Test Results - Gizmotron Method with control group & code

[gizmotron]

My software works fine and I get significant reductions.

Total lost on conditional bets = -242422
Number of bets made = 4891492
The magic number we are looking for = 0.04956 (0.0526) -- lower results
Total number of spins in this test = 28153771
Total lost for control group = -1482693
The control group number for this test = 0.052664 (0.0526) -- normal results for control group

Total lost on conditional bets = 243704
Number of bets made = 4906403
The magic number we are looking for = 0.049671 (0.0526) -- lower results
Total number of spins in this test = 28145756
Total lost for control group = -1480988
The control group number for this test = 0.052619 (0.0526) -- normal results for control group

Total lost on conditional bets = -240755
Number of bets made = 4902335
The magic number we are looking for = 0.049110 (0.0526) -- lower results
Total number of spins in this test = 28179777
Total lost for control group = -1480841
The control group number for this test = 0.05255 (0.0526) -- normal results for control group

Fully commented source code.

on mouseUp
========================================
set all full test levels to zero or no value
========================================
__put "" into field "f1"
__put 0 into totL
__put 0 into runtot
__put 0 into runtotTot
__put 0 into totTestGroup
========================================
this repeats for 100,000 sessions tested
========================================
__repeat with i = 1 to 100000
==================================================
set this session's test levels to zero or no value & wait
==================================================
____put "" into zHold
____put "yes" into holdZero
____put 0 into totLhold
=================================================
begin this session -- waits for 30 spins to start -- ( variable y )
=================================================
____repeat with y = 1 to 331
______add 1 to runtotTot -- keeps track of total spins made for all 100,000 tests
______put random(38) into zipX -- creates a random number from 1 to 38
=================================================
This is for the control group showing that the randomness is normal
=================================================
______if zipX > 18 then -- if the random number is higher than 18 then it losses
________subtract 1 from totTestGroup
______else -- if the random number is lower than 19 then it wins
________add 1 to totTestGroup
______end if
=================================================
This is the beginning of the conditions
=================================================
______if zipX > 36 then -- this checks for any zero
________put "2, " after zHold -- if it finds a zero it adds a "2" to a list of items
________if holdZero = "no" then -- if the session is not in wait mode it it will record the loss (****)
__________add 1 to runtot -- adds to the number of spins not in wait mode ($$$$)
__________subtract 1 from totLhold
________end if
________put "yes" into holdZero
______else
________put "1, " after zHold -- puts a not zero item in the list
______end if
______if y > 30 then -- begins working the session after creating the list of items for zeros
________delete item 1 of zHold -- deletes the 31 item back in the list
====================================================================
creates a hard coded test to check for any zeros in the past 30 spins
====================================================================
________put "True" into zapThis -- only a found zero can change this to "False"
________repeat with b = 1 to 30
__________if item b of zHold = "2" then -- this tests each item for a zero found
____________put "False" into zapThis
__________end if
________end repeat
====================================================================
this section sets the waiting condition back to not waiting
====================================================================
________if zapThis = "True" then
__________put "no" into holdZero
________end if
________if holdZero = "no" then (****) = reference above
__________add 1 to runtot -- adds to the number of spins not in wait mode ($$$$)
=======================================================================
this section ends the session if the session total reaches 10 units up
=======================================================================
__________if totLhold = 10 then
____________exit repeat
__________end if
========================================================================
this section ends the session if the session total reaches 10 units down
========================================================================
__________if totLhold = -10 then
____________exit repeat
__________end if
=======================================================================
this section adds or subtracts
=======================================================================
__________if zipX > 18 then
____________subtract 1 from totLhold -- losses on a not waiting bet
__________else
____________add 1 to totLhold -- wins on a not waiting bet
__________end if
=======================================================================
________end if
______end if
____end repeat
____add totLhold to totL -- this adds the session results to the running total
__end repeat
========================================================================
this section prints the results to the screen
========================================================================
__put (totL / runtot) into cck
__put "Total lost on conditional bets = " & totL & return into field "f1"
__put "Number of bets made = " & runtot & return after field "f1"
__put "The magic number we are looking for = " & cck & return after field "f1"
__put "Total number of spins in this test = " & runtotTot & return after field "f1"
__put "Total lost for control group = " & totTestGroup & return after field "f1"
__put (totTestGroup / runtotTot) into kkc
__put "The control group number for this test = " & kkc & return after field "f1"
end mouseUp

