File Simulation

[Joe]

Hi all,

I need more explanation about how to work with WN History->File Simulation.
What is Back-test Tickets in this page?.
Whats the need of it?.

[sliver]

Yeah Stan,
I would also like to know more about the File Simulation feature, how does this work and what is it used for ? Looks interesting...

Sliver

[stan]

the file simulation is just an experiment. it is similar to 'simulation back-test' except that it tests tickets from a specified files instead of selected past draws.

here's how it works:
* first you have to select winning numbers range in the selection panel and press 'update' button to recalculate wn history values
* then select simulation tickets (same as in 'simulation back-test'). i recommend using rand5if6 file (or an equivalent) that is mentioned in our lottery strategy in the jackpot section
* then select a file with tickets that will be tested, the file is limited to 1000 tickets only
* click 'start simulation' and expert lotto will calculate new wn history sums for each ticket from the back-test tickets file as if the ticket was drawn after the latest draw in your selection of winning numbers
* +/-20 is added to the new sums (or any other value specified in the file simulation tab)
* the simulation tickets (i.e. from the package or from the generator) are filtered through these ranges and number occurrence statistics are calculated for tickets that passed
* number stats are sorted in descending order and numbers from the back-test ticket are compared with the stats. the summary results table show how many ticket numbers were in the first 10, 15, 20, ... sorted numbers
* the details table shows the number occurrences, back-test ticket numbers are highlighted with red background

basically, i created this new tab to verify our (expert lotto team) strategy - if we applied the correct wnh ranges then the filtered tickets should have several numbers in the first 20-25 sorted numbers....

[Red Devil]

Stan,

I have never been able to verify this strategy in the UK Lotto, not sure if I am doing anything wrong but have followed all instructions given by both you and Bobes.

In this file simulation for the file to be tested could you use the last 500 or so winning draws?

[stan]

[quote=Red Devil]
In this file simulation for the file to be tested could you use the last 500 or so winning draws?
[/quote]

yes, that's fine. in fact you should use all draws to have the complete wn history in the calculation

[Red Devil]

Stan,

Ok, I'll try again

[stan]

[quote=Red Devil]
I have never been able to verify this strategy in the UK Lotto, not sure if I am doing anything wrong but have followed all instructions given by both you and Bobes.
[/quote]

i'll try to find some free 'screen camera' software to record a 'movie' to better illustrate what steps need to be done to use our strategy.
if anybody knows some good and free software for screen recording, pls let me know

[george]

Stan you can use CamStudio from camstudio.org

[stan]

[quote=george]
Stan you can use CamStudio from camstudio.org
[/quote]

thanks for the tip, it looks good

[Bobes]

Due to our open policy the "File Simulation" function is available to everybody. So we owe an explanation. As you remember I recommend certain strategy couple of weeks ago. The final result of this strategy is sorted line of numbers according to their frequency of occurrence. When you use condition "at least 9 to 11 column must pass" and the margins were approximately correct then usually the first two numbers are the expected one. In addition the (expected) drawn numbers show certain pattern. Now the purpose of the File SImulation function is to use file remaining in the package after WNH filtration (or after any combined filtration) to identify those tickets, which are the best concerning how they meet certain pattern after simulation . To use it you must at first create such file in package using any out of available filters (according to your strategy). Then save the file under suitable name and load the file commonly used for simulation into the package. Load the tested file to the Back-Test field. Start simulation. Then the function takes tickets (step by step) from the saved file and performs simulation for each particular ticket. Such way you can compare sorted lines of numbers (frequency of occurence) for many tickets in one step without need to simulate them manually. However I recommend to use this operation for small files only when you use full wheel of possible tickets.

This way you can also use limited pool of numbers. For example: Create full wheel from the first twenty numbers of the appropriate sorted line to the package and then compare the tickets against such package. You must keep in mind that in such case could be better to use more strict conditions for simulation (closer distance between margins and more strict condition "at least 10 (11) to 11 must pass").