circularity, or if you want, circular patterns. This is the case for a permutations builder, but it applies also in this filter. If you look at a pattern like +++-- and aplply it to the last draw you allways will have combinations to remove, but it must permite circularity and rearranjement of those combinations:


Example:

Last draw is: 1, 13, 22, 24, 45

applying the movement (+++--) will give:

2, 14, 23, 25, 46
3, 15, 23, 26, 47
4, 16, 24, 27, 48,
5, 17, 25, 28, 49
6, 18, 26, 29, 50

but also:

7, 19, 27, 30, 1 or what is the same: 1, 7, 19, 27, 30 *
8, 20, 28, 31, 2 .......................: 2, 8, 20, 28, 31 *

These are combinations that are worth to remove (that is 5, 4 or 3 numbers from each of them), as the rule icontinues the same. If you try this transformations and applied to your last draw and look at the next, you will see that normally only 1 or 2 numbers (exceptionally 3) appear in that next draw. Mind that this also is observable for the combinations identified by * ... as if they were part of the same family of transformations, so i think they should be considered.

Sorry if english isn´t very good or to my difficulty to express an idea in a more clear way. I think, though, this is something to explore. A filter that looks at the movement shouldn´t consider this? Isn´t this logical?

Regards,

F.

[Jerzy]

Gatsby,

I like the latest Number Movement Patterns filter. It is simple but powerful. But your new idea is also very attractive.

It seems to me that you want an entirely different number movement filter based on an entirely different principle.

The starting point to the number movement would be the last draw result or any combination selected by the user. The numbers would move from the starting point according to +++++ or ++++++ pattern, depending on the lottery, forming combinations until they return to the starting point and then stop. Some positional number rearrangement would be necessary. The count of combinations in the circle which would be filtered, including the starting combination = total numbers in the pool. The filter would identify these combinations and would accept/reject them from the package according to the user's instruction. And the instruction would be "from x to y positions must pass".

Your idea is worth implementing. Thank you for your explanations.

Regards

Jerzy

[Jerzy]

Stan,

The latest Number Movement Patterns filter is simple and powerful. There is no point in changing it.

Gatsby has an idea of building a different number movement filter and it should be implemented. The number movement would be trigerred by the last draw combination or, if possible, by any combination selected by the user. The numbers would move according to +++++ or ++++++ pattern forming combinations until the starting combination is reached. These combinations would be identified and processed according to the user's instruction.

I am looking forward to using such filter.

Regards

Jerzy

[Gatsby]

Hi Stan,

I´m glad Jerzy agrees on my approach, but please let me clarify on this:

"The latest Number Movement Patterns filter is simple and powerful. There is no point in changing it." - I think it does. The same ideia of circularity is present in the last number movement pattern. If it´s not implemented then not all combinations are generated and the filter won´t be working at it´s full potencial. (BTW, the same for the number movement filter).

"It seems to me that you want an entirely different number movement filter based on an entirely different principle" - yes, but what i´m aiming here it´s not a new number movement filter (which exists, though it doesn´t contemplate circularity) and adresses the same movement in each position as the movement observed from the two last draws or other that we choose (it simply doesn´t generate combinations when some of the numbers are at the extremes of the number pool...). What i´m aiming here is a Permutations Builder. It will do as Jerzy put it "The count of combinations in the circle which would be filtered, including the starting combination = total numbers in the pool. The filter would identify these combinations and would accept/reject them from the package according to the user's instruction. And the instruction would be "from x to y positions must pass"." This Permutations Builder will allow any transformation you think of. If not, you would have to implement for each transformation a new filter Stan. A Number movement filter for "+++++", for "+=+=+", for "=-=-=". Understand that the case Jerzy refers is a particular case where as he well put it "The numbers would move from the starting point according to +++++ or ++++++ pattern, depending on the lottery, forming combinations until they return to the starting point and then stop. Some positional number rearrangement would be necessary. The count of combinations in the circle which would be filtered, including the starting combination = total numbers in the pool".

Number movement filters (NMF) as they are are fine, only need to be adjusted to generate all possible combinations allowed by circularity. And mind that this change wouldn´t in any way distorce the use of those filters, would only permit to filtrate more combinations.

Finally, NMF it´s not the same as a Permutations Builder, which would give for each rule established in a 50 pool game 50 new lines with combinations to filter, just as Jerzy put it : it "would identify these combinations and would accept/reject them from the package according to the user's instruction. And the instruction would be "from x to y positions must pass" (that is, only accept combinations in the package that share 1 to x (normally 3) numbers with the combinations generated by the builder).

Best Regards,

F.

[teoman]

Hi Stan and Jerzy.

for sometime i´ve been advocating the use of circularity, or if you want, circular patterns. This is the case for a permutations builder, but it applies also in this filter. If you look at a pattern like +++-- and aplply it to the last draw you allways will have combinations to remove, but it must permite circularity and rearranjement of those combinations:


Example:

Last draw is: 1, 13, 22, 24, 45

applying the movement (+++--) will give:

2, 14, 23, 25, 46
3, 15, 23, 26, 47
4, 16, 24, 27, 48,
5, 17, 25, 28, 49
6, 18, 26, 29, 50

but also:

7, 19, 27, 30, 1 or what is the same: 1, 7, 19, 27, 30 *
8, 20, 28, 31, 2 .......................: 2, 8, 20, 28, 31 *

These are combinations that are worth to remove (that is 5, 4 or 3 numbers from each of them), as the rule icontinues the same. If you try this transformations and applied to your last draw and look at the next, you will see that normally only 1 or 2 numbers (exceptionally 3) appear in that next draw. Mind that this also is observable for the combinations identified by * ... as if they were part of the same family of transformations, so i think they should be considered.

