Colin,
Your missing my main point. There is a vast difference between pseudo-random and pure random.
http://random.mat.sbg.ac.at/~ste/dipl/node10.html
BUt I will leave that argument up to the math heads. My point was given the fact that your RNG is actually a PRNG, and your prediction engine was performing better than random expectation, then you can deduce that your engine appears to indicate that any lottery tested (which is assumed to be purely random) is similar to a PRNG. You even state that you believe the same...
"The fact is this method can be emulated by a Computer..."
That was the point I was trying to make, which would confirm my suspicions in this thread...
https://www.lotterypost.com/thread/70915/10
And it seems you are misquoting me somehow, I find that I said in that thread
"The balls are supposed to be relatively IDENTICAL, thus no bias exists in which ones will be drawn."
https://www.lotterypost.com/thread/70915/3
And I also said
- The balls are all the SAME size and weight
- The balls are always loaded in the same order
- The balls are always drawn from the same location
- The machine is always the same configuration/design
- The machine always runs from the same power source
- The balls are always mixed via the same method
Where again did I ever say something about minute imperfections in balls? The point I have always tried to point out, is the belief that a lottery machine is biased, just like a PSEUDO-RANDOM number generator. It has every appearance of being random, but in fact is not.
Here, let me give you a case in point (and this is funny too, because it deals with SQL, ACCESS and databases...) I use the following code inside MSAccess to generate random Alpha numeric strings (6 characters in length).
Public Function strScrambled() As String
Dim intUpperbound%, intLowerbound%, intChar%, strHold$, intPick%, strFinal$
intPick = 0
Do
'RetryPick:
'randomize the random number generator seed
Randomize ((intChar * Timer) / Rnd)
intPick = Int(intPick * Rnd + 1)
'here we will select cap or num
Select Case intPick
Case Is < 6
intLowerbound = 65
Case Is > 5, Is < 9
intLowerbound = 0
Case Is < 16, Is > 8
intLowerbound = 65
Case Is > 15, Is < 19
intLowerbound = 0
Case Is < 29, Is > 18
intLowerbound = 65
Case Is > 28
intLowerbound = 0
Case Else
intLowerbound = 65
End Select
'now that we know what we've picked
'we can set the rest of what we need
'for the selection routine
If intLowerbound > 10 Then
intUpperbound = intLowerbound + 25
Else
intUpperbound = 9
End If
'here we return a number based on the UPPERBOUND and LOWERBOUND
'Lower = 0 and Upper = 9 for Integers
'Lower = 65 and Upper = 90 for Uppercase
Randomize (intChar * Timer)
intChar = Int((intUpperbound - intLowerbound + 1) * Rnd + intLowerbound)
If intChar < 10 Then
strHold$ = strHold$ & intChar 'just tack the Character on the end
Else
strHold$ = strHold$ & Chr$(intChar) 'just tack the converted Character on the end
End If
intPick = 30
Loop While Len(strHold$) < 100 'Loop until we are done
strFinal$ = Left(strHold$, 1)
Do
Randomize ((Rnd * Timer) / (intChar + 1))
intPick = Int(95 * Rnd + 1)
intChar% = Int(3 * Rnd + 1)
strFinal$ = strFinal$ & Mid(strHold$, intPick, intChar%)
Loop Until Len(strFinal$) >= 6
If Len(strFinal$) > 6 Then
strFinal$ = Left(strFinal$, 6)
End If
strScrambled = strFinal$
End Function
This code will always start with an Alpha character, then will follow with a blend of 5 other alpha/numeric combinations.
Since each character location is unique, the total number of combinations can be calculated as such...
1st place holder (26 choices), 2nd-5th place holders (36 choices).
26 x 36 = 936
x 36 = 33,696
x 36 = 1,213,056
x 36 = 43,670,016
x 36 = 1,572,120,576
That yields just over 1.5 billion combinations. Now my dillema. I am populating a table using a text field as a Primary key, which I need to generate unique keys with the above attributes (first character is alpha, the remaining alpha/numeric with a total of 6 characters). How many repeats should I have had while generating 40,000+ keys out of 1.5 billion possible? Well over 4,0000 repeats occured during the generation of records and Keys! Seems like that calculates out to every 1 in 40 was a repeat, out of 1.5 billion possible? When was the last time a 6 ball lottery repeated itself?
And lastly, why are you bent on beating me over the head with SQL? I use it all the time, but I didnt want to burden WINHunter with the extra overhead of the MSJet database engine. WINhunter would generate hundreds if not thousands of queries in a few hours, and somehow you think using a database engine will speed it up? I think not. Besides that, not all lottery histories are formatted the same. You see, WINHunter has the most simplistic history interface, you simply use what the lottery corps provide. You are always guarenteed one fact about a lottery histoy file, that it is either in ascending or descending order. But hey, you stick to your SQL and have fun.
BTW, maybe you should go somewhere "Quiet" and read a dictionary. (And he said "attack the issues", seems more like personal attacks to me).