Here in MO keno draws every 4 minutes so short of having a super computer one would need to work on a method that does
not require brute force. Generating every possible combination, IMHO is out of the question. What I would suggest is to analyze
decades and ones place digits looking for patterns then assemble the numbers afterward. Not saying what your attempting is
bad, it's the time factor. To calculate the digits or numbers by position within the matrix can be done very quickly but this would
not help with finding patterns within the draw history. Loading the entire game history into arrays would speed up the processing
time and certain patterns could then be extracted.
Lets say that we dim two arrays as integers, Decades(0 to 8) and OnesPlace(0 to 9). Now lets say we load the history into a string
using something like this
Simple basic code example
File$="My games history"
OPEN File$ FOR BINARY AS #1
L = LOF(1)
temp$ = SPACE$(L)
GET #1, , temp$
CLOSE #1
We could then do something like this, simple (do-loop) to get the counts for both decades and ones place digits. One could
also use string arrays which could be mined later.
i=1
Do While i <= Len(Temp$)
Dec1 = Val(Mid$(Temp$,i,1))
OneP = Val(Mid$(Temp$,i+1,1)) This assumes zero padded with single space between numbers
Decade(Dec1)=Decade(Dec1)+1
OnesPlace(OneP)=OnesPlace(OneP)+1
i=i+3
loop
The arrays could be multi-dimensional so that the data could be stored positional. This is just some basic ideas that may help you to
speed up processing time.
Now al we need to know the number of times a digit showed is to do something like this.
Print OnesPlace(5) 'will return the number of times the digit 5 has shown, anyway these are just some simple examples that might
spark some new idea to help you in what looks like very hard task. If you keep the arrays formatted so that the data lines up then
it's a simple process to extract just about anything of use.
RL