Sorry if english isn´t very good or to my difficulty to express an idea in a more clear way. I think, though, this is something to explore. A filter that looks at the movement shouldn´t consider this? Isn´t this logical?

Regards,

F.

If after your example (1, 13, 22, 24, 45) is to propose a new draw (3, 10, 23, 29, 41) then we have:

47, 09, 18, 20, 41 => 1 number
48, 10, 19, 21, 42 => 1 number
49, 11, 20, 22, 43 => 0
50, 12, 21, 23, 44 => 1 number
01, 13, 22, 24, 45
02, 14, 23, 25, 46 => 1 number
03, 15, 24, 26, 47 => 1 number
04, 16, 25, 27, 48 => 0
05, 17, 26, 28, 49 => 0
06, 18, ​​27, 29, 50 => 1 number

where we have tickets to 0, 1 numbers of proposed draw, while filters Adjacent Numbers and Knight's Move will choose even itself draw 3, 10, 23, 29, 41. In conclusion, your proposal is not more efficient than existing current filters in EL 5.

[Gatsby]

Hi Teoman,

In first place let start by apologizing how i started posting in a post started by you in "Adjacents". It wasn´t my intention to divert that discussion or to break your line of reasoning. It just seemed appropriated to give my opinion.

Maybe you didn´t like what i posted above that Stan would have to create a new filter for each movement that someone imagines, i will let Stan be the judge on that.

About your example, though i didn´t see where you´ve gone to present combination 47, 09, 18, 20, 41 i can see the logic. I can´t agree with you, though, on the question of efficiency. If by that you mean that only o or 1 commun number appear, i remember that i said normally o, 1 or 2 are commun to the transformed combination (of which adjacents and knight move are examples). In some draws the transformation will give 0 or 1 number, in others it will give 3 numbers (very exceptionally), but we do not expect that draw after draw it will give 2 numbers or even 1, normally it will give sequences of 0 or 1. In a further apllication of this filter (builder), it even could come to reject all the numbers if we have a sequence of several 1´s, but this would have to be an kind of "Transformations History filter...." that i guess Stan isn´t quite interested in implementing, considering the work it may involve, and because we are just coming around with the ideia of a Permutations Builder.

Well, that is, i hope you don´t think i was by some way dismissing your idea, just trying to look at it from another perspective. In fact i´m glad you put the question about adjacents, because that gave me a opportunity to present by another light this question about permutations.

Best Regards,

F.

[teoman]

Hi Teoman,

In first place let start by apologizing how i started posting in a post started by you in "Adjacents". It wasn´t my intention to divert that discussion or to break your line of reasoning. It just seemed appropriated to give my opinion.

Maybe you didn´t like what i posted above that Stan would have to create a new filter for each movement that someone imagines, i will let Stan be the judge on that.

About your example, though i didn´t see where you´ve gone to present combination 47, 09, 18, 20, 41 i can see the logic. I can´t agree with you, though, on the question of efficiency. If by that you mean that only o or 1 commun number appear, i remember that i said normally o, 1 or 2 are commun to the transformed combination (of which adjacents and knight move are examples). In some draws the transformation will give 0 or 1 number, in others it will give 3 numbers (very exceptionally), but we do not expect that draw after draw it will give 2 numbers or even 1, normally it will give sequences of 0 or 1. In a further apllication of this filter (builder), it even could come to reject all the numbers if we have a sequence of several 1´s, but this would have to be an kind of "Transformations History filter...." that i guess Stan isn´t quite interested in implementing, considering the work it may involve, and because we are just coming around with the ideia of a Permutations Builder.

Well, that is, i hope you don´t think i was by some way dismissing your idea, just trying to look at it from another perspective. In fact i´m glad you put the question about adjacents, because that gave me a opportunity to present by another light this question about permutations.

Best Regards,

F.

Stan.decide what will make.
Permutations means:

01 => 01, 13, 22, 24, 45
02 => 01, 13, 22, 45, 24
03 => 01, 13, 24, 22, 45
04 => 01, 13, 24, 45, 22
05 => 01, 13, 45, 22, 24
06 => 01, 13, 45, 24, 22
07 => 01, 22, 13, 24, 45
08 => 01, 22, 13, 45, 24
09 => 01, 22, 24, 13, 45
10 => 01, 22, 24, 45, 13
11 => 01, 24, 13, 22, 45
12 => 01, 24, 13, 45, 22
"
"
120=> 45, 24, 22, 13, 01

and what you explain would be the number movement or circular movement.

[Gatsby]

No argue about this, i guess Teoman interpreted permutations well as all possible changes of place with 5 different numbers. And all that combinations are still the same draw.

I´ve checked the definition about Permutations in Wiki, and it says that "in In mathematics, the notion of permutation is used with several slightly different meanings, all related to the act of permuting (rearranging in an ordered fashion) objects or values. Informally, a permutation of a set of objects is an arrangement of those objects into a particular order. For example, there are six permutations of the set {1,2,3}, namely [1,2,3], [1,3,2], [2,1,3], [2,3,1], [3,1,2], and [3,2,1]. One might define an anagram of a word as a permutation of its letters. The study of permutations in this sense generally belongs to the field of combinatorics.

So i guess i´m addressing to a particular case of permutations... . I´m adressing permutations not of the numbers themselves, but of their structure, their group (their symmetric group to be more exact).
In this sense, 1, 3, 5, 7, 9 is the same as 2, 4, 6, 8, 10 (the subjacent group is that the diferences between all of them is 2, or 2-2-2-2, or "adjacents" +2).

In some of my previous posts, conscient of this confusion i tried to use also the term "Transformation". Wether it should be called Permutations Builder is in fact dubious, as Teoman posted, specially for those that have some mathematic notion, maybe Stan as a better idea.

Best regards,

F.