Filter Normal Distribution (FND)

[sysp34]

Greeting EL members

i just found some idea after helping neighbour with statistic for her Bachelor Thesis, maybe it already discuss earlier. it's about normal distribution with significant level / alpha 5%.



Thanks to ImageShack for Free Image Hosting

for example FL 5-36 have filter normal distribution (sum of numbers) between sum 58 to 127 an statistics package shown sum 54 to 131 (using mean square error)

so how it use for lottery? well Stan, Joe and anyone who talented java programming maybe we can add some syntax sequence to filter out "filter normal distribution" i mean it applied for any filter that had filter normal distribution

i dunno if this will to get the jackpot because if we add more filters then it will reduce the probabilities due to the multipliers

stan, it would be good tool if EL have default setting for filter normal distribution, i mean for any lotteries EL will create FND by default. then EL user reduced with favorite numbers or numbers that look good to play

gl

[kdsjeter]

[quote=sysp34]
Greeting EL members

i just found some idea after helping neighbour with statistic for her Bachelor Thesis, maybe it already discuss earlier. it's about normal distribution with significant level / alpha 5%.



Thanks to ImageShack for Free Image Hosting

for example FL 5-36 have filter normal distribution (sum of numbers) between sum 58 to 127 an statistics package shown sum 54 to 131 (using mean square error)

so how it use for lottery? well Stan, Joe and anyone who talented java programming maybe we can add some syntax sequence to filter out "filter normal distribution" i mean it applied for any filter that had filter normal distribution

i dunno if this will to get the jackpot because if we add more filters then it will reduce the probabilities due to the multipliers

stan, it would be good tool if EL have default setting for filter normal distribution, i mean for any lotteries EL will create FND by default. then EL user reduced with favorite numbers or numbers that look good to play

gl
[/quote]

Sysp34,

Why not use the sum filter which already exists?

From my memory of statistics, in a normal distribution, most sums will occur (95% or more) will occur within two standard deviations of the mean.

For example for my game Bonus Match 5, the mean sum is 100. The standard deviation of the distribution over 2,174 drawing remains 24.

Two deviations from below the mean is 52, and two deviations above the mean is 148. From this sum range of 52 - 148, 2,084 occurences from a total possible of 2,174 to date have occurred, or 95.86%.

One deviation from the mean or, 76 - 124, produces 1,454 sums out of 2,174 drawings or, 66.88% of the time, or 1 in 1.495 draws.

Not a bad filter.

[stan]

[quote=sysp34:1200765638]
stan, it would be good tool if EL have default setting for filter normal distribution, i mean for any lotteries EL will create FND by default. then EL user reduced with favorite numbers or numbers that look good to play
[/quote]

it's been discussed before. the problem is i'd have iterate all possible combinations to find out what the normal distribution is and that's not very practical, especially for keno lotteries

[sysp34]

[quote=kdsjeter:1200767331]
Why not use the sum filter which already exists?

Not a bad filter.
[/quote]

kdsjeter, thank for replying and i agree the sum of number good enough to reduced the JP numbers, although it little bit tricky using skip and hit methods.

well using a normal distribution would be an advantages because 95% the next JP would be there with +/- 250K JP numbers.

[quote=stan]
it's been discussed before. the problem is i'd have iterate all possible combinations to find out what the normal distribution is and that's not very practical, especially for keno lotteries

[/quote]

yup i know it not practically using the normal distribution, and i mean is EL provide some compound filters and let user to choose several compound filter that have a normal distribution.

btw how about creating a groups pattern? although creating that pattern would be :wand: because creating a looping condition based on selected group. for example a unique 18 groups for 5/36 game created 8568 pattern unfortunately i can't created a double pattern due to limited knowledge using a macro in spreadsheet :wand: :vogel:

but the good new is FL 5/36 will have a new pattern for next JP with +/- 225K #'s and :wand: :wand: :wand: i'm not carry the files while i come to a friend home that have a dsl connection.

maybe later i post the information.

thank and good luck

[stan]

maybe joe will be able to help you out with your pattern filters...

[sysp34]

stan, joe or el user who had java knowledge

here is some idea with group patterns


well i hope it an enhanced group filters

good luck

[CARBOB]

[quote=sysp34]
stan, joe or el user who had java knowledge

here is some idea with group patterns


well i hope it an enhanced group filters

good luck
[/quote]

Ragwi, could you give a little more detail on your group patterns theory?
Thanks.

[sysp34]

[quote=CARBOB]

Ragwi, could you give a little more detail on your group patterns theory?
Thanks.
[/quote]

Senor Bob

sorry for not having a crystal clear explanation, here is a little bit explanation



i hope it little bit help about number group

gl