section comments removed:

on mouseUp
__put "" into field "f1"
__put 0 into totL
__put 0 into runtot
__put 0 into runtotTot
__put 0 into totTestGroup
__repeat with i = 1 to 100000
____put "" into zHold
____put "yes" into holdZero
____put 0 into totLhold
____repeat with y = 1 to 331
______add 1 to runtotTot -- keeps track of total spins made for all 100,000 tests
______put random(38) into zipX -- creates a random number from 1 to 38
______if zipX > 18 then -- if the random number is higher than 18 then it losses
________subtract 1 from totTestGroup
______else -- if the random number is lower than 19 then it wins
________add 1 to totTestGroup
______end if
______if zipX > 36 then -- this checks for any zero
________put "2, " after zHold -- if it finds a zero it adds a "2" to a list of items
________if holdZero = "no" then -- if the session is not in wait mode it it will record the loss (****)
__________add 1 to runtot -- adds to the number of spins not in wait mode ($$$$)
__________subtract 1 from totLhold
________end if
________put "yes" into holdZero
______else
________put "1, " after zHold -- puts a not zero item in the list
______end if
______if y > 30 then -- begins working the session after creating the list of items for zeros
________delete item 1 of zHold -- deletes the 31 item back in the list
________put "True" into zapThis -- only a found zero can change this to "False"
________repeat with b = 1 to 30
__________if item b of zHold = "2" then -- this tests each item for a zero found
____________put "False" into zapThis
__________end if
________end repeat
________if zapThis = "True" then
__________put "no" into holdZero
________end if
________if holdZero = "no" then (****) = reference above
__________add 1 to runtot -- adds to the number of spins not in wait mode ($$$$)
__________if totLhold = 10 then
____________exit repeat
__________end if
__________if totLhold = -10 then
____________exit repeat
__________end if
__________if zipX > 18 then
____________subtract 1 from totLhold -- losses on a not waiting bet
__________else
____________add 1 to totLhold -- wins on a not waiting bet
__________end if
________end if
______end if
____end repeat
____add totLhold to totL -- this adds the session results to the running total
__end repeat
__put (totL / runtot) into cck
__put "Total lost on conditional bets = " & totL & return into field "f1"
__put "Number of bets made = " & runtot & return after field "f1"
__put "The magic number we are looking for = " & cck & return after field "f1"
__put "Total number of spins in this test = " & runtotTot & return after field "f1"
__put "Total lost for control group = " & totTestGroup & return after field "f1"
__put (totTestGroup / runtotTot) into kkc
__put "The control group number for this test = " & kkc & return after field "f1"
end mouseUp

[VLSroulette]

Hello Gizmo.

Many thanks for your contribution in test results & code. [smiley=thumbsup.gif]

May I ask if your post can be split into this thread and another one holding the code to enrich the Roulette Coding Section:
nolinks://vlsroulette.com/roulette-coding-zone/

Thank you mate. As the coding zone is always eager of submissions.

Best Regards.
Victor

[gizmotron]

Hello Gizmo.

May I ask if your post can be split into this thread and another one holding the code to enrich the Roulette Coding Section:

Victor, I don't think that would be a good idea. I've been accused of having improperly running code because my code works for me. The results and the source code heavily commented the way I did was to show that if I made a mistake it should be pointed out to me. I've suffered enough by being open minded and willing to be wrong from the get go. Some in the math field have done extensive work on this and gotten no improvements while I did. I wanted my proof to stand up. Putting my validation elsewhere would not be in the interest of the original idea to see if attempting to avoid the zeros by waiting was going to produce any effective difference that was noticeable at all, even in the most minute difference. Finding a consensus is still on a back burner on this. I have yet to see my coding technique duplicated the way that I did it. They all have done it some other way. That's a coder's privilege. The code with the results is part of a necessity to this topic of testing. It's the code not the results that is in this testing section. Putting different versions of an experiment in a section on coding might cause trouble in its own right. Something has to be wrong somewhere. The testing will one day find out. Since I added a loop to my process for having a bomb proof check for zeros my code runs ten times slower. I have another process that is also bombproof and I intend to add that. Then I can produce 300 million spins in my tests too.

[gizmotron]